In the frequently damp environment of early medieval Britain and western Europe sheaths were essential for valuable ferrous blades, including swords, war-seaxes and smaller utility knives. The sophisticated and labour-intensive construction of sword scabbards, formed of animal-hair lined carved wooden plates enclosed in leather or hide (“skin product”) is evidence of the priority given to protecting blades. Well preserved examples of seax sheaths, mainly from productive leatherworking sites of the mid-to-late Anglo-Saxon period, at York and Gloucester (Cameron, 2000) as well as more trace evidence from earlier grave finds, demonstrates the high level of craftsmanship which went into these, usually of a single piece of folded and moulded skin-product, sometimes intricately decorated, as might be expected given that the sheath, and not the blade, was the most visible component of any seax assemblage day to day and therefore an important surface for wealth display. Vastly more common but nevertheless valuable personal possessions and often of sophisticated smithcraft, smaller utility and eating-knives also required tight-fitting sheaths of valuable skin-product, but despite the abundance of such knives in the archaeological record, for a wide range of reasons clues about the sheaths of smaller knives are scarcer than for larger seaxes.
It might be reasonable, and has been the practice among more detail-oriented living historians and reenactors, to assume that sheaths of smaller knives in, for example, the early Anglo-Saxon period, might be miniature analogues of those of larger seaxes, yet fragmentary sheath remains from a handful of well-studied early Anglo-Saxon sites appear surprisingly crude, with amateurish stitching having unattractively contorted the seams in a way that might disgust a modern leatherworker. There is no reason not to think these examples are not well representative, and indeed, many of the later (Viking-Age) knife-sheath remains from York, though often skilfully decorated, bear the tell-tale marks and contortions of this same rudimentary stitch work.
It is always tempting to base the sheaths of our knives on the very fanciest, and neatest archaeological examples to hand, and perhaps neaten them up with some more modern handiwork, but this can lead to a creeping departure from what is truly known of the historic craft culture purportedly represented. In contrast, replicating (to our modern eyes) “unbecoming” examples might provide useful, practical insights into why they were made this way. Such is the case with these apparently crudely stitched knife sheaths – our experiments in replicating them have revealed what might be a cunning Dark Age leatherworker’s “life-hack” which made the tricky shaping of sheaths vastly quicker and more reliable.
It might be reasonable, and has been the practice among more detail-oriented living historians and reenactors, to assume that sheaths of smaller knives in, for example, the early Anglo-Saxon period, might be miniature analogues of those of larger seaxes, yet fragmentary sheath remains from a handful of well-studied early Anglo-Saxon sites appear surprisingly crude, with amateurish stitching having unattractively contorted the seams in a way that might disgust a modern leatherworker. There is no reason not to think these examples are not well representative, and indeed, many of the later (Viking-Age) knife-sheath remains from York, though often skilfully decorated, bear the tell-tale marks and contortions of this same rudimentary stitch work.
It is always tempting to base the sheaths of our knives on the very fanciest, and neatest archaeological examples to hand, and perhaps neaten them up with some more modern handiwork, but this can lead to a creeping departure from what is truly known of the historic craft culture purportedly represented. In contrast, replicating (to our modern eyes) “unbecoming” examples might provide useful, practical insights into why they were made this way. Such is the case with these apparently crudely stitched knife sheaths – our experiments in replicating them have revealed what might be a cunning Dark Age leatherworker’s “life-hack” which made the tricky shaping of sheaths vastly quicker and more reliable.
The scarcity of evidence to inform our understanding of knife sheaths is due, in part, to the smaller size of the blades providing little surface for the preservation of sheath traces, the use of thinner skin-product, and the “commonness” of small knives leading to them being overlooked, historically, by archaeologists and conservators. These challenges greatly impacted Esther Cameron’s review of Sheaths and Scabbards in England AD400-1100 (2000) – arguably the most useful work on this subject – which despite great effort has far less to say on the sheaths of smaller knives than of seaxes and swords. The most abundant evidence comes from her own study of the knives from the 6th century Anglo-Saxon cemetery at Snape, which were of 1-1.5mm thick skin product folded, as we would expect, about the back of the blade and stitched along the blade edge. Rudimentary linear tooled decoration on some sheaths from Snape, Dover Buckland, and Sutton Hoo, are consistent with insights from larger seax sheaths, though of course simpler due to the limited space for such details. One of the main stand-out features of many of this still admittedly small sample are the “sinuous curves” of the sheath along the blade edge, concertinaing as a result of “loose serpentine stitch” using leather thong or in some cases, possibly other stronger fibres. It is this stitching – also evidenced on some of the Coppergate leathers, which is puzzling.
Sheath remains from Snape, Suffolk (left) and Broomfield, Essex (right) both showing widely spaced "serpentine" running stitch, and resulting concertinaed / wavey seam edge. (Cameron, 2000)
It is desirable that closure of the seam along the blade edge be extremely tight, neat and secure, to prevent ingress of moisture and so that the clenching of the front and back faces of the sheath together actually protects the stitches inside the sheath from being cut by the sharp blade edge. To achieve this, any skilled modern leatherworker might be expected to resort to the trusty saddle-stitch – effectively two running-stitch “snakes” passing through the layers alternately, creating a continuous line of stitches visible on both sides, tightened throughout, and locking each-other. It is precisely this stitch which was predominantly used for the sewing of shoes in the middle to late Anglo-Saxon period / Viking Age, as evidenced by abundant, well studied remains from York (Mould et. al. 2004). As stitches on both sides are pulled tight simultaneously there is no contortion of the leather – the tension of the stitches manifests only in the tightness of the seam, and the plane of the leather remains beautifully flat.
Instead, many Anglo-Saxon and Anglo-Viking knife sheaths (edit: including knives from the 6th century princely burial and cemetery at Prittlewell, Essex, see addendum below) appear to have been stitched crudely with a single running-stitch – a snake which burrows onto the back, then onto the front, then onto the back again, creating an interrupted and alternating stitches on both sides. Weaker and unlocked, tensioning of these stitches causes the leather to concertina, forming the wavy edge seen on the archaeological sheath remains. Why would skilled Anglo-Saxon leather-workers do this?
Instead, many Anglo-Saxon and Anglo-Viking knife sheaths (edit: including knives from the 6th century princely burial and cemetery at Prittlewell, Essex, see addendum below) appear to have been stitched crudely with a single running-stitch – a snake which burrows onto the back, then onto the front, then onto the back again, creating an interrupted and alternating stitches on both sides. Weaker and unlocked, tensioning of these stitches causes the leather to concertina, forming the wavy edge seen on the archaeological sheath remains. Why would skilled Anglo-Saxon leather-workers do this?
Fragments of an Anglo-Scandinavian sheath from York showing crude, widely spaced running-stitching (Groenman et. al. 2004)
As the preserved leathers are often in various ways shrunk and contorted it is not unreasonable that observers might dismiss the wavy edge of individual knife sheaths as a result of degradation, yet the feature is common across sheath remains from a variety of contexts. Though Cameron (2000) describes this as a “loose” serpentine stitch, in fact, for it to cause the observed wave-pattern contortion of the leather edge, the stitching must have been deliberately pulled quite tight - certainly tighter than the leather beneath, leading to the observable constriction and undulations. That the tension of thread and layers being sewn, should be matched, is one of the most basic and intuitively obvious rules of sewing, which if ignored can lead to unsightly bunching (though over-tensioning stitches provides the basis for pleating – such as is evidenced for some Viking Age under-dresses (Grömer, 2013) and drawstrings/laces, such as those used to fasten shoes (as abundantly evidenced among the York remains; Mould et. al. 2004). These problems are always a risk with running stitch, yet the makers of these sheaths appear to have taken no care at all in carefully managing the tension and mitigating the risk of unsightly bunching.
Quite the opposite - this almost corrugated outcome appears to have been a deliberate choice. In fact, the widely spaced running-stitch and wavy contorted edge evidenced on Anglo-Saxon knife-sheath remains might not represent sloppy, over-tensioned seam-closure, but rather, might have been a crucial feature of the design and manufacture of such sheaths, overcoming a number of technical challenges and greatly improving production reliability and efficiency.
Quite the opposite - this almost corrugated outcome appears to have been a deliberate choice. In fact, the widely spaced running-stitch and wavy contorted edge evidenced on Anglo-Saxon knife-sheath remains might not represent sloppy, over-tensioned seam-closure, but rather, might have been a crucial feature of the design and manufacture of such sheaths, overcoming a number of technical challenges and greatly improving production reliability and efficiency.
Anglo-Saxon and Viking knives came in a variety of shapes (Blakelock, 2013), and anyone who has studied sheaths and/or attempted to make one quickly becomes aware of the technical challenges these shapes present. An Anglo-Saxon sheath MUST be folded along the back of the blade all the way to the tip, but as these backs are often not straight (curving to a spear-like point, or changing angle sharply in the famous “broken back” style) the leather fold must, itself, curve to match. A curving fold, without cutting or stitching, requires the squashing and/or stretching of material, and thankfully, rawhide and/or traditionally tanned leathers are mouldable when wet, approximately retaining their new shape when dry. Modern (chromate or oil-tanned) leathers lack this ability, but even with traditional materials this procedure is tricky. Tannic acid, liberated from leather by the water, is staining and slowly burns the workers skin, and dissolves other substances (especially metal deposits!) it comes in contact with, which can in turn stain the leather.
Reproduction small utility knife and sheath (Æd Thompson) inspired by finds from Viking Age York. Made from 2mm veg-tan leather; upward-curving fold, matching the "broken back" shape of the blade, was achieved by wet-moulding with a wooden former / blade facsimile. Fine stitching with a dense, conventional two-needle saddle stitch, was undertaken after shaping.
For this reason, moulding the sheath onto the blade itself is unwise unless the blade is protected in some way – a common reenactor “hack” is to use cling-film / saran-wrap and grease, though the former was obviously not available to our ancestors, and the latter will, of course, absorb into and stain the leather too. Historically, quickly carved wooden facsimiles of the blade may have been used to mould the sheath (Mould et. a. 2004) which could then be quickly stitched or pinned (through the stitch or rivet-holes “to be”) along the “blade edge” of the wooden former, to allow the leather to shrink around it as it dries, and acquire the shape. Clamping the leather along the edge works too.
Nevertheless, this process is fiddly, messy and can be frustrating. Even despite the moulding characteristics of the leather, it still is inclined to return to its natural shape, and so the intended curvature of the spine fold will often reduce as it fully dries, is worked further, and ages. On smaller sheaths, made of thinner (and therefore less 3D-mouldable) leather, the technique is proportionally more difficult, and results more badly affected by its tendency to “return”.
Aware of this, it is possible that Anglo-Saxon and Viking leatherworkers achieved the curving fold on the spines of their sheaths with an entirely different technique, which is faster, easier, more reliable, and does not require wetting of the leather at all, and that the crude, concertina-inducing running stitch was an instrumental part of this approach.
In fact, it appears that the curving fold of knife sheaths could have been achieved entirely dry, with no need for moulding, using the following procedure; the leather is folded around the knife, such that the fold is, naturally, straight, and stitch holes are positioned along the blade edge not straight, as they are intended to be in the final product, but rather, in an arc which curves downward to the blade tip where it meets the fold. With the folded back of the sheath straight, and the stitch-line curving, this is the inverse of the final form. The chord is secured at the tip of the sheath, threaded in a serpentine path through the holes along the edge of the sheath forming a crude running-stitch, and pulled excessively tight at every hole, with the result that the path of the cord is shorter than the stitch line on the leather, and must necessarily straighten, while the leather concertinas along it.
Nevertheless, this process is fiddly, messy and can be frustrating. Even despite the moulding characteristics of the leather, it still is inclined to return to its natural shape, and so the intended curvature of the spine fold will often reduce as it fully dries, is worked further, and ages. On smaller sheaths, made of thinner (and therefore less 3D-mouldable) leather, the technique is proportionally more difficult, and results more badly affected by its tendency to “return”.
Aware of this, it is possible that Anglo-Saxon and Viking leatherworkers achieved the curving fold on the spines of their sheaths with an entirely different technique, which is faster, easier, more reliable, and does not require wetting of the leather at all, and that the crude, concertina-inducing running stitch was an instrumental part of this approach.
In fact, it appears that the curving fold of knife sheaths could have been achieved entirely dry, with no need for moulding, using the following procedure; the leather is folded around the knife, such that the fold is, naturally, straight, and stitch holes are positioned along the blade edge not straight, as they are intended to be in the final product, but rather, in an arc which curves downward to the blade tip where it meets the fold. With the folded back of the sheath straight, and the stitch-line curving, this is the inverse of the final form. The chord is secured at the tip of the sheath, threaded in a serpentine path through the holes along the edge of the sheath forming a crude running-stitch, and pulled excessively tight at every hole, with the result that the path of the cord is shorter than the stitch line on the leather, and must necessarily straighten, while the leather concertinas along it.
Sheath (1mm veg-tan leather, and crude running-stitch) by Æd, shown in progress. Beginning as simply a flat rectangle, folded with a straight fold, and a downward curving line of stitch holes. You can see the pulling of the stitches tight, causing serpentine bunching of leather along the stitch line, has begun to pull the fold line upward, creating the characteristic back curve.
The tension in the stitching, as well as concertinaing the edge of the sheath, pulls its tip upward and into line with the rest of the stitching, forcing the folded back of the sheath to stretch and curve upward, resulting in a sheath which matches the shape of the blade. The line of stitching thus not only serves as a means of closure, but also as a “draw-string” to pull the sheath into the desired shape. The result, though perhaps less neat and attractive than more sophisticated stitching methods, makes the characteristic back-curve to tightly enclose the blade achievable in a matter of minutes rather than hours, and in a way which is reliable, with little risk of failure or staining which might waste valuable materials. Interestingly, the concertina effect along the stitched edge also appears to prevent the blade edge from cutting the thread, when the blade is inserted, as the edge will be blocked by the undulations of leather flesh inside before encountering the stitches. Tension in the cord, and the corrugation of the edge, also appears to contribute additional stiffness to the sheath - much needed when constructed of extremely thin skin-product.
Reproduction knife sheath based on 6th century traces from sites including Snape, Suffolk, and Broomfield, Essex. This is the same sheath shown above during stitching; the drawing of the stitches have pulled the stitch-line straight, and forced the back-fold to curve. Thanks to member Beth Woolman
One challenge to this interpretation is that it relies substantially on the tensile strength of the chord, both during manufacture and afterwards. Strong cordage such as waxed linen thread, sinew or even horse-hair would be obvious candidates, and as threads typically do not survive among the leather traces, which typically bear empty stitch-holes and their legacy marks, these are certainly plausible. However, a number of cases (including some of the 6th century knife sheaths from Snape) were sewn with very fine leather thong. In practice single-strand veg-tan leather thong of a gauge suitable for sewing is rather prone to breaking under tension, and effectively cannot be used with a needle. Stiffening the tip is necessary to allow it to be threaded directly through awled holes, perhaps using an organic glue, and it is plausible that doing so with an entire length of thong might give it the required tensile strength. Greater tensile strength of leather thong for sewing might also be provided by alum-tawed or dressed, rather than vegetable tanned leather, as suggested by Mould et. al. (2004). Strain on the thread can be reduced further by awling the holes not perpendicular to the plane of the leather, but rather, at the oblique angles which the serpentine path of the chord would “prefer” to take, thus reducing friction and allowing it to “draw” more readily without breaking.
Interestingly Cameron (2000) reports stitching associated with the (thicker leather) early Anglo-Saxon knife sheaths from Broomfield, Essex, as a sinuous “tunnel stitch” – a technique where the thread takes a curving path through a “tunnel” in the flesh of a thicker piece of leather and back out the same side, rather than passing right through, so that it is invisible on the outer face. In practice such stitching is impossible with thinner leathers, which lack enough depth of flesh through which to tunnel, but suggests that knife sheaths were indeed stitched through holes awled at oblique, rather than perpendicular angles, as described above.
In summary the crude running stitch and concertinaing effect observable on many Anglo-Saxon and Viking-Age knife sheaths might have served not only as a means of closure, but to pull the sheath into its curving shape. Rather than dealing with shaping and stitching in a conventional sequential fashion, this approach – vastly quicker, more reliable and more practical than wet moulding, especially at smaller scales - would require both the shaping and closure of sheaths to be planned and addressed concurrently. To what extent sheaths of larger seaxes were manufactured in this way remains to be seen, though various c8-10th sheath leathers from York and Gloucester do not appear to feature the characteristic wavy edge – a great many of them bear the holes for (non-extant) rivets which are also a not uncommon feature of early Anglo-Saxon seax sheaths. Although there is precedent for both tunnel-stitching and rivets used in combination for seax sheath closure (eg. Buttermarket, Ipswich, g3243), rivets serve only as an alternative means of closure and cannot have contributed to forming and maintaining the shape of the sheath. It might be that larger sheaths, therefore, were shaped using the more familiar wet-moulding method, perhaps using the rivet holes to temporarily fix the leather around the former, which is proportionally easier and yields better results with larger sheaths and thicker skin product.
Interestingly Cameron (2000) reports stitching associated with the (thicker leather) early Anglo-Saxon knife sheaths from Broomfield, Essex, as a sinuous “tunnel stitch” – a technique where the thread takes a curving path through a “tunnel” in the flesh of a thicker piece of leather and back out the same side, rather than passing right through, so that it is invisible on the outer face. In practice such stitching is impossible with thinner leathers, which lack enough depth of flesh through which to tunnel, but suggests that knife sheaths were indeed stitched through holes awled at oblique, rather than perpendicular angles, as described above.
In summary the crude running stitch and concertinaing effect observable on many Anglo-Saxon and Viking-Age knife sheaths might have served not only as a means of closure, but to pull the sheath into its curving shape. Rather than dealing with shaping and stitching in a conventional sequential fashion, this approach – vastly quicker, more reliable and more practical than wet moulding, especially at smaller scales - would require both the shaping and closure of sheaths to be planned and addressed concurrently. To what extent sheaths of larger seaxes were manufactured in this way remains to be seen, though various c8-10th sheath leathers from York and Gloucester do not appear to feature the characteristic wavy edge – a great many of them bear the holes for (non-extant) rivets which are also a not uncommon feature of early Anglo-Saxon seax sheaths. Although there is precedent for both tunnel-stitching and rivets used in combination for seax sheath closure (eg. Buttermarket, Ipswich, g3243), rivets serve only as an alternative means of closure and cannot have contributed to forming and maintaining the shape of the sheath. It might be that larger sheaths, therefore, were shaped using the more familiar wet-moulding method, perhaps using the rivet holes to temporarily fix the leather around the former, which is proportionally easier and yields better results with larger sheaths and thicker skin product.
Two reproduction Anglo-Saxon small knife sheaths based on the same finds - above, using the technqiue described, and below, an earlier version that was made conventionally - wet-moulded first, and then sewn with a "loose serpentine" running stitch with leather thong, not contorting the sheath edge. Property of members Beth Woolman and Æd Thompson respectively
Of course, I do not rule out the possibility that this interpretation might have already ocurred to others, might seem obvious, or that many analogues might exist in other historical or modern leatherworking traditions. Nevertheless, as this interpretation and method, which I have found experimentally to be highly effective, quick and reliable, appears to have not hitherto been discussed in any of the main publications concerning Anglo-Saxon and Viking sheath archaeology, I offer it here in the hope that it might plug a gap in the understanding of this leatherwork, and provide some assistance to anyone attempting the sometimes frustrating task of sheathing Anglo-Saxon and Viking knives.
References
Cameron, E.A., 2000. Sheaths and scabbards in England AD400-1100. Archaeopress.
Mould, Q., Carlisle, I. and Cameron, E. 2004. Leather and Leatherworking in Anglo-Scandinavian and Medieval York.
Blakelock, E.S., 2013. The Early Medieval Cutting Edge of Technology: An archaeometallurgical, technological and social study of the manufacture and use of Anglo-Saxon and Viking iron knives, and their contribution to the early medieval iron economy (Doctoral dissertation, University of Bradford).
Grömer, K. and Rast-Eicher, A., 2019. To pleat or not to pleat–an early history of creating three-dimensional linear textile structures. Annalen des Naturhistorischen Museums in Wien. Serie A für Mineralogie und Petrographie, Geologie und Paläontologie, Anthropologie und Prähistorie, 121, pp.83-112
References
Cameron, E.A., 2000. Sheaths and scabbards in England AD400-1100. Archaeopress.
Mould, Q., Carlisle, I. and Cameron, E. 2004. Leather and Leatherworking in Anglo-Scandinavian and Medieval York.
Blakelock, E.S., 2013. The Early Medieval Cutting Edge of Technology: An archaeometallurgical, technological and social study of the manufacture and use of Anglo-Saxon and Viking iron knives, and their contribution to the early medieval iron economy (Doctoral dissertation, University of Bradford).
Grömer, K. and Rast-Eicher, A., 2019. To pleat or not to pleat–an early history of creating three-dimensional linear textile structures. Annalen des Naturhistorischen Museums in Wien. Serie A für Mineralogie und Petrographie, Geologie und Paläontologie, Anthropologie und Prähistorie, 121, pp.83-112
Addendum: Only a few days after first sharing this article, while looking through the excellent report "The Prittlewell Princely Burial; Excavations at Priory Crescent, Southend-on-Sea, Essex, 2003" (MOLA, 2019) by Lyn Blackmore, Ian Blair, Sue Hirst and Christopher Scull, we came across discussion of the knife from the late 6th century princely burial, and also of the knives from the small cemetery of 6th century high-status burials nearby. Unlike most knives from early AS cemetery contexts, these have been subject to thorough analysis, from which clues about sheath technology can be gleaned. The knife from the princely burial, "largely complete but broken" into five fragments corroded into three layers, with a blade that would have totalled 138mm in length, had traces of a horn handle and remains of a thin skin or leather sheath that extended over both the blade and handle, and was seamed along the cutting edge, though no stitching is extant to examine. This is entirely consistent with expectations for an early Anglo-Saxon knife and sheath based particularly on insights from Broomfield and Snape. The knife from grave 121 (excavated, unlike most of the Prittlewell cemetery not during the Pollitt excavations of the 1920s, but rather, as part of Trench 2 of the 2003 work) found with remains of a patternwelded sword, a shield and spear, was also fragmentary but had sufficient organic remains for Esther Cameron to determine that it, again, had a horn handle and was sheathed with thin calfskin seamed along the cutting edge with running stitch, and which also enclosed part of the handle, in a way wholly consistent with finds from Broomfield, Snape, and Anglo-Scandinavian 10th century York.
At the time of writing, the Prittlewell Report represents not just a valuable treasure-trove of Early Anglo-Saxon archaeological insights but also the most comprehensive up-to-date reviews of early AS material culture, and with respect to the sheaths of smaller knives provides a useful extra couple of datapoints, which, when combined with others, point to running-stitched calf-skin being the norm. It is worth emphasising that these knives come from elite burials - we have no reason to suppose that they are "cheap" or "shoddy" work. Rather, it seems that, even for elite knifes, running-stitched calf-skin was the sheathing approach of choice, for positive reasons and not for necessity, lack of materials or skill. This seems to support the hypothesis we offer here - that there are real, non-obvious practical benefits to this approach. "It works".
At the time of writing, the Prittlewell Report represents not just a valuable treasure-trove of Early Anglo-Saxon archaeological insights but also the most comprehensive up-to-date reviews of early AS material culture, and with respect to the sheaths of smaller knives provides a useful extra couple of datapoints, which, when combined with others, point to running-stitched calf-skin being the norm. It is worth emphasising that these knives come from elite burials - we have no reason to suppose that they are "cheap" or "shoddy" work. Rather, it seems that, even for elite knifes, running-stitched calf-skin was the sheathing approach of choice, for positive reasons and not for necessity, lack of materials or skill. This seems to support the hypothesis we offer here - that there are real, non-obvious practical benefits to this approach. "It works".
RSS Feed