Over 20 years on from the publication of this still critically important work, this observation has gradually exerted influence on some modern impressions of warriors from the period, and even beyond, given the limited evidence for late Anglo-Saxon shields, and limited availability of information on 'Viking' ones. It is further, not uncommon to hear, repeated by respected historians, the assertion that most early Anglo-Saxon shields were “little more than bucklers”. To what extent is this statement accurate? Just how small were Anglo-Saxon shields?
(Article originally published in March 2016, Thegns of Mercia blog. Updated March 2021)
The 9th-10th century Middleton “warrior” cross depicting a figure in funereal display shows a shield barely half the size of the figure's head, while depictions of shields on pressblech foils from the Sutton-Hoo mound 1 helmet, and their Vendel-culture parallels are more variable.
The fact that later depictions of shields show them considerably larger need not invalidate the notion that early shields were small. Indeed, the archaeological evidence hints that shields grew, gradually, towards the end of the furnished burial rite in the 7th century – a trend which, if continued, would be consistent with considerably large shields by the end of the 8th (Dickinson & Härke, 1992).
What unites these earlier depictions of shields is the need to constrain proportions of certain items to fit into a particular frame. A shield of any practical size, designed to “cover the body”, will naturally obscure details one might wish to show on a portrait, be it in silver foil or in stone. In the case of the Repton Stone and many of the pressblech foils, shields are held aloft by the figures depicted precisely for the purpose of occupying empty space above the shoulder, and allowing less of the warriors' bodies to be obscured, exposing more visually interesting details such as belts, armour, and weapons. It's highly likely, too, that shield depictions were sized with this aim in mind.
Judging shield size from degraded archaeological remains is fraught with difficulty; soil stains – not always present or reliable, can give an estimate of probable diameter (Dickinson and Härke, 1992). Maximum diameter can be inferred from the position of any edge-fittings, although these fittings are extremely rare anyway, and can shift. Alternatively, maximum shield diameter can be inferred from position of the shield boss relative to the edge of the grave cut; this depends on the cut having a definable edge, and depends on the boss' position not having shifted. Unfortunately the rarity of such circumstances make probaable, or maximum diameter impossible to infer in most cases.
Minimum shield diameter is easier to infer; the length of a grip, or position of any board fittings, can be used. The problem is the validity of minimum shield diameter estimates; as I. P. Stephenson (2002) remarks; “Determining the minimum diameter provides no information on the actual size of the shield and, as a result, no meaningful conclusions as to how the shield was used can be drawn from such a calculation”.
The reason for this, is that 'short' and 'medium' iron grip remains (Dickinson types I and II) could not have extended across the full diameter of the shield board – the former typically being no longer than the diameter of the boss, and the latter barely being twice that length. Even 'long' grips (Dickinson type III) should be treated with suspicion; the one case of a long-gripped (albeit rather atypical) Anglo-Saxon shield for which the board diameter is conclusively known – the shield from Sutton Hoo mound 1, had a board which clearly extended well beyond the ends of it's elaborately embellished grip. Long gripped, elaborately decorated shields of the contemporaneous Vendel culture of Sweden, again with metal rims to provide the actual board diameter (varying from 0.84 to 1.1m) show a similar pattern, with the long grip spanning approximately the middle two thirds of the board.
With Dickinson and Härke's minimum diameter data from their national sample (102 cases), most were between 35cm and 45cm (69 cases), with some finds as small as 20cm, but none larger than 60cm. However, if the minimum diameter data is discarded and only more reliable and informative 'maximum diameter' estimates are used, as suggested by Stephenson (2002), the vast majority of shields from early Anglo-Saxon burials (68% of 112 cases) are between 49cm and 73cm. Of the remaining 32%, most are larger than this range – up to 92cm, while relatively few are smaller; between 32 and 45cm. Crucially, shields of a size-range which seemed to be overwhelmingly most common when inferring from the minimum diameter data (the so-called “small shield”, 34-42cm) become little more than a rare oddity when one switches to using more reliable, maximum diameter data.
One possibility; that early Anglo-Saxon warriors were simply much smaller than us, can be dismissed. As previously discussed (click here), we can tell from skeletal remains that weapon-bearing Anglo-Saxons often achieved similar heights to modern men, probably thanks to nutritious and reliable diets during upbringing, made possible by their membership of an advantaged elite. Strong bones with substantial muscle attachments show many warriors were quite muscular. Although no doubt substantially leaner, an Anglo-Saxon warrior would have, broadly, as much 'body' to protect with his shield as would a typical modern man.
It is not impossible that smaller shields served some kind of purpose in looser, skirmishing combat. A small shield has a number of advantages; certainly cheaper and easier to manufacture, requiring fewer planks, it also has fewer plank-joins, therefore potentially being stronger. All shields trade off protection against weight, and a smaller shield could allow a warrior to be more manoeuvrable in loose combat, and could conceivably be made thicker and therefore more reliable in battle, while still being light enough to not exhaust the warrior using it. Dickinson and Härke note, however, that as shields grow larger over time, they also grow thicker. There would seem to be no evidence that smaller shields were indeed built more robustly; the opposite seems to be true. It is, further, worth repeating that a small shield can protect little of the body at any given moment, and certainly cannot be interlocked with neighbours to form a shield-wall. While it is conceivable that a small shield could be used in a dynamic, 'point-defence' fashion, even for the well trained, such protection would surely be more taxing for the user while also introducing a greater element of chance. Why not just use a larger shield?
The ideal diameter for a shield is at least equal to the width of a body, but arguably would be wider; this provides that more of the body can be protected while static, and allows for room for the edges of boards to be interlocked in a shield-wall (if this practice was indeed as dominant in the early period as is often assumed). Yet precisely the opposite is desirable where a shield is being buried in a grave. Here, a shield would ideally be no wider than the body, so that a grave-cut need not be enlarged to accommodate it. Inclusion of a 90cm shield would increase the effort needed to dig a typical grave-cut by at least a third, and it may be no coincidence that the median shield diameter in the Dickinson and Härke sample – around 60cm, is very close to the minimum width of a grave-cut necessary to accommodate an average male. Smaller-than-average shields are also, overwhelmingly, of earlier date. More easily included in the burial rite, a smaller, more symbolic shield, perhaps re-boarded in preparation for the grave, would have saved valuable resources – seasoned and carved timber and valuable leather, and allowed a more valuable combat shield to be passed down.
Perhaps, then, burial with large and sturdy combat shields was a rite usually reserved for the most wealthy elites, who's relatives, companions and subjects could afford such lavish and wasteful conspicuous consumption, exemplified by cases such as Sutton-Hoo mound-1, Taplow and Ford.
The theory that, as a whole, the corpus of shields in Anglo-Saxon burials might not be perfectly representative of shields borne into battle, is difficult to prove without studying assemblages of well-preserved gear recovered from actual battle contexts. No such finds exist from Anglo-Saxon lowland Britain, but many such assemblages have been discovered in northern Germany and Denmark dating to the few centuries immediately prior to the Anglo-Saxon age in Britain. These productive sites appear to represent a long-held custom of sacrificing war-gear of defeated enemies in lakes; the finds often feature battle-damage, fit neatly into phases representing distinct depositional events (battles) often separated by decades or centuries, and so the status of the finds as genuine war-gear can be assumed, although the precise origin / identity of those who used the gear is almost never clear. Occasional metal rims, and survival of shield planks where ground conditions allow, show Iron-Age Germanic round shields were often large, almost always in excess of 70cm, and of similar construction and thickness to that which is evidenced from Anglo-Saxon burials, suggesting a high degree in conservatism of design over time. Of particular interest, an array of at least 25 circular shields from Nydam, associated with the pine-boat deposit (4th century CE) all bearing remarkable similarity to what is known of early Anglo-Saxon shields, varied from 83 to 104cm in diameter (Holst & Neilsen, 2020), but were mostly strongly clustered around a median of 90cm - as big as only the very largest shields from 5-7th century Anglo-Saxon graves.
While the debate regarding the intended purpose and utility of these small-to-medium sized shields will no doubt continue to rage, the notion that the majority of early Anglo-Saxon shields were tiny can be comprehensively dismissed. To reiterate, it would seem that the vast majority of early Anglo-Saxon shields were between 49 and 73cm diameter, with some rare cases confirmed to be smaller, and quite many substantially larger.
Although when held as if ready for battle, this shield looks quite absurd, it by no means represents the smallest of its kind; a number of finds have been confirmed to have been far smaller – examples such as Stretton-on-Fosse II g96, Westgarth Gardens g50 and g60 had diameters of 38cm, 34cm and 36cm respectively. This reconstruction's diameter sits in the middle of the “small shield” class defined by Dickinson and Harke (1992) and, were it not for more reliable 'maximum' diameter data, would sit close to the average board size for their sample of early Anglo-Saxon shields.
Dickinson, T.M., Harke, H., 1992. Early Anglo-Saxon Shields (Archaeologia). First Edition Edition. Society of Antiquaries.
Dickinson, T.M., 2005. Symbols of protection: The significance of animal-ornamented shields in early Anglo-Saxon England. Medieval Archaeology, 49(1), pp.109-163.
Holst, S. and Nielsen, P.O., 2020. Excavating Nydam-Archaeology, Palaeoecology and Preservation: The National Museum's Research Project 1989-99. Det Kgl. Nordiske Oldskriftselskab & Syddansk Universitetsforlag.
Stephenson, I.P., 2004. The Anglo-Saxon Shield. illustrated edition Edition. Tempus Pub Ltd.
Neilsen, K.H,. 2010 Style II and all that: the potential of the hoard for statistical study of chronology and geographical distributions. [ONLINE] Available at: https://finds.org.uk/staffshoardsymposium/papers/karenhoilundnielsen. [Accessed 25 March 2016].
Beatson, P. 2016. Viking Shields. [ONLINE] Available at: http://members.ozemail.com.au/~chrisandpeter/shield/shield.html. [Accessed 25 March 2016].
Thompson, A.,: Size Matters. 2013. Thegns of Mercia: Size Matters. [ONLINE] Available at: http://thethegns.blogspot.co.uk/2013/04/size-matters.html. [Accessed 25 March 2016].
Thompson, A.,: Grave Problems. 2013. Thegns of Mercia: Grave Problems. [ONLINE] Available at: http://thethegns.blogspot.co.uk/2013/12/grave-problems.html. [Accessed 25 March 2016].