Looms and Weaving Tools

Along with food and shelter, clothing is one of the most basic human needs. In our modern society, with machinery and synthetic fabrics, clothing is almost always something that is taken for granted. Very few people in today's world make their own clothing, much less weave their own fabric, or even have the slightest idea how to approach such a project. In Roman times, however, textiles were much more a part of an average person's life. Many households would have owned a loom on which to weave, and many country families would have had sheep from which to shear wool.

The fact that weaving was a widespread and common practice does not translate into widespread evidence today for how and what Romans wove. There are several reasons for the rather sparse evidence about weaving we have left today. The first reason is the materials that were used for weaving are perishable. The looms, and often the other tools, were largely made out of wood. Wood will not survive for 2,000 years except under certain conditions. Thus, many sites that have yielded a great many artifacts made out of terra cotta, stone, or metal do not yield a wooden loom or a spindle. A second reason for the small number of artifacts that has been found is because of the actual product created from weaving. Fabric is also perishable, unless it is kept very dry, submerged in water, or frozen. The pieces of Roman fabric which remain for us to study were either not completely burned by fire or destroyed in other ways such as dampness or decay.

Keeping these limitations in mind, it is possible to recreate a great deal about the process of Roman weaving using the evidence that is left. This evidence comes from several ancient sources. The ancient evidence is present in archaeological remains of loom tools, which were often made out of terra cotta, stone or bone, as well as in the surviving pieces of cloth. There are a few remaining depictions, in the art of the Romans, of looms and weaving. The art of the Greeks, most of it done before the Roman period, also helps to fill out our picture of ancient weaving. In addition, there are several important Roman literary sources who mention or describe spinning and weaving. Gaps in this ancient evidence can tentatively be filled through the use of modern comparative evidence. This evidence can be found in the modern day equivalent of the warp-weighted loom which is still in use in parts of Scandinavia.

Roman Looms

The first subjects to examine are the looms themselves and the products associated with them. There are three main types of loom to consider when dealing with the Roman period: the warp-weighted loom, the two-beam vertical loom and the horizontal loom. The warp-weighted loom was the earliest developed and most commonly used loom in Italy and the more northern and western provinces during the Roman empire. It continued to be used in the northern and western provinces after its use had declined in the east. The type of loom known as the two-beam vertical loom was being used widely in the empire by the end of the first century A.D. (Wild 1970, p71) The horizontal loom was common, at least in the eastern provinces, by the middle of the third century A.D. (Wild 1987, p459)

The warp-weighted loom is constructed of two upright beams connected at the top by a bar called the cloth bar. The upright beams may have been between 150 and 220 cm tall, and the cloth bar between 180 and 240 cm long. This conclusion is reached by examining the sizes of the remaining Scandanavian looms, because no Roman looms survive. (Hoffmann, 24-29) From the cloth bar of a warp-weighted loom are hung the warp threads. These are the threads which run vertically. The number of warp threads on the loom depends on the desired width of the fabric to be woven. The warp threads are divided into two groups for a simple weave.  They alternate, even and odd threads.  A beam parallel to the cloth beam is called the shed rod and helps to keep the evens and odds apart. It is positioned one third of the height of the uprights from the ground. The even numbered warp threads will hang perpendicular to the ground, behind the shed rod. The odd numbered warp threads are placed in front of the shed rod, and are parallel to the uprights, which lean against a wall. (Strong and Brown, 171) The open space between the two sets of warp threads is called the natural shed.   In the diagram below, the even threads are represented by the letter A, odd threads by B.

(The above is an example of a warp-weighted loom, emphasizing the heddle rods)
These warp threads are gathered and tied at the bottom to a cord which is strung through a warp-weight. (Barber, 104) There will be a later discussion of warp-weights along with other tools used with a loom. The warp weights insure that the warp threads do not blow in the breeze or get tangled and in the wrong order. The even set of threads (in our case, thread A) is tied to another rod which is known as the heddle rod. The heddle rod is placed halfway up the height of the uprights and sits on supports, from which it can be removed.  The heddle rod can be lifted forward to bring the set of threads represented by A in front of the threads represented by B.  This creates a different shed than the natural shed. The thread that is being used to weave, called the weft or woof thread, can then be passed between the two sets of warp threads, through the shed. After this pass, the heddle rod is returned to its bracket and the even threads will again fall behind the odd threads. The weft thread is then passed back through the natural shed.  This process of passing the weft thread through the natural shed, then lifting the heddle rod and passing the weft back the other direction is repeated countless times when weaving on a warp-weighted loom.

The second major type of loom which should be examined in the context of Roman weaving is the two-beam vertical loom. (Wild 1970, p69) In many ways, it is similar to the warp-weighted loom. There are two uprights and a cloth beam at the top. There may be a heddle bar, or more than one heddle bar in order to open multiple sheds. The difference between the two looms is how the warp threads are placed on the loom. The warp threads are not anchored in groups at the bottom with weights as they are in a warp-weighted loom. Instead there is a second beam at the bottom, like the cloth beam at the top. A long piece of thread, or "one or two balls of yarn" is used to warp the loom. (Wild 1970, p70) Although it is not clear which way the Romans warped the two beam vertical loom, the method depends on the type of weave desired.

One method is to run the warp thread down the back of the loom to the lower beam. The warp thread will come up the front of the loom to a rod placed in the middle of the loom, wrap around that rod, back to the lower rod and up the back of the loom. The thread is then brought up to the cloth beam, down and around the rod in the middle of the loom and back up to the top cloth beam. (Wild 1970, p 70) This loop is repeated until the whole loom had been warped. Thus, a two-beam vertical loom is a loom in which the warp threads are under tension, not simply weighted by gravity. The two-beam loom was in use throughout the Roman empire by the first century AD, and even in part replaced the warp-weighted loom.

There are differences in the products of the two types of looms, as well as their uses. The woven pattern created by a warp-weighted loom is a different type than that created on a two-beam vertical loom. The cloth woven on a warp-weighted loom can also be considerably larger than the cloth woven on a two-beam vertical loom for two reasons. The warp-weighted loom is larger and can be used to make a wider piece of cloth. Also, the shed on a warp-weighted loom can open wider and a larger and longer "dolly" of weft is able to be passed through the shed. The two-beam vertical loom can be used as a tapestry frame, in a way the warp-weighted loom can not. Another difference between how the two looms are used is that when actually weaving, the weaver at a two beam loom stands, so she is able to pass the weft across the width of the warp. She also beats the weft up, thus weaving from top to bottom. A weaver at a two-beam vertical loom sits and weaves from bottom up.

There is also a third type of Roman loom which is a horizontal loom. It is a much more complex loom which probably originated in the area of Syria. (Wild 1987, p471) It seems to have been developed and in use by around AD 250. (Wild 1987, p459) The evidence for the horizontal loom is much scarcer than evidence for the two types of vertical looms. It is not
mentioned by any ancient Roman writer, and the first depictions date from the Middle Ages. (Wild 1970, p75) It is believed, though, that some type of horizontal loom was used in the later Roman empire because of the types of fabrics which have survived. For example, certain fabrics, made out of silk, produced after AD 250 have very complex patterns. These would have required more than the one or two sheds that were usually set up on a warp-weighted or two-beam loom. (Wild 1970, p77) It is easier to have multiple heddle rods and thus more sheds, on a horizontal loom. These horizontal looms, even during the late empire, would not have been as widely used as vertical looms. Only skilled weavers would have known how to use them, and the fabric woven on horizontal looms would have been too expensive for many people to afford.

Loom Tools

The warp-weights themselves are the most commonly found objects associated with weaving. Those that were created out of terra cotta or stone have survived conditions that destroyed the dried mud warp-weights and the wooden looms. Warp-weights are usually shaped as pyramids, or spheres, with a hole cut through the weight. This hole allows a cord to be strung through which attaches the weight to the warp threads. Warp-weights have been found at many Roman sites, both in Italy and the provinces. Weights found which date from the early imperial period at the site of Herculaneum were made out of baked clay. (Wild 1970, p62)  Warp-weights have been found at many sites in Italy, for example, from the Iron Age site called Monte Loffa. (Barber, 389) They have also been found at numerous sites outside of Italy. Very few warp-weights have been found still in place because looms were set up only when someone needed to weave and taken down when the fabric was finished. (Hoffmann, p31) In order for warp-weights to have been found in place, the loom would have had to be destroyed, most likely by fire, while it was in use.

Other tools associated with weaving have also been found at various archaeological sites. Weaving combs, which were made out of bone during the Iron Age, were made out of wood by the time of the Romans. (Strong and Brown, p171) This tool is used to push the weft threads up, to create a tighter weave. Another tool for this purpose is called a weaving sword, or spatha. (Strong and Brown, 171) It is a flat, long tool that was put in the shed and pushed the weft up. Pin beaters, which are also used to push up the weft and untangle knots, are shaped roughly like a cigar. (Hoffmann, p140 fig 61) Pin beaters are not slipped into the shed to push the weft up, as weaving swords are, but rather pushed between warp threads, or pulled along in front of the weft. (Wild 1970, p66)

One tool, not known to have been used in Roman weaving, that is now commonly associated with weaving is a shuttle. The weft thread was instead passed through the shed using a weft bobbin. This is simply a bundle of thread with a "firm head." The "firm head" is formed by wrapping the thread around the weaver's hand and then looping it around itself to create a tightly bound, rather pointed ball of thread. A weft bobin is easier to pass through the shed than a simple ball of thread, or a loose clump would be. (Hoffmann, p66-67) If the width of the warp was too wide for the weaver to reach across it, this weft bobbin, which can also be called a "ball, hank, or 'dolly' " can be tossed from one side of the loom to the other. (Strong and Brown, p171)

Ancient Depictions and Literary Sources

In addition to the actual remains of the art of weaving, there are depictions of looms and weaving which can be studied. There is one depiction of a warp-weighted loom from a gravestone found at a site in Spain known as Lara de los Infantes and dating from the first or second century A.D. (Wild 1987, p 460) There are no other pictorial representations of warp-weighted looms which have been found from the Roman period. Earlier Greek depictions, though, particularly those on vases from the areas of Attica or Boeotia are helpful when studying warp-weighted looms. There are several fifth-century B.C. vases which show a warp-weighted loom set up and being woven upon. A particularly well preserved vase shows Telemachus and Penelope in front of Penelope's loom. This representation has several extra horizontal rods pictured, which have been the basis for some questioning. The beam higher up than the cloth beam seems to be used to hold extra weft thread. The beam below the cloth is thought to be an unrealistic feature while the two in the middle may be multiple heddles, or a heddle and a shed rod, placed higher than many others. (Hoffmann, p305-306) A lekythos, a vase for storing oil or perfume, from the sixth century, also shows a weaving scene and the warp-weighted loom clearly. There are two women standing in front of the loom who are weaving on it; the one on the left may be beating up the weft and the figure on the right may be opening the shed with a heddle rod. (Hoffmann, p306-310)

There are also a number of depictions of two-beam looms that still exist today. The gravestone of a Roman woman named Severa Seleuciana, from the third century AD, has a two-beam loom pictured on it. The forum of Nerva in Rome, from the late first century AD, depicts the weaving contest between Arachne and the goddess Minerva in stone relief and shows three two-beam looms. (Wild 1970, p 69) There is also a wall painting in the hypogeum of the Aurelii in Rome showing a seated male figure, a standing female figure, and a partially woven two-beam loom. (Wild 1970 plate XIb) A later illustration "in the Vatican Virgil manuscript" (Wild 1987, 460) also shows a loom with a woman standing in front of it, facing away from the loom. (Hoffmann, p 326)

In addition to artistic representations, there are written descriptions of looms and weaving. Ovid's Metamorphoses describes in poetry the weaving scene between Minerva and Arachne that is shown on the Forum of Nerva. (Ovid, Book VI, lines 53-58) In Virgil's epic poem The Aeneid, there are a number of comparisons to weaving and spinning. (Scheid and Svenbro, 43) The most complex are the comparisons of the movements of the Trojan boys at the funeral games for Anchises to weaving. There are several other Roman sources that discuss looms and weaving. "The Edict of Diocletian on Maximum Prices" includes a listing of the types of looms and loom parts that were available for purchase in the early fourth century A.D.  (Wild 1987, 465) It shows the widespread use of the two-beam loom by the time of Diocletian. The work of Seneca, who wrote in the first century A.D., has valuable information regarding the features of a warp-weighted loom and helps to date the emergence of the two-beam loom.(Wild 1970, 67)  Seneca writes that weaving on a two-beam loom involved a "more sophisticated technique," implying earlier warp-weighted looms were considered old fashioned by the time Seneca was writing. (Seneca, Letter 90, 20) These warp-weighted looms were being "rapidly displaced by the two-beam loom in Italy and the Mediterranean provinces." (Wild 1970, p67)

Comparative Evidence

There is also comparative evidence which aids us as we seek to know about Roman weaving. Today in parts of Scandinavia, there are still looms which are almost identical in appearance and function to the warp-weighted looms used by the Romans. These looms have helped to determine how Roman looms looked, in terms of materials needed to build the looms, as well as the size of an actual warp-weighted loom. They show that such looms can be roughly put together and do not need sophisticated carpentry. These looms also demonstrate which weave patterns could and could not have been done on the warp-weighted loom. Marta Hoffmann's book, The Warp Weighted Loom, extensively describes the process of warping, and weaving on the loom, as well as documenting all the surviving warp-weighted looms still in Scandanavian museums.


Works Cited

Barber, E.J. W. Prehistoric Textiles. Princeton: Princeton University Press, 1991.

Hoffmann, Marta The Warp-Weighted Loom. Oslo: Universitetsforlaget, 1964.

Ovid. Metamorphoses. Trans. Mary M. Innes. London. Penguin, 1995.

Scheid, John & Svenbro, Jesper. The Craft of Zeus. Trans. Carol Volk. Ed. G.W. Bowersock.

    Cambridge: Harvard University Press, 1996.

Seneca. Seneca 17 Letters. Trans. Costa, C.D.N. Warminster: Aris & Phillips Ltd, 1988.

Strong, D. and Brown D. Roman Crafts. New York, 1976.

Wild, John Peter. "The Roman Horizontal Loom." The American Journal of Archaeology 91.

    1987: 459-471.

Wild, J. P. Textile Manufacture in the Northern Roman Provinces. Cambridge Classical Studies.

    London: Cambridge University Press, 1970.


Please direct any Comments, questions, or suggestions to Kathy Laxton: klaxton@email.unc.edu
All material on this website copyright Kathy Laxton 1997
Graphics for this page were taken from: Infinite Fish
Best Viewed in Netscape