In a world where most outer clothing was made of wool, a base layer to work as a barrier between it and the wearer’s skin was nearly a necessity. Later centuries would merge layers together, seamstresses and tailors putting linings in doublets and chauces, but in the 12th century these were distinct items of clothing.
Wool was the primary fiber used for textiles throughout medieval Europe. This product of sheep was sorted into grades which ranged from the very fine, similar to modern merino wool, to wool of more primitive breeds of sheep which yield a coarse and hairy overcoat and fine undercoat. (Crowfoot 15) All but the finest grades would require an intermediate layer between the wool and wearer’s skin. In the 12th century this took the form of a long sleeved garment with an overall length depending on the specific fashion of the region and social status of the owner. The relationship between the length of the undergarment likewise changes across the century, cultures, and class.
Undertunics may be seen occasionally in artistic depictions such as the figure to the right. In this early 12th century context, the outer layer was distinctly shorter than the under, giving us a glimpse of the latter. In the back half of the same century, the hems had maintained their relative positions, but fallen past the knees (Figure 2). This image however, is ecclesiastic in nature, being a depiction of Jacob (left) and his son Joseph (right). Such images always need to be taken with a grain of salt. These are biblical figures depicted as a period artist imaged they would be, or in a way that the period artist would expect their audience would recognize. My purposes for this image is to illustrate that the figures are wearing both an under layer and outer layer of clothing. This is a detail that I think was so self evident to the illustrator that it was included without specific consideration to the culture or region of the world being depicted.
Undertunics also appear in literary sources. According to Alexander Neckham around 1190, a full compliment of dress included “a cote or bliaut provided with sleeves and openings, slit at the crotch. Braies are needed to cover the lower limbs and stockings or chauces should be worn around the legs, while covering the feet with laced boots or leather shoes. An undershirt of muslin, silk, or cotton, or linen-the fur of the outer mantle should be gris or vair, or rabbit, or lérot, and the mantle’s edging can be of sable or martin, or beaver, or of otter, or of fox fur”(Holmes 159-160). Neckham is even kind enough to specify the materials an undertunic may be made of.
A couple sources which I am surprised do not refer to undertunics or undershirts are the Rules of St. Benedict(Benedict) and the Templars(Wojtowicz). Both refer to the amount and types of clothing approved for wear by the brothers of the orders, but curiously do not reference undergarments at all. The closest is the Rule of the Templars, or Latin Rule, which was laid down in 1129. It states “Among other things, on account of the excessive heat of the eastern region we compassionately consider that one linen shirt only be given to each brother from Easter until the Feast of All Saints; not out of obligation but from grace alone— to that one, I mean, who may wish to use it. But at other times let them generally have woolen shirts…Moreover, let them always sleep clothed in shirts and breeches. Also, let there be a lamp continually until morning for the brothers who are sleeping.”(Wojtowicz 65) The use of the word “shirt” to describe both linen and wool garments is not a feature only of the translation, but were derivatives of “camisi” in all instances.
There is one extant garment I am aware of which can be considered an undertunic and is relatively close to the culture and time period I am focused on. The St. Louis tunic is a garment now stored in the Notre Dame Cathedral Treasury, Paris dated to the middle of the 13th century. It is a fragmentary garment made of linen. Unfortunately, a detailed analysis of the garment has not been allowed due to the religious nature, so the best that can be made is estimations such as in Figure 3. The pattern can be described as a single piece body with sleeves that are wide at the shoulder and narrow toward the wrist, and front and back godets. The relative dimensions of the pieces suggest that they could have all been cut out of one width of fabric for the sake of efficiency of material. This layout is similar to the Kragelund tunic from Denmark dated to 1045-1155 CE, that garment, also having a single piece body with sleeves which are wide at the shoulder and narrow at the wrist, and godets adding fullness to the skirt.(Ostergard 195) The main difference is that Kragelund is wool, has pleats at the top of each godet, godets are added the sides as well as the front and back, and the rear godet is only split about two thirds of the way up to the top of the piece.(Figure 4)
When deciding on how I should pattern this tunic, I first needed to determine what information I could take from what sources and what material to use. The length of sleeves, length of skirt, fit, and shape of neck line I took sources which are as close to my place and time as possible. A good example of this is displayed in Liber Ad Honerem Agusti (Figure 5). This image is dated to 1194-1196 Italy and represents individuals who were contemporary to the author. To be precise, Liber ad Honerem Augusti details the events of a rebellion a couple years prior to the creation of the document in the same area that it was written and illustrated. The slinger in the bottom right of the image is wearing an outer garment which reaches his knees, full length sleeves, and a keyhole neck line.
The next set of cues I could take from extant examples. From these I was informed how the garment should be assembled; the type of seams, type and spacing of stitches, proportions of pieces, and patterning cues. The relevant garments I used were the St. Louis shirt (Figure 3) and Kragelund tunic (Figure 4). Unfortunately, I have not been able to find a detailed analysis of the stitches used for the St. Louis shirt. Most linen garments have been annihilated by time and the seams used for assembly have gone with them. We therefore have a wealth of information on how wool was put together and little to none about linen. Wool as a fabric tends to stay together, unlike linen which tends to fray. Because of this, I decided to adapt the common seam techniques detailed in Woven into the Earth (Figure 6). On the wool cloth, the seam was first sewn together with a running stitch, both sides of material folded together, and a row of overcast stitches secures the raw edges. For the linen of my undertunic, I determined most seams and hems would require a second fold of one layer of the linen before the overcast stitches to protect both raw edges, creating what is known as a flat felled seam. I do not know of direct evidence for this stitch in my period. The other seam type used was for the bottom hem. This is a low wear area, and so I determined a single fold before the overcast stitches was sufficient. The cuffs of garments are high wear areas and I decided to fold them twice. Another practice was to add a line of stitching along the edge of a neck line, possibly to stiffen the edge and prevent it from curling. In woolen garments, this was done with a series of stab stitches which passed through the medium at angles so that the stitches were very densely packed.(Figure 7) This is not practical for much thinner linen, so a very dense running stitch must suffice.
The final decision was what material to choose for the undertunic. As I already quoted Alexander Neckham in approximately 1190 CE, they could be made of “muslin, silk, or cotton, or linen.”(Holmes 160) This is a matter of ongoing research for me. I am attempting to portray an individual from the Norman Kingdom of Sicily who has re-settled in the Principality of Antioch. The question of cultural integration between such settlers in the subject of clothing and material culture is somewhat open ended for me. I have attempted to find relative prices of trade goods in the levant without success so far. A 15th century source states the price for a yard of silk as between 120 and 144 denars per yard. A silken tunic would require approximately 2.5 yards of material, making a material price between 300 and 360 denars for a silken tunic. For context, a 13th century high quality riding horse was valued at 200 denars.(Hodges) This sort of expense puts the option of silk well outside the price range of the social status I aim to portray. Muslin and cotton were simply different types of cotton cloth. In the beginning of the 14th century, cotton was a trade good produced in Syria, the Byzantine Empire, Apulia, Calabria, Sicily, and Malta at the least.(Lopez 109) Extant examples are also prevalent in the area, including finds at Jesiret Fara’un Carbon 14 dated between the 9th and 14th centuries. Cotton examples were most common with 289 pieces, linen second with 265, and there were 133 fragments with a linen warp and cotton weft, called fustian.(Boas) Because of this and the fact that cotton was produced in such close proximity to Antioch, I feel I could justify the use of cotton except for the cultural and societal norms from Europe. I am not aware of any sumptuary laws in the region besides the aforementioned Rules of religious orders, but because I am still not confident about the level of integration with the locals and adoption of dress, I am more comfortable not choosing cotton as a fiber. This leave me with linen as the medium for my undertunic.
Linen shows up infrequently in Europe due to a generally unfavorable conditions for preservation. The fragments I just referenced from Jesiret Fara’un are only counted and poorly photographed, I have found no detailed study about their weave structure whatsoever to my great frustration. A few examples of linen textiles have been preserved in Britain fortunately, giving us at least a glimpse into the weaves used. Tabby cloth is common, but also represented is “…a type of huckaback with warp floats on one face and weft floats on the other, a fabric which remains to this day popular for towels…Fragments preserved…include herringbone, lozenge and rosette twills”(Crowfoot 80-81) Unfortunately, almost all linen produced today is a simple tabby weave. For this project I was fortunate enough to find a linen 2/1 twill with both 19 warp and 19 weft threads per cm, both threads Z spun. While the specific weave does not show up in the limited extant examples I have information about, all surviving warp and weft threads are Z spun and the average thread density is about 20 threads per cm, both warp and weft.(Crowfoot 80) The main point is that it is something other than a tabby weave and allows me to depict at least a hint of the rich weaving culture present in the middle ages.
The pattern I arrived at is something of a compromise between the St. Louis tunic and Kragelund tunic.(Figures 3, 4) It is intended to go with a supertunic with a skirt which is split front and back, so I reasoned increased fullness around the bottom hem would be beneficial. An important aspect of this pattern was efficient use of material. I organized the pattern such that the side godets are triangles taken out of the fabric width, then pieced together at the sides. The front and back godets which are completely split in the front and partially split in the back were likewise organized to fit across the same width of material as the body.(Figure 8) The torso was made in a single piece, just as both the St. Louis and Kragelund tunics.(Figures 3, 4) At the initial patterning, I was unsure how high to place the split. I consider pictorial references to be not completely reliable in the matter, because the proportions can be rather exaggerated. I instead returned to the Kragelund tunic dated 1045-1155 CE (Ostergard 125) and the Moselund tunic dated 1050-1155 CE (Ostergard 136-137). Both of these garments are split up the front and rear of the skirt. While the hem line of tunic is variable according to fashion, I reckon the distance between the shoulder and the front/rear split to remain consistent in proportion to the length of the sleeve. This ratio on both garments is approximately 5/6 the length of the sleeve to the skirt split, both front and back. On my garment, applied this same proportion, placing the split close to my belt line and the final length of the garment right at the middle of my knee, the target length.
I began assembly by attaching the sleeves to the body and side godets to their places. In the pondering of piecing the garment, I had a thought that I could square off the material removed for the neckhole and use it as an underarm guesset, then use another scrap for the opposite arm. I can only guess this was some form of temporary insanity, as none of my source extant garments have underarm guessets and the sleeves are full enough at the shoulder that they are completely unnecessary. I later came to my senses and removed the guessets. Seams were constructed as I described with flat felled seams and the natural 16/2 linen thread I typically use. Sewing was done initially with a reproduction brass needle, but this was blunted through use and I switched to a reproduction steel needle.
I initially gave the body some extra length because of my uncertainty with how well I had measured myself. Certainly not a period practice, but it made costly mistakes less likely. An important part of this was the order of operations and assembly of seams to make them all lie properly. A second seam type was used for the piecing of the side godets. The sides of the seam were felled opposite directions before being secured by an overcast stitch. This seam type shows up in both Woven into the Earth(p99) and Textiles and Clothing(p156). The front and back godets were assembled with pleats, as the godets of the Kragelund tunic were. The rear was made with a box pleat with knife pleats above it, and the front has a series of knife pleats. One point of uncertainty is the termination of the split of the keyhole neck line and the top of the split skirt. I have never found a good source of information on how these cuts through the weft ended without removing material to facilitate rolling or folding back the edge. The solution I used was to finish the raw edge with a buttonhole stitch. This significantly reinforces the weak point of the pattern. The remainder of the garment was finished as I have detailed above. It has satisfied all of my expectations, and in all I think is a fair attempt at representing an undergarment suitable for a working class individual from Norman Sicily.
Benedict of Nursia, Leonard J. “St. Benedict’s Rule for Monasteries.” Translated by Leonard J Doyle, Project Gutenberg, THE LITURGICAL PRESS, 22 Sept. 2015, http://www.gutenberg.org/files/50040/50040-h/50040-h.html.
Boas, Adrian J. CRUSADER ARCHAEOLOGY: the Material Culture of the Latin East. ROUTLEDGE, 1999.
Crowfoot, Elisabeth, et al. Textiles and Clothing C.1150-C.1450. Boydell Press, 2016.
Hodges, Kenneth. “List of Price of Medieval Items.” Medieval Price List, UC Davis, medieval.ucdavis.edu/120D/Money.html.
Holmes, Urban T., and Alexander Neckham. Daily Living in the Twelfth Century, Based on the Observations of Alexander Neckham in London and Paris. University of Wisconsin, 1980.
Lopez, Robert Sabatino, and Irving Woodworth Raymond. Medieval Trade in the Mediterranean World Illustrative Documents Translated with Introduction and Notes. Oxford University Press, 1955.
Wojtowicz, Robert Thaddeus. “The Original Rule of the Knights Templar: A Translation with Introduction.” The Original Rule of the Knights Templar: A Translation with Introduction, Western Michigan University, June 1990, https://scholarworks.wmich.edu.
Østergård Else. Woven into the Earth: Textiles from Norse Greenland. Aarhus Univ. Press, 2009.