by Antti Ijäs
The study of historical combat arts is comprised of archaeology, philology and practical experimentation. All of these domains have their own methodology and they provide answers for markedly different questions. On one hand, archaeology tells us what kinds of weapons existed at a given time period; on the other hand, philology gives us an idea of how they were used based on textual evidence of historical or instructional nature. Naturally, damage to weapons, armour and bones uncovered by archaeologists will also provide important data on how different armaments were used.
Experimentation and actual physical exercise, which in this context could also be termed “experimental archaeology”, should be used to verify the plausibility of using the weapons and armour in the assumed way. Sometimes insufficient knowledge and lack of practical experience can result in rather strange conclusions, which could in the worst case be perpetuated in subsequent literature, thus contributing to the ever-growing corpus of common misconceptions. In this essay, I will give an example of the importance of synthesizing knowledge by pointing out an error in the article “Archaeology and Beowulf” by Leslie Webster originally published in an edition of Beowulf (Mitchell & Robinson 1998: 183-194) and later reprinted with the Present-Day English verse translation by Seamus Heaney (Donoghue 2002: 212-236). This essay is loosely based on my earlier off-the-cuff blog post “Anglosaksien miekoista, lähdeviittaamisesta ja relatiivipronominista” (Ijäs 2014), in which I also discuss the only vaguely related topic of Old English relative pronoun.
In the natural sciences, experiments must be documented so that they may be replicated by others. Similarly, even in the most humanistic kinds of historical studies, one should indicate the data used so that others may review it and follow the reasoning used to draw the conclusions presented. There are essentially two distinct kinds of referencing in scientific writing. The first kind entails giving a source for the data used for drawing specific conclusions. To give an example from linguistics, this means that instead of claiming that one expression is used more often than the other based on personal assumption (or “reflection”), one would refer to a specific text corpus where the instances of one expression outnumber those of the other. The validity of such a statement is then subordinate to the quality (or, better said, quantity) of the corpus used. Citing a secondary source also falls under this category, since eventually the trail will lead to the primary source. The second kind of referencing is purely ethical and entails giving credit to previous research for having made certain conclusions. Failing to do this can be construed as plagiarism and is punished by the most severe methods available to the academic community. From a methodological point of view, this kind of referencing is, strictly speaking, irrelevant, for a logical conclusion is no less valid no matter who presents it or takes credit for it. Giving credit where credit is due is, however, of utmost importance for the social dynamics of the academia.
In the worst case scenario, an innocuous offhand reference to an authority generally considered to be credible might lead to the accumulation of misinformation mentioned earlier. This is especially true of monographs and articles that are likely to be read by a wide audience of students or researchers. However, if referencing to primary sources is done properly, or if at least the trail of references does not end before primary sources, there exists at least the possibility for checking the facts.
The purpose of this essay is, ultimately, to provide the reader with an example of an (arguably minor) scholarly lapse and the method of critical reading necessary to correct it. Below is the relevant quotation from Webster’s article (1998: 191):
They [the swords] are archetypally the ealde lāfe, powerful heirlooms to be prized for their power and might; some are the work of marvellous smiths (l. 1681), or eotenisc (l. 1558), the work of giants long ago; they may have names, or carry owner inscriptions and images which tell of the mythic past (ll. 1688-98). They have richly decorated hilts (l. 1698) and bear twisting and branching patterns (wyrmfāh l. 1698, ātertānum fāh l. 1459); their iron blades are fearsome double-edged weapons, which need two hands to swing them (l. 1461). Supporting archaeological evidence for much of this has long been recognized.
Line numbers refer to the primary source, i.e. Beowulf, or, more precisely, Mitchell’s and Robinson edition (1998) of Beowulf the article was published with. Interestingly, a footnote for line 1461 in the text proper refers to the passage quoted above, effectively lending Webster’s claim the authority of the said editors. As for “supporting archaeological evidence”, a passim reference is given to a secondary source, in this case H. R. Ellis Davidson’s book The Sword in Anglo-Saxon England: Its Archaeology and Literature (1962).
For anybody familiar with the Anglo-Saxon or similar Germanic swords (though attested in a single late 10th to early 11th century Old English manuscript, the story of Beowulf is set in Scandinavia, not England), the idea that their use would require two hands should strike as rather peculiar. According to the article, however, this proposition seems to be supported by both textual evidence and archaeological findings. To be fair, Webster does not specify which details are supported by archaeological evidence, but weight and handedness are certainly such basic and concrete qualities of a sword that they would be among the first characteristics inferred from surviving specimens.
I will start my analysis from the reference to the primary source. Below are lines 1459 to 1463 describing Unferth’s sword Hrunting with a literal translation:
ecg wæs īren, ātertānum fāh
āhyrded heaþoswāte; nǣfre hit æt hilde ne swāc
manna ǣngum þāra þe hit mid mundum bewand
se ðe gryresīðas gegān dorste,
‘the edge was iron, with poison-twigs decorated,
hardened with battle-sweat; never had it in battle betrayed
any man who grasped it with hands,
who dared to go on perilous expeditions
to the enemies’ meeting place.’
Webster seems to assume that the plural mid mundum ‘with hands’ refers to using the sword exclusively with both hands. It should be noted that even though the antecedent manna ǣngum ‘any man’ (lit. ‘of-the-men any’) could arguably be either singular or plural, the verb bewand is singular, so it is indeed obvious that two hands of a single man are meant. But nevertheless, since we are dealing with poetry, two things should be considered. First, it could very well be that the poet has chosen words based on acoustic qualities such as rhythm, in this case bewand instead of the trisyllabic plural bewundon. Secondly, the verb bewindan is by no means the regular word for using, wielding or swinging a sword: the literal meaning is ‘to wind or bind around or about’ (in Gothic biwindan is used in the sense ‘to swaddle’). Its usage here seems rather poetic and could refer to grasping the handle with one hand and the sheath with the other when starting to use the sword. There is, of course, a very remote possibility that we are dealing here with a technical term (or a usage inspired by a technical term) related to the Middle High German winden ‘to turn (the sword)’ familiar to practitioners of historical European martial arts.
Next I will consider the “supporting archaeological evidence”. Because no specific page number is provided, I had little choice but to look up the relevant passages myself. The one quoted below concerns the size of the hilt (Davidson 1962: 61):
The grips of Anglo-Saxon and Viking swords often seem surprisingly small. One sword found at Reading (now lost) is said to have had a grip too small for a grown man’s hand, and it may be noted that the other swords in this museum, including the one from Shifford, have small grips. […] Some of these swords may have been made for boys or slender men, but another possible explanation was given by R. E. Oakeshott, who suggested that grip and pommel might be grasped together, and confirmed this by illustrations from early manuscripts; the brazilnut pommel in particular fitted easily into the hand in this way.
In this case, archaeological evidence would definitely suggest single-handedness. In another passage Davidson says that “a good swordsman might be capable of using both hands as the occasion demanded” (ibid. 201), referring to the passage from the Droplaugarsona saga (written probably in the 13th century) quoted below:
Grim could fight equally well with both hands; he brandished a sword in his left hand, but with the right he struck at Gauss and cut off his leg above the knee.
Again, there is no indication whatsoever about the sword requiring two hands to swing it in an effective manner. The Old Norse text (Jónsson 1878) makes it even more obvious that Grímr has indeed two swords, one in each hand:
Grímr hafði tvau sverð, því at Gauss kunni at deyfja eggjar. Grímr vá jafnt báðum höndum. Hann brá upp sverði með vinstri hendi, enn hjó með hinni hœgri til Gauss ok af fótinn fyrir ofan kné. Nú fell Gauss, ok í því veifði hann sverðinu at Grími ok kom á fótinn, ok varð þat svörðusár. Nú flýði víkingrunn á brott, enn Grímr tók silfrit, ok fekk góðan orðstír af verki þessu.
‘Grímr had two swords, because Gauss could make edges blunt. Grímr was equally good with both hands. He brought up his sword with his left hand, and with the right he struck at Gauss and cut off his leg above the knee. Then Gauss fell and waved his sword at Grímr and hit his foot causing a surface wound. Now the viking fled away, but Grímr took the silver and got a good reputation of this work.’
(My emphasis and translation.)
With the sword in his left hand Grímr lures his opponent into lifting his defences and immediately uses the opening thus created to strike with the other sword in his right hand at his opponent’s leg, chopping it off. It should be borne in mind that sagas in general do not qualify as a reliable source and the descriptions contained in them should not be taken at face value. In any case, Grímr’s apparent ambidexterity needs not be considered a super-human feat conjured up by the imagination of the anonymous author. On a related side note, in Speculum regale (Konungs skuggsjá), a Norwegian educational textfrom the late 13th century, it is stated that an expert in the use of weapons should be trained to use both hands alike, though swords are not explicitly mentioned in this context (chapter 37).
The typical size of the grip and the fact that the sword was used together with a shield constitute conclusive and, as far as I am aware, widely undisputed evidence that the Germanic swords of the period discussed were single-handed. Whereas an interpretation of the Beowulf verses quoted earlier might tentatively suggest otherwise, a simple experiment with any replica sword made according to specifications provided by archaeological specimens will reveal that only a child would actually require two hands to swing the Anglo-Saxon sword or any other similar Migration period Germanic sword based on the Roman spatha. Further experimentation will also confirm that it is indeed within human capabilities to chop off a limb using only one hand, even though such a technique might not have played a decisive part in the ancient battlefields.
But despite the overwhelming facts regarding swords enumerated above, it is nevertheless worth discussing whether dismissing the plural mid mundum ‘with hands’ as a product of poetic licenceis a valid approach.Davidson points out that oral poetry can be expected to yield more accurate details than historical works, since the audience would have been intimately familiar with the objects described by the performer (1962: 3). There actually is reason to question the identification of Hrunting as a sword in the modern sense: inaddition to sweord, mēce ‘sword’ (passim)and waegesweord ‘wavesword’ (1489), it is once referred to as a hæftmēce (1457), glossed in Bosworth-Toller as ‘a hilted sword’ (also in Mitchell & Robinson 1998, but more interpretatively ‘sword with a long hilt’ in Wrenn 1958). This rather trivial extension (all swords have hilts) could be taken as a poetic byform of mēce, but in the supplement volume a comparison is drawn to the Old Norse hefti-sax mentioned in Grettis saga. In the saga, the weapon in question is wielded by a giant and is referred to as a fleinn ‘pike’ with a tréskapt ‘wooden shaft’ (66.4):
En er Grettir kom at honum, hljóp jǫtuninn upp ok greip flein einn ok hjó til þess, er kominn var, þvíat bæði mátti hǫggva ok leggja með því. Tréskapt var í; þat kǫlluðu menn þá heptisax, er þannveg var gǫrt.
‘But when Grettir came near him, the giant leapt up, grabbed a pike and struck at him, who had just come, for with [the pike] he could both cut and thrust. A wooden shaft it had, and [a weapon] that was made in such a way men would back then call a haft-seax.’
(My emphasis and translation.)
In a footnote to his critical edition, Boer notes that the word heptisax is not attested outside Grettis saga, where is appears twice, but the striking similarity with hæftmēce would suggest a connection with Beowulf (1900: 231). Davidson, who identifies the weapon as a ‘knife fastened to a wooden shaft’ points out that the weapon would seem to have been unfamiliar to the intended audience of the saga (1962: 134).
Interestingly, a similar Old English word, stæfsweord ‘staff-sword’ appears in Ælfric’s glossary as a translation (possibly coined by the author) of Latin dolones, plural of dolo ‘iron-pointed staff, pike’ (see Wright 1857: 35ff. for this and other Latin and Old English weapons terms). An exact cognate of the compound stæfsweord is also attested in Old High German, where stabaswert ‘staff-sword’ is used to gloss Latin framea, which according to Tacitus is the Germanic word for hasta ’spear’ (Germania 6.1), but the word was used by later Christian writers to signify ‘sword’ (Green 1998: 185). Accordingly, Ælfric’s glossary gives sweord as the Old English equivalent of Latin framea.
Following the lexical trail further is beyond the scope of this article. In any case, this etymological excursion would seem to indicate that if the weapon referred to in Beowulf was wielded with both hands, as mid mundum would suggest, and if indeed the name Hrunting is derived from hrung ‘rung, staff’, it would seem that the weapon – at least in the archetypal story – was something akin to a pike or a glaive, not the typical Anglo-Saxon or Viking sword or any of their known predecessor.
The linguistic surface usually passed through as quickly as possible by non-philologists offers a rich and, admittedly, confusing domain of data for studying not only language but the physical world language has been used to describe. In conclusion, I hope to have demonstrated the importance of not only checking second-hand references and studying the primary sources whenever possible, but also of inter-disciplinary approach combined with practical experimentation. This is especially true when it comes to weapons, since, after all, they are quite concrete tools and should not be treated and studied as philosophical abstractions.
Boer, R.C. (ed.). 1900. Grettis saga Ásmundarsonar. Halle A.S.: Max Niemeyer.
Bosworth, Joseph & T. Northcote Toller (eds.). 1989 (supplement 1921). An Anglo-Saxon Dictionary.
Mitchell, Bruce & Fred C. Robinson (eds.). 1998. Beowulf: An Edition with Relevant Shorter Texts, including ’Archaeology and Beowulf’ by Leslie Webster. Oxford: Blackwell.
Davidson, H. R. Ellis. 1962. The Sword in Anglo-Saxon England: Its Archaeology and Literature.Oxford: Blackwell.
Donoghue, David (ed.). 2002. Beowulf: A Verse Translation. Translated by Seamus Heaney. New York: W.W. Norton & Co.
Green, D. H. 1998. Language and History in the Early Germanic World. Cambridge: Cambridge UP.
Ijäs, Antti. 2014. “Anglosaksien miekoista, lähdeviittaamisesta ja relatiivipronominista.” Filologogrammata. <http://filologogrammata.blogspot.fi/2014/07/anglosaksien-miekoista.html>
Jónsson, Þorleigr (ed.). 1878. Droplaugarsona saga. Reykjavík: Forlag Kristjáns ó. Þorgrímssonar.
Petersen, Jan. 1919. De Norske Vikingesverd: En Typologisk-kronologisk Studie over Vikingetidens Vaaben. Kristiania: Jacob Dybwad.
Webster, Leslie. 1998. “Archaeology and Beowulf.” In Michell & Robinson 1998; 183–194, reprinted in Donoghue 2002: 212–236.
Wrenn, C.L. (ed.) 1958. Beowulf with the Finnesburg Fragment. Revised and enlarged edition. London: George G. Harrap.
Wright, Thomas (ed.). 1857. A Volume of Vocabularies. Privately printed.
To learn more about Combat Archaeology Click Here.
Antti holds a MA degree in English Philology from the University of Helsinki, with a minor in pedagogy and an advanced minor in Latin language and Roman literature. Having rounded up his classical education with advanced studies of Greek language and literature, he specializes in historical linguistics and philology with a particular interest in the Germanic language family. He served his military service in the Guard Jaeger Regiment and holds the rank of 1st lieutenant (infantry) in reserve. Currently he is working on his doctoral dissertation on Royal Armouries MS I.33 at the University of Helsinki. Antti is the chief instructor of Wing Tsun Finland, an instructor in Weapons Combat Systems and the Regional Head of KDF Finland.