This article addresses the relatively under-researched subject of Early Medieval double-edged swords discovered within the territory of present-day Serbia. These finds are dated to the 10th and 11th centuries and consist exclusively of stray finds. This work builds upon two previous studies in which the author examined double-edged swords from Romania and Bulgaria spanning the 8th to 12th centuries (Viskupič 2023a; 2023b).
The primary objective of this study is to present a systematic inventory of these swords and their fragments, classified according to the typological criteria established by Jan Petersen and Alfred Geibig, while incorporating relevant international analogues. In contrast to the corpus of Croatian specimens, which is significantly more extensive, Serbian finds have historically received less attention in the academic literature.
All items discussed herein have been previously published. The first comprehensive inventory of Early Medieval double-edged swords was compiled by the Croatian archaeologist Zdenko Vinski in 1983, recording twelve specimens from the former Yugoslavia. In the 1990s, the Hungarian archaeologist László Kovács analyzed swords from the Carpathian Basin, including five finds from the autonomous province of Vojvodina. The most recent assessments of specific finds from the study area were published by Silviu Oța (2008) and Milica Radišić (2022).
Typology
The following analysis of 10th- to 11th-century double-edged swords includes specimens categorized as “Western-type” swords. For these, we have adopted Jan Petersen’s classification system, based on the morphology of their hilts. In his thesis, Petersen analyzed nearly one thousand swords from Norway, dated between the 8th and 11th centuries. His typology was constructed using morphological-decorative criteria, designating individual types with 27 letters of the Norwegian alphabet (A through Æ). Furthermore, he identified an additional 20 variants classified as “special types.” After more than a century, it has become evident that not all of Petersen’s designated types remain analytically valid. For instance, Special Type 20 is now recognized as a torso of Type H/I or V (Petersen 1919: 174-175, Fig. 135; Androshchuk 2014: 59-60).
It is also necessary to emphasize that his typology focused primarily on Norwegian material; consequently, its application to hilt types outside the Norwegian or broader Scandinavian milieu is not always seamless. This is evidenced by sword production that, while perhaps inspired by traditional Northern variants, deviates in technical parameters, form, or hilt ornamentation. Although this typology may not fully reflect the total diversity of swords from the period under consideration, it remains one of the most widely utilized frameworks for the typological classification of European swords from the 8th to 11th centuries.
Fig. 1. Petersen’s sword typology (after Košta 2021: Figs. 1–2).
Concurrently, consideration was given to the fact that for certain swords dating to the 11th century, the application of Petersen’s classification is insufficient; consequently, it was substituted with the combinatorial typology developed by Alfred Geibig. His research involved the formulation of a new typology for the structural components of the hilt (pommel and crossguard) as well as the blades. This system was applied to swords from West Germany dating between the 8th and 13th centuries. The researcher distinguishes nineteen pommel types and fourteen blade types. Furthermore, he proposed three potential methods for mounting pommels onto the blade tang (Geibig 1991: 90–100, Fig. 24). Subsequently, Geibig’s typology was expanded by Jiří Košta and Jiří Hošek to include several additional variants and transitional types during their analysis of a sword corpus from the Czech territory (Hošek – Košta – Žákovský 2021: 261–274, Fig. 94). Regarding the specimens in our study that preserve the pommel, we have identified the third construction variant – specifically variants Ib and IIIa as defined by J. Košta and J. Hošek.
The investigation identified nine specimens that meet the established chronological and typological criteria. Five items display definitive typological features. For the remaining four specimens, a more precise typological classification was precluded by their fragmentary state of preservation; nonetheless, their morphological characteristics permit their assignment to the defined chronological horizon. We cannot dismiss the possibility that further swords or fragments may be held in museum repositories that remain unidentified or have yet to be published.
Fig. 2. Geibig’s combinatoin sword typology: a. frontal view of the pommel; b. frontal view of the crossguard (after Geibig 1991).
Fig. 3. Diagram of construction types illustrating various methods of mounting the pommel to the blade tang (1. after Geibig 1991: Fig. 24; 2. after Hošek – Košta – Žákovský 2021: 94).
Petersen Type X
This type constitutes a group of swords characterized by a semicircular pommel fashioned from a single piece of iron. These pommels exhibit a uniform width and rounded extremities, while appearing relatively narrow in both horizontal and lateral profiles (Petersen 1919: 159; Hošek – Košta – Žákovský 2021: 302). The crossguards are straight and rectangular in section, though in certain instances, they may be slightly curved toward the blade. In horizontal view, the ends are either rounded or squared (Hošek – Košta – Žákovský 2021: 302). The tang extends through the pommel and is secured by peening at its apex; this method of attachment corresponds to Geibig’s combination type III (Geibig 1991: Fig. 24). Jan Petersen subdivided this type into two variants – early and late (Petersen 1919: 159) – distinguishing them based on the divergent morphology of the pommels and crossguards. The earlier variant is described as having a taller, wider, and less massive pommel paired with a longer, more robust crossguard, which is occasionally slightly inclined toward the blade. Conversely, the later variant is characterized by a smaller but more massive pommel and a thinner crossguard of equal length (Petersen 1919: 159, Figs. 124–129).
Fig. 4. Petersen Type X sword from an unprovenanced Norwegian locality, Telemark County (C8073). Source: unimus.no.
In Geibig’s classification, this type corresponds to combination type 12, variant I (Geibig 1991: 57, Fig. 13). The pommels and crossguards are typically undecorated, with the exception of several specimens featuring incised lines that divide the pommel into a base and a three-lobed crown. These incisions serve to imitate more prestigious swords with Type V, W, or E pommels (Hošek – Košta – Žákovský 2021: 303). Examples of such ornamentation can be found on a sword from an unknown Norwegian locality (C280) and on a specimen from the East Prussian site of Zohpen (present-day Suvorovo, Kaliningrad Oblast) (Petersen 1919: Fig. 125; Petri 2017: 145–150, Pl. 4b). Furthermore, there are recorded specimens with inlaid decoration or precious metal plating, as well as a specific find from Oljonsbyn (Orsa parish, Dalarna, Sweden; (SHM 11097), featuring a pommel decorated with stamped triangles (Hošek – Košta – Žákovský 2021: 303).
Petersen Type X swords are encountered across a vast geographical range, spanning from the British Isles through Scandinavia and Western/Central Europe to Eastern and Southeastern Europe. Their total number is estimated at approximately 370 specimens. It is hypothesized that they evolved from Petersen’s types K and N within the Frankish Empire or territories under its influence. The combination of structural simplicity and innovative technological solutions in blade manufacture contributed to their popularity among higher social strata (Hošek – Košta – Žákovský 2021: 304–305). Their production, and the subsequent fitting of hilts onto imported or local blades, likely occurred in local workshops outside the Frankish heartland (Košta et al. 2019: 220–221). Evidence for the repair, supplementation, or replacement of hilt components is provided by finds such as the three swords from the Great Moravian site of Pohansko near Břeclav (Košta et al. 2019: 220–221).
Jan Petersen dated the earliest Type X specimens to the first half of the 10th century and later examples to the 11th century (Petersen 1919: 182). However, it is possible that their origin dates back to the first two-thirds of the 9th century, although finds from the former Frankish Empire are frequently stray or riverine, complicating precise dating (Hošek – Košta – Žákovský 2021: 304–305). In this regard, swords from Great Moravian contexts are of paramount importance, as they primarily originate from secure grave contexts dated from the second half of the 9th to the early 10th century (Hošek – Košta – Žákovský 2021: 305). In the territory of the former Kyivan Rus’, their occurrence is placed between the 10th and early 11th centuries (Kainov 2012: 55–56), and in the Baltic region between the 10th and 12th centuries (Kazakevičius 1996: 138). Hungarian finds, dated to the late 10th and early 11th centuries, represent one of the most widespread sword types in that region (Bakay 1965: 31, 33). From the territory of present-day Croatia, seven finds are known: Brodski Drenovac, Dabar, Koljane, Šoderica, Jarun-Zagreb, and two specimens in a private collection (Vinski 1983: Cat. Nos. 6, 7, 10; Demo 1984: 218–225, Pl. 1.2, 3.2; Radić 1991: 77–80; Kovač 2003: Cat. Nos. 7, 8). Their chronological horizon is established from the second half of the 9th century to the beginning of the 11th century (Vinski 1983: 24–25; Bilogrivić 2019: 53–54).
Fig. 5. Petersen Type X sword from the Jarun–Zagreb area (after Bilogrivić 2019: Fig. 2).
Only one find from the territory of Serbia corresponds to the type under consideration. This sword originated from an unidentified location in the vicinity of Vršac (South Banat District, Vojvodina) and is a stray find (Vinski 1983: Note 18). The sword measures 85 cm in length and is preserved in relatively good condition. The pommel is of single-piece construction, semicircular in frontal view with a straight base. The blade tang is trapezoidal, widening from the pommel toward the crossguard. The guard is relatively long and straight with rounded edges. The blade is straight and double-edged; however, it is not possible to determine from the available photography whether it features a fuller. The object is reportedly curated at the City Museum (Gradski muzej) in Vršac (Kovács 1994–1995: Cat. No. 89).
Fig. 6. Sword from the vicinity of Vršac (after Radišić 2022: Fig. 1. 1).
Petersen Type Y
Jan Petersen identifies this type as the second most common and structurally simplest among swords of the Late Viking Age (Petersen 1919: 167). The pommels of Type Y swords are forged from a single piece of iron, featuring a slightly concave base with upturned ends. The upper portion of the pommel is characterized by three rounded lobes separated by saddle-shaped depressions. The crossguard is relatively long and may be curved in the opposite direction of the pommel base – with the terminals directed toward the blade – or straight, as seen in Petersen Type X. Both the pommel and crossguard taper toward their extremities, which may be either squared or rounded (Androshchuk 2014: 83–84). In some instances, the pommel is demarcated into a base and crown by decorative grooves.
Petersen subdivided Type Y into two variants. The first variant bears a resemblance to Petersen Type P but is distinguished by more robust hilt components lacking non-ferrous metal ornamentation (Petersen 1919: 168, Fig. 130). The second variant is characterized by a low crossguard and base, as well as a more rounded central lobe on the crown (Petersen 1919: Figs. 131–132). Within Geibig’s combinatorial typology, Petersen Type Y corresponds to Combination Type 13, subdivided into variants I and II based on the morphology of the pommels and guards (Geibig 1991: Fig. 14). The tang passes through the entire pommel and is secured at the apex (Hošek – Košta – Žákovský 2021: Fig. 103. f-h), aligning with Geibig’s construction variant III (Geibig 1991: Figs. 14, 24). Pattern-welded blades are also frequently associated with this type (Hošek – Košta – Žákovský 2021: 309). Approximately 118 specimens of Type Y are currently recorded; beyond Scandinavia, they have been recovered in Poland, Germany, the Netherlands, Czechia, Austria, Hungary, the Baltic states, Finland, France, the territory of the former Kyivan Rus’, and Slovakia (Bakay 1965: 31; Moilanen 2015: 258-259; Hošek – Košta – Žákovský 2021: 309-314). In the British Isles, finds typically consist of transitional variants closely related to Type L or other Anglo-Saxon forms (Hošek – Košta – Žákovský 2021: 313).
Fig. 7. Petersen Type Y sword from Bøen Østre, Tinn, Telemark, Norway (C21038/a). Source: unimus.no.
Only a single specimen of Petersen Type Y is known from Serbia. It is a stray find discovered during excavation works in the village of Banatski Brestovac (South Banat District, Vojvodina) and is held by the Historical Museum of Serbia in Belgrade (Inv. No. 426) (Aleksić 2004: 251). The museum acquired the item from Branko Ostojić in 1983 (Aleksić 2004: 251, Note 1). According to the provider, the sword was unearthed in the center of the village at an unspecified depth, with no other associated artifacts recovered (Aleksić 2004: Note 1).
The sword survives as a blade lacking its tip, with a damaged edge and the pommel still attached to the tang. The pommel is a single-piece construction, divided by two saddle-shaped grooves into three lobes with rounded edges; the base is oval. The blade is straight and double-edged, featuring a fuller that remains discernible only in the upper half. The blade was manufactured using pattern-welding, as evidenced by a herringbone pattern with points oriented toward the hilt, located a few centimeters below the shoulders; an identical pattern repeats further down the blade (Aleksić 2004: 256). On one side of the fuller, in its upper portion, there is a shallow circular indentation, which likely originally contained an inlaid motif (Aleksić 2004: 256).
The total preserved length of the sword is 87 cm, of which the blade accounts for 75.4 cm and the tang with pommel 11.6 cm (Aleksić 2004: 251). The blade width at the shoulders is 5 cm, and the width of the preserved fuller is 2.8 cm. The tang measures 9.2 cm in length, with its width varying between 1.1 and 3.3 cm from the pommel to the shoulders. The pommel is 6.5 cm long and 2.4 cm high (Aleksić 2004: 251–252).
Fig. 8. Sword fragment from Banatski Brestovac (after Aleksić 2004: Fig. 1).
Fig. 9. Detail of the tang and pommel of the sword from Banatski Brestovac (after Aleksić 2004: Fig. 2).
Petersen Type Y swords are generally dated from the first half of the 10th century to the turn of the 11th century (Petersen 1919: 182; Hošek – Košta – Žákovský 2021: 312). However, finds from Great Moravian contexts demonstrate their presence as early as the 9th century. Radiocarbon analysis of skeletal remains from specific Great Moravian graves supports adjusting the dating of these swords to the final third of the 9th century, extending at the latest to the turn of the 10th century (Hošek – Košta – Žákovský 2021: 312). Conversely, Type Y finds from Hungary and Slovakia are typically dated to the late 10th or early 11th century (Bakay 1967: 169).
In Hungary, three Type Y finds are recorded. One is a grave find from Székesfehérvár (Sóstó district, Vízművek site, Grave 1) (Bakay 1965: Cat. No. 54, Pl. V). The remaining two are contextless finds: one from a brickyard in Miskolc, which is reportedly lost (Bakay 1965: Cat. No. 36), and another from an unknown location, likely in the former Esztergom County, which is also either lost or unidentified (Kovács 1994–1995: 180). Kornél Bakay notes a further specimen from Solomonovo (Tiszasalamon) in Western Ukraine, near the borders of Hungary and Slovakia (Bakay 1965: Cat. No. 45, Fig. 15).
In Slovakia, three published specimens are currently identified (Jócsik 2024). Two are associated with Old Hungarian cemeteries: Grave 3 at Hurbanovo-Bohatá (Rejholcová 1976: 203, Pl. II.2) and an alleged find from a skeleton inhumation in Marcelová (Ruttkay 1975: Cat. No. 91, Fig. 6.3). A third specimen was recovered as a stray find during dredging operations on the Danube (Turčan 1997: Fig. 1. a–d). Furthermore, a fourth Type Y sword was discovered during systematic archaeological excavations in 2023 but remains unpublished.
The sword from Banatski Brestovac is dated between the second half of the 10th century and the early 11th century (Aleksić 2004: 257). Marko Aleksić posits that, given its pattern-welded blade, the sword originated in a Western European workshop – likely in the Middle Rhine region. He suggests that the object may have reached southern Banat through military engagements between the Hungarians and the Duchy of Ajtony (Ahtum) during the late 1020s, or alternatively via trade routes (Aleksić 2004: 257, 263).
Fig. 10. Petersen Type Y swords: a. sword dredged from the Danube (after Turčan 1997: Fig. 1. d); b. Marcelová (after Ruttkay 1975: Fig. 6. 3); c. Hurbanovo-Bohatá (after Rejholcová 1976: Pl. II. 2); d. Székesfehérvár-Vízművek (after Bakay 1965: Pl. V); e. Solomonovo (after Bakay 1965: Fig. 15).
Petersen Type Z
Swords of Petersen Type Z are likewise assigned to the Late Viking Age (Petersen 1919: 175). Jan Petersen defined these specimens by their curved crossguards and pommel bases paired with trilobate crowns, where the central segment is both the tallest and the thickest, similar to Type S (Petersen 1919: 175). Fedir Androshchuk contends that the criteria established by Petersen for this type lack sufficient clarity. He considers the curved guards and bases to be the primary diagnostic feature, identifying a significant number of variants which he classifies into three subtypes (Androshchuk 2014: 84–86).
The classic type, according to Petersen (1919: Figs. 136–137; Androshchuk 2014: Pls. 134, 136, 139), is distinguished by a pommel base curved upward and a crossguard curved in the opposite direction (downward toward the blade). These curvatures may range from subtle to pronounced. The terminal edges of both components are rounded, and some may expand, reminiscent of Type S. The crowns are trilobate, secured to the base by two rivets or a tang. The division of the crown is often accentuated by twisted silver wires, and the lateral lobes frequently take the form of zoomorphic (animal) heads. In some instances, the lobes are decorated with “eyes, eyebrows, or ears” made of copper alloy (similar to Type R), and the hilt ornamentation is often further embellished with silver inlay (Androshchuk 2014: 84; Hošek – Košta – Žákovský 2021: 316, Note 196). Other variants feature a pommel with a semicircular upper side, a convex base, and a downward-curved guard (Androshchuk 2014: Pl. 159). Hilt components may be manufactured from either copper alloy or iron (Androshchuk 2014: 85). A rarer variant, known from only two cases, features a five-lobed pommel and a downward-curving guard (Androshchuk 2014: 85–86, Pl. 135).
Fig. 11. Petersen Type Z sword. Løken, Hole, Buskerud, Norway (C5402). Source: unimus.no.
The National Museum of Banat in Timișoara houses a sword (Inv. No. 3203) from the site of Temesch – Cubin on the Danube. Located in the former Torontál County of the Habsburg Empire, the site corresponds to present-day Kovin in the Vojvodina region of Serbia (Pinter 2007: 73). Given the administrative nomenclature cited in early records, the discovery likely predates 1918. According to Zdenko Vinski, the specimen represents a stray find (Vinski 1983: 44, Note 19) and is preserved in relatively good condition. According to Pinter (2007: 73), the pommel was likely manufactured as a single piece, though it is visually bipartite, delineated into a base and a crown by a slightly curved horizontal groove. The base is moderately curved, with its terminals oriented upward. The crown is divided into three lobes by two oblique grooves; the lateral segments are lower, while the central lobe is the most prominent and widest, featuring an oval superior margin. The tang expands from the pommel toward the crossguard, which is notably long and massive with rounded ends and a slight curvature toward the blade.
The blade is almost entirely preserved, albeit lacking the tip and exhibiting damage to its inferior section. A fuller runs along the central axis; while Vinski posits that it was pattern-welded, Karl Z. Pinter maintains the contrary, stating that the blade shows no visible evidence of pattern-welding, inscriptions, or marks (Vinski 1983: 44, Note 19; Pinter 2007: 73). The technical parameters of the sword are as follows: the total length is 82.2 cm, with the pommel measuring 6.9 cm in length, 3.4 cm in height, and 1.6 cm in thickness. The tang is 9.1 cm long, with a width ranging from 1.8 to 3.1 cm and a thickness of 0.31 cm. The crossguard has a length of 13.6 cm and a thickness between 1.13 and 1.8 cm. The extant blade length is 69.1 cm; its width tapers from 4.5–4.7 cm at the guard to 2.3 cm toward the tip, with a blade thickness of 0.34 to 0.13 cm and a fuller width varying between 2.1 and 1.4 cm.
Regarding its classification, Pinter argues that the sword lacks direct analogies within the Petersen typology (Pinter 2007: 73). He notes hilt components reminiscent of Type L or Special Type 15, a crossguard similar to Type Y, and a trilobate pommel showing affinities with Petersen Type V, Kirpichnikov Type Z, and Geibig Construction Type 7. He also highlights early forms of Petersen Type X (or Geibig Type 12) characterized by decorative grooves dividing the pommel, similar to the Kovin specimen. Thus, Pinter interprets it as a transitional variant between Types L and X, potentially a “Special Type L-X” or Geibig’s Combination Type 11 (Pinter 2007: 74). Conversely, Vinski and Kovács classify the sword as Petersen Type Z, a designation supported by the aforementioned morphological traits (Vinski 1983: 25; Kovács 1994–1995: 167). Vinski dates the specimen to the 11th century, regarding it as an isolated example within the Danube region and Southeastern Europe (Vinski 1983: 27). A close analogue is the find from Bøen Nordre, Telemark, Norway (C60769). Another Type Z specimen from Southeastern Europe was recovered at a medieval settlement near Gradešnica (Vratsa region, Bulgaria), dated between the second half of the 10th and mid-11th centuries (Zlatkov 2014: 142–143, Cat. No. 41).
Fig. 12. Petersen Type Z sword, Bøen Nordre, Telemark, Norway (C60769): a. current state of preservation; b. X-ray radiograph. Source: unimus.no.
In general, Type Z swords are dated to the late 10th and early 11th centuries (Androshchuk 2014: 86). According to Żak, this type combines elements of Type L or R-S, with guards potentially derived from the curved crossguards of nomadic sabers (Żak 1960: 333–334; Kirpichnikov 1966: 35). The origins of Type Z remain debated; they are variously attributed to Scandinavian or Rus’ manufacture, or hypothesized to have originated in the Eastern Baltic region (Moilanen 2015: 260).
Fig. 13. Sword from Kovin (after Radišić 2022: Fig. 1. 3).
Geibig Combination Type 16
Two specimens recovered from the historical region of Banat correspond to this type within the territory of present-day Serbia. Alfred Geibig subdivided the pommels of Combination Type 16 into two variants, which were defined by their proportional characteristics given their morphological identity (Geibig 1991: 70–71, Fig. 17). This analysis focuses on the first variant, to which the aforementioned swords can be assigned. The hilts of this type are characterized by lenticular pommels and long, straight, rectangular crossguards. Some specimens exhibit a slight thickening at the center of the guard, which subsequently tapers toward the terminals (Geibig 1991: Fig. 17). While often plain, decorated examples have been documented in the Czech Republic, France, and Poland (Hošek – Košta – Žákovský 2021: Fig. 123. a–c).
The method of securing the pommel to the blade tang conforms to Geibig’s Construction Type III (Geibig 1991: 71, Fig. 24). Lenticular pommels belonging to Combination Type 16, frequently referred to as “Brazil nut” pommels, were categorized by Ewart Oakeshott as Type A (Oakeshott 1964: 80, 93). Their chronological horizon is relatively broad; the earliest examples appeared in the mid-10th century and remained widespread until the mid-12th century, with sporadic occurrences persisting into the mid-13th century (Oakeshott 1964: 80; Hošek – Košta – Žákovský 2021: 359).
Iconographic evidence for this type is found in Ottonian art, specifically in the manuscript produced for Emperor Otto III in Reichenau between 983 and 991 (Oakeshott 1964: 83). One folio depicts the Emperor enthroned, flanked by a male figure – likely a nobleman – holding a sheathed sword featuring a straight crossguard and a Brazil nut pommel (Oakeshott 1964: Fig. 8). Marko Aleksić suggests that the production of swords with lenticular pommels may have originated in smithies located in the Middle Rhineland (Aleksić 2007: 33).
As previously noted, only two finds are recorded from Serbian territory. The first originated from the vicinity of Zrenjanin and is a stray find with a total length of 89 cm (Vinski 1983: 44, Note 20). The sword is curated in the National Museum in Zrenjanin (Vinski 1983: 42, Cat. No. 3). It is preserved almost in its entirety, with the exception of the tip. The pommel is lenticular and appears relatively massive, mounted on a short blade tang that widens from the pommel toward the crossguard. The guard is massive, straight, and long, characterized by a rectangular form with parallel lateral edges. Neither the guard nor the pommel exhibits discernible decorative elements. The blade is double-edged and straight, tapering toward the missing tip. A fuller runs along the central axis of the blade, showing damage in two distinct areas near the center; additional wear is evident in the lower sections of the blade.
Fig. 14. Sword from the vicinity of Zrenjanin (after Radišić 2022: Fig. 1. 2).
The second sword was recovered from the vicinity of Sombor (West Bačka District, northwestern Serbia) and is held in the collections of the Sombor City Museum (Vinski 1983: 42, Cat. No. 5). This is also a stray find for which detailed discovery circumstances are unavailable (Vinski 1983: 44, Note 21). The hilt components are preserved intact. The pommel is of lenticular form, forged from a single piece of iron and fitted onto a short, expanding blade tang. The crossguard is straight and long, featuring terminals with arched edges. Only a fragment of the blade remains; its inferior portion, including the tip, is absent (Vinski 1983: 44, Note 21).
Fig. 15. Sword from the vicinity of Sombor (after Vinski 1983: Pl. IV. 2).
A further specimen warrants discussion: a sword from Braničevo in eastern Serbia, situated near the Romanian border within the Braničevo District. While its morphological features align with certain Geibig construction types, its chronological horizon likely extends beyond the primary period under consideration.
The sword is a stray find, and its historical context is frequently associated with the Crusades, which, according to contemporary sources, traversed this region in 1096, 1147, and 1189 (Stojić 2022: 441, 463). Goran Stojić classifies the weapon as Oakeshott Type XI based on the blade’s morphology, which is notably long and narrow, featuring a slender central fuller (Stojić 2022: 441). The blade features an inscription on one side, positioned just below the crossguard. The text is flanked by two crosses with flared terminals. The identified characters + INIO . IN + have been interpreted as an abbreviation for In Nomine Iesu Omnipotentis (“In the name of Jesus the Almighty”), while a second portion of the inscription, consisting of the letters IND, is identified as a possible abbreviation for In Nomine Domini (“In the name of the Lord”) (Stojić 2022: 441). Religious epigraphy of this nature reached the height of its popularity on sword blades during the 12th century; the most frequent formulaic groups included variants such as In Nomine Domini, In Nomine Dei, and Homo Dei (Aleksić 2007: 120). Analogous inscriptions in the form INOINI are documented elsewhere, such as on a sword from Zlēku Pasilciems, Latvia, dated to the late 11th or early-to-mid 12th century, typologically assigned to one of the forms of Type T according to A. Tomsons (Tomsons 2018: 185, Fig. 23). Furthermore, two additional examples of unknown provenance are recorded in Poland, dated between the second half of the 12th and the early 13th century (Głosek 1984: 106, Pls. XII. 218, XV. 369).
The pommel of the Braničevo sword is oval in form, bearing a resemblance to the previously discussed specimens from Zrenjanin and Sombor, though it appears taller and narrower (Fig. 16). Typologically, it aligns with Geibig Construction Types 15 II, III, IV, or V, or alternatively, Oakeshott Type B (Geibig 1991: 65–68, Fig. 16; Hošek – Košta – Žákovský 2022: 359–360). Pommels of this type appeared across nearly the entire European continent from as early as the mid-9th century until the end of the 12th century, with potential overlap into the first half of the 13th century (Hošek – Košta – Žákovský 2022: 359). While the dating of the Braničevo specimen is generally estimated to fall between the 11th and 12th centuries, Goran Stojić favours a 12th-century attribution (Stojić 2022: 441, Cat. Nos. 35, 463).
Fig. 16. Sword from Braničevo (after Stojić 2022: Cat. No. 35).
Specimens without typological attribution
Four specimens are categorized under this heading. One remains in a state of preservation that precludes a precise assignment within the Petersen or Geibig typological frameworks. Two additional objects consist of blade fragments that, due to the absence of hilt components, cannot be typologically classified. The final item is a find of alleged Carolingian origin.
The first specimen originates from an unknown locality in the vicinity of Novi Bečej (Central Banat District, Vojvodina). Discovered after the Second World War, it was donated to the museum in Zrenjanin under inventory number 314; it is currently curated under the new inventory number 103 (A. Šalamon, pers. comm.). Zdenko Vinski initially characterized the sword as late- to post-Carolingian, assigning it to a combined W-X type (Vinski 1977-1978: 170) and subsequently to Type W, a classification adopted by other researchers (Vinski 1983: 42, Cat. No. 4; Oța 2008: 132).
The sword is preserved in its entirety with a length of 90 cm. The pommel, reportedly manufactured from a single piece of material, consists of a base and a trilobate crown intended to be decorated with non-ferrous metal. The crossguard is described as short, wide, and massive, with straight lateral edges (Vinski 1983: 43, Note 14, Pl. II. 1). However, its typological attribution remains contentious. Vinski noted that his data and drawings were provided by R. Radišić, suggesting that Vinski may not have examined the object firsthand. Furthermore, S. S. Zozulya and S. Yu. Kainov point out that the hilt components of the Novi Bečej sword were likely forged from iron rather than bronze (Kainov – Zozulya 2022: 111, Note 7), the latter being the standard material for Petersen Type W swords (Androshchuk 2014: 79). Conversely, Jiří Košta identifies the specimen as Petersen Type V (Košta 2021: Table 10, Cat. No. 2472). The sword is currently reported to be in a very poor state of preservation (M. Radišić, pers. comm.).
Fig. 17. Sword from the vicinity of Novi Bečej (after Vinski 1983: Pl. II. 1).
The second find is a blade from the village of Lučica (near Požarevac, Braničevo District) in eastern Serbia. It was discovered in the autumn of 1991 in a gravel pit on the banks of the Morava River (Milenković 1992: 57). Only the double-edged blade is extant, lacking both the crossguard and pommel. Its total length is only 76 cm, of which 5.7 cm comprises a fragment of the tang. The blade width measures 5.7 cm at the shoulders and 3 cm at the tip (Milenković 1992: 57). The blade was forged without a fuller and features a Latin inscription on both sides. The characters are identified as DOICTANH, appearing as DOⅠCTΛNH on one side and DOICTAИИ on the reverse (Milenković 1992: Fig. 1. a-b). The inscription is interpreted as an abbreviation for D(ominus) O(mnipotentis) I(esus) C(hristus) T(er) A(litissimi) (i)N N(omine) or DO(minus) O(mnipotentis) I(esus) C(hris) T(us) A(litissimi) (i)N N(omine) (Milenković 1992: 58). According to Kirpichnikov and Medvedev, this belongs to the “IN NOMINE” group of inscriptions, dated between the 11th and 13th centuries (Milenković 1992: 58).
While M. Janković and Z. Vinski suggested the blade might be a product of Byzantine workshops, Vinski also noted it could belong to the post-Carolingian milieu, such as Petersen Type X, where some Norwegian or Polish specimens also lack fullers. Marina Milenković posits that the blade may be a local imitation of West European products, citing two fragmentary swords from Gamzigrad as the closest analogues (Milenković 1992: 58). Theoretically, this may be an intentionally shortened blade that underwent repair; rather than forging a new blade, the original may have been reground at the point of damage. Analogies for this practice are found in two Finnish Type V swords from Hämeenlinna Vanaja (KM 370) and Hämeenlinna Tuulos (KM 708), with lengths of 49.8 cm and 55.7 cm, respectively (Moilanen 2015: 346). Such modifications might also suggest the weapons were intended for younger individuals (M. Moilanen, pers. comm.). Notably, both Finnish examples also feature inscriptions and geometric marks within the fuller (Moilanen 2015: 351–352).
Fig. 18. Inscribed sword blade from Lučica (after Milenković 1992: Fig. 1. a–b).
The third find is a double-edged blade of unknown provenance, held in the Military Museum in Belgrade (Inv. No. 16080) (Peković 2006: 115; 2015: Cat. No. 399). The blade is long, tapering toward the tip, with a central fuller. The tang tapers from the shoulders and features a perforation. The total length of the object is 95 cm, with the blade itself measuring 83 cm (Peković 2015: Cat. No. 399). Mirko Peković provides a broad dating between the 9th and 11th centuries. However, the tang appears relatively thin, showing morphological affinities with swords from the Steppe-nomadic milieu of the Migration Period. The presence of the perforation in the tang further supports this possible earlier classification; consequently, it cannot be ruled out that this specimen falls outside the chronological scope of the current study.
Fig. 19. Sword blade of unknown provenance from Serbia (after Peković 2015: Cat. No. 399).
The fourth specimen refers to an alleged Carolingian sword mentioned by Gavro Škrivanić in his 1957 study. This sword is reportedly curated in the National Museum in Belgrade under inventory number 2284 (Škrivanić 1957: 41). Unfortunately, specific data regarding this specimen are unavailable. Beyond the basic parameters of the weapon, no further details – such as the find-spot, circumstances of discovery, or state of preservation – are known. The item does not appear to have been published in other scholarly works. Its total length reaches 69.8 cm, of which the blade itself measures 54.3 cm, with a width (presumably at the shoulders) of 5.8 cm. The hilt, including the pommel, measures 10.5 cm (Škrivanić 1957: 41, Note 186). Given the absence of illustrations or photographs in the cited work, a precise typological identification is impossible. However, as the blade width is 5.8 cm, one may theoretically hypothesize that the sword belongs to the period under consideration. Notably, Škrivanić situates this sword within the timeframe of the Slavic migration to the Balkans (Škrivanić 1957: 41).
To conclude this chapter, we wish to briefly address several other thrusting and edged weapons, some of which were erroneously identified as swords, while others could not be verified or were incorrectly attributed to a discordant chronological horizon.
A second alleged Carolingian sword mentioned by Škrivanić was reportedly found in the vicinity of Belgrade. It consists of a blade fragment with a crossguard and tang (Škrivanić 1957: Fig. 8. 3). Škrivanić provides no further information beyond its location in the Museum of the Yugoslav People’s Army in Belgrade (now the Military Museum – Vojni muzej Beograd) under inventory number 74.251 (Škrivanić 1957: 41). From the extant drawing, it can be inferred that the blade is likely double-edged with a relatively narrow fuller. In frontal view, the crossguard appears straight and long, though possibly damaged at both terminals. The tang is relatively thin and similarly damaged. Based on these features, the crossguard could theoretically be assigned to Petersen Type X. A very close parallel to this specimen is found within the collections of the Military Museum in Belgrade itself. Under inventory number 16072, the museum records a similar fragmentary double-edged blade with a tang and a long crossguard, reaching a total length of 62 cm (Peković 2015: Cat. No. 401). According to Peković, this sword fragment was recovered from Kotor (Montenegro) and dates to the 13th–14th centuries. Due to the quality of the photograph in his work, the fuller cannot be clearly identified. Owing to the lack of detailed information regarding the sword mentioned by Škrivanić, it is difficult to ascertain whether these represent the same object or two distinct specimens of similar morphology.
Fig. 20. a. Sword fragment from the vicinity of Belgrade (after Škrivanić 1957: Fig. 8. 3); b. sword fragment from Kotor (after Peković 2015: Cat. No. 401).
Zdenko Vinski identifies a singularly unique sword find from the site of Ritopek in the municipality of Grocka, located approximately 20 km east of Belgrade on the right bank of the Danube. This iron sword is a stray find lacking detailed discovery circumstances (Vinski 1983: Note 120). Its total extant length is 70 cm. This relatively short weapon features a wide, double-edged blade without a fuller, marked by a cross-shaped motif located in its superior section just below the shoulders. From the shoulders, the blade tapers into the hilt tang, which expands toward its midpoint before narrowing again toward the pommel. The specimen has been preserved without a crossguard (if indeed one originally existed). A small guard-like element separates the pommel from the grip itself. The pommel is spherical, surmounted by a decorative element in the form of a smaller sphere with a concave depression (Vinski 1983: Note 120).
At the time of his study, Vinski noted that this was the first published account of the find. He considered it a potential Byzantine product deriving from the tradition of the earlier Roman gladius (Vinski 1983: 32–33). Furthermore, he remarked that the specimen lacks any closely related analogues. Given that the sword has never undergone detailed material analysis and is known only through a drawing by M. Ljubinković-Čorović and M. Tatić (Belgrade) (Vinski 1983: Note 120), definitive conclusions regarding its provenance remain elusive. However, the morphology of the pommel shows similarities to certain 9th-century swords of Byzantine or Oriental origin (Szőke 1992: Pl. 20; Yotov 2011: 38, Fig. 2c; 2022: Pl. I). A significantly later dating for the object cannot be ruled out (V. Turčan and R. Čambal, pers. comm.).
Fig. 21. Sword from the Ritopek site (after Vinski 1983: Pl. VI. 2).
Jiří Košta records one unclassified sword from Serbian territory (Košta 2021: Table 10, Cat. No. 2473), originating from the site of Novi Kneževac in the northeast of the country (North Banat District, Vojvodina). Literature regarding this find is virtually non-existent; the only reference to the object or its find-spot appears in László Kovács’s catalogue. The Hungarian archaeologist notes that an unspecified quantity of finds from the estate of Béla Tállián originated from Törökkanizsa (Serbian: Novi Kneževac). These entered the collections of the Hungarian National Museum (Magyar Nemzeti Múzeum) in 1899, either as donations or as a result of excavations by László Éber (Kovács 1994–1995: 178). Without providing explicit evidence, József Hampel assigned several of these items to a specific chronological horizon. The assemblage included a spearhead, a stirrup, a bridle, and three fragments of a single-edged sword. Kovács attributes these items to an Avar context (Kovács 1994–1995: 178), a classification also supported by József Szentpéteri (Szentpéteri 2002: 62).
Additionally, Kovács mentions two other items from Vojvodina as potential swords. One consists of sword fragments from a destroyed site in the village of Ada (North Bačka District), which are reportedly lost; Kovács notes it was impossible to verify whether these were swords or sabers (Kovács 1994–1995: 153). The second unverified item reportedly came from the village of Svilojevo (Szilágyi), Apatin municipality (West Bačka District). These were allegedly fragments of a corroded sword blade from a cemetery dated to the Hungarian Conquest period (honfoglalás) (Kovács 1994–1995: 176).
In the works of Mirko Peković, a single-edged sword from Stara Pazova – Surduk (Syrmia District, Vojvodina) is recorded and dated to the 8th or 9th century (Peković 2006: 115; 2015: Cat. No. 398). Its blade is long and straight, with a damaged edge and a missing section at the tip. The tang is also incomplete and features two rivets on its surface. The guard arms are short, featuring two elongated triangular projections extending toward the blade and the tang. Due to a low-quality older photograph, some authors previously classified the weapon as a Byzantine-type sword (Rabovyanov 2011: 78, Fig. 20). However, based on more recent, higher-quality photography (Fig. 22. a), the specimen appears more closely related to certain sabers and pallasches from the Avar Khaganate period, such as those discovered in Croatia (Fig. 22. b), the Danubian region, and Bulgaria. Such weapons are relatively common in graves associated with the Onogur Bulgars (Aralica – Ilkić 2012: 179-180).
Fig. 22. a. Sword from Stara Pazova – Surduk (after Peković 2015: Cat. No. 398); b. pallasch from Struga Nartska, Croatia (after Aralica – Ilkić 2012: Fig. 12. a–b).
Zdenko Vinski also incorporated a sword originating from the site of Tekija (Kladovo municipality) on the banks of the Danube into his corpus. The site is located near the Iron Gates in eastern Serbia, on the border with Romania (Vinski 1983: Cat. No. 12, Note 25). The find consists of a double-edged blade preserved with a crossguard of rhomboidal cross-section. It was discovered within a grave pit situated above a skeleton in the precincts of a former Roman military building; the find was originally dated to the 10th–11th centuries (Radišić – Bilogrivić 2024: 273). Typologically, this specimen was previously associated with Petersen Type X or interpreted as a Byzantine imitation of a West European sword (Janković 1983: 60; Radišić – Bilogrivić 2024: 278). However, given the morphology of the crossguard and blade, as well as the orientation of the skeleton within the grave, it is highly probable that the sword dates to the Migration Period, with a proposed chronology in the 5th–6th centuries (Radišić – Bilogrivić 2024: 273).
Similar characteristics are exhibited by a fragmentary sword find with a rhomboidal crossguard from the Danube bank at Batajnica (near Belgrade). This specimen is dated to the first half of the 5th century and is associated with a steppe-nomadic milieu (Radišić – Bilogrivić 2024: 265–271, 278, Figs. 2–4). A find of a bent blade with a rhomboidal crossguard from Horgoš (North Banat District) likely belongs to the same period. This sword was reportedly discovered in a grave alongside artifacts typical of 10th-century Old Hungarian cemeteries, such as stirrups and a bucket (Radišić – Bilogrivić 2024: 271). Milica Radišić contends that the sword may originate from the Migration Period, noting that ritually bent militaria are not characteristic of grave assemblages from the Hungarian Conquest period (honfoglalás). She further observes that the deposition of buckets in graves is also documented during the Migration Period (Radišić – Bilogrivić 2024: 271).
Fig. 23. Swords with rhomboid crossguards: a. Tekija (after Vinski 1983: Pl. VI. 1); b. Batajnica (after Radišić – Bilogrivić 2024: Figs. 3. b, 4. b); c. Horgoš (after Radišić – Bilogrivić 2024: Fig. 7. 4).
Gavro Škrivanić records two additional swords recovered during archaeological excavations in 1947 at the site of Caričin Grad (Justiniana Prima) near Lebane (Jablanica District) in southern Central Serbia. He attributed both to the 9th century (Škrivanić 1957: 41). Both swords were found with bent blades; one measures 98 cm in total length and is preserved with a rhomboidal crossguard and a ring-pommel (Ringknauf), while the second consists only of a tang and reaches a total length of 80 cm (Deroko – Radojčić 1950: Fig. 39). According to Škrivanić, the crossguard of the first specimen is an analogue to the Horgoš find, though he notes that while some authors consider it a Migration Period artifact, others date it to the 10th–11th centuries (Škrivanić 1957: 41). In light of the aforementioned data regarding swords with rhomboidal crossguards, it is plausible that the crossguard from Caričin Grad indeed dates to the Migration Period. The ring-pommel shows affinities with the Ringknaufschwert type of the 2nd–3rd centuries (Mráv 2006: Figs. 9–10; Miks 2009; Žákovský – Bárta – Hošek 2023: 23). It is also worth noting that the blade of the first sword in the work of A. Deroko and S. Radojčić appears to be single-edged. A compelling analogy is provided by a single-edged sword from an Early Avar grave at the Hungarian site of Biharkeresztes–Lencséshát (Mesterházy 1987: Fig. 6). The separate extant blade appears to be double-edged, of uniform width from the shoulders to the tip without discernible tapering, and likely lacks a fuller.
Two specimens of double-edged blades are known from the eastern Serbian site of Gamzigrad (Zaječar District). Both blades are relatively short, preserved only with their tangs, and were forged without fullers (Srejović et al. 1983: 156; Milenković 1992: 58). Their total lengths are 74.4 cm and 68.4 cm respectively (Srejović et al. 1983: Cat. Nos. 316, 317), with a maximum width at the shoulders of less than 4 cm (Srejović et al. 1983: 156). It is hypothesized that these blades may be products of Byzantine workshops (Srejović et al. 1983: 156), similar to the previously discussed blade from Lučica. Although some artifacts recovered from Gamzigrad are dated to the 10th–11th centuries, the unreliable stratigraphy of the find-spots and the absence of hilt components preclude a more definitive chronological assignment for these blades (Srejović et al. 1983: 154–156).
Fig. 24. a. Double-edged sword blade from Gamzigrad (after Srejović et al. 1983: Cat. No. 316); b. double-edged sword blade from Gamzigrad (after Srejović et al. 1983: Cat. No. 317).
The summarized catalogue data is available for download via the link provided below:
The distribution of double-edged swords from the 9th to 11th centuries within the territory of present-day Serbia.
Chronology and Historical Context
Given that the aforementioned swords consist exclusively of stray finds lacking detailed stratigraphic or contextual data, their assignment to a specific chronological horizon remains a significant challenge. Zdenko Vinski dated post-Carolingian swords from the territory of the former Yugoslavia (including several specimens from Croatia) to the 10th and 11th centuries (Vinski 1983: 7). Generally, the chronology of such artifacts within Eastern and Southeastern Europe is situated between the second half of the 10th and the 11th century.
Similarly, Milica Radišić attributes the post-Carolingian double-edged swords discovered in present-day Serbia to the period spanning the latter half of the 10th to the 11th century (Radišić 2022: 43). Their occurrence – alongside other bladed weapons and polearms whose origins are hypothesized to lie in Frankish or Kyivan Rus’ workshops – may be linked primarily to the political-military activities of the Carolingian Empire or Great Moravia during the 9th century. Furthermore, their presence may correlate with the southward and southeastward expansion of the early Hungarian monarchs into Central Europe; one prevailing hypothesis posits the involvement of foreign mercenary contingents within their military forces (Radišić 2022: 43). This theory is potentially corroborated by the uneven distribution of swords across the Serbian portion of the Pannonian Basin.
Valeri Yotov suggests that certain medieval weapons found in the Middle and Lower Danube regions are of Byzantine origin, indicating Byzantine influence in the area. He further notes that recent archaeological scholarship highlights potential links to the Avars, Bulgarian dominance north of the Danube, and the migration of the early Hungarians into the Danube region (Yotov 2011: 35–36).
Acknowledgements
In conclusion, I would like to express my gratitude to Milica Radišić (Institute of Archaeology, Belgrade) and Aleksandar Šalamon (National Museum, Zrenjanin) for the information they graciously provided regarding the sword from Novi Bečej. My thanks also go to Vladimír Turčan and Rado Čambal (Slovak National Museum – Archaeological Museum, Bratislava) for their expert consultation concerning the sword recovered from the Danube and the specimen from the Ritopek site. Finally, I am grateful to Tomáš Vlasatý for the opportunity to publish this contribution on the platform of the Forlǫg project.
Bibliography
Aleksić 2004 = Алексић, Марко (2004). Посткаролиншки мач из Банатског Брестовца. In: Гласник Српског археолошког друштва 20, 251-265.
Aleksić, Marko (2007). Medieval Swords from Southeastern Europe. Material from 12thto 15th century, Belgrade.
Androschchuk, Fedir (2014). Viking Swords: Swords and Social aspects of Weaponry in Viking Age Societies, Stockholm.
Aralica, Tomislav – Ikić, Mato (2012). Križnica sablje ili paloša Drugog avarskog kaganata iz Nina, in: Prilozi Instituta za arheologiju u Zagrebu 29, 167-184.
Bakay, Kórnel (1965). Régészeti tanulmányok a magyar államalapítás kérdéséhez, Pécs.
Bakay, Kornél (1967). Archäologische Studien zur Frage der ungarischen Staatsgründung. Angaben zur Organisierung des fürstlichen Heeres. In: Acta archaeologica Academiae Scientiarum Hungaricae 19, 105-174.
Bilogrivić, Goran (2019). Oružje i ratovanje u doba kneza Branimira. In: Bunčić, Maja – Dugonjić, Anita (eds.). Zbornik predavanja održanih tijekom Branimirove godine u arheološkom muzeju u Zagrebu, Zagreb, 52-58.
Demo, Željko (1984). Srednjovjekovni mačevi u muzeju grada Koprivnica. In: Vjesnik Arheološkog muzeja u Zagrebu XVI-XVII, 211-239.
Deroko – Radojčić 1950 = Дероко, А. – Радојчић, С. (1950). Откопавање Царичина града 1947 године. In: Старинар I, 119-142.
Geibig, Alfred (1991). Beiträge zur morphologischen Entwicklung des Schwertes im Mittelalter, Neumünster.
Głosek, Marian (1984). Miecze środkowoeuropejskie z X-XV w., Warszawa.
Hošek, Jiří – Košta, Jiří – Žákovský, Petr (2021). Ninth to mid-sixteenth century swords from the Czech Republic in their European context. Part II, Praha – Brno.
Janković, Milica (1983). Implements and weapons from the 9th-11th centuries found at Ključ Dunava, In: Balcanoslavica 10, 55-70.
Jócsik, Kristián (2024). Včasnostredoveké meče z územia dnešného Slovenska v kontexte európskej archeológie, Nitra: Univerzita Konštantína Filozofa v Nitre.
Kainov, S. Yu. (2012). Swords from Gnёzdovo. In: Acta Militaria Mediaevalia VIII, 7-68.
Kainov – Zozulya 2022 = Каинов, С. Ю. – Зозуля, С. С. (2022). Мечи типа «W» на территории Древней Руси. In: Археологические вести 32, 107-129.
Kazakevičius, Vytautas (1996). IX–XIII a. baltų kalavijai, Vilnius.
Kirpichnikov 1966 = Кирпичников, А. Н. (1966). Древнерусское оружие: Вып. 1. Мечи и сабли IX–XIII вв, Москва – Ленинград.
Košta, Jiří (2021). Early Medieval Swords from the Czech Republic (9th–10th Centuries), Praha: Univerzita Karlova.
Košta, Jiří et al. (2019). Velkomoravské meče z Pohanska u Břeclavi a okolí – nová revize. In: Památky Archeologické 110, 173 – 235.
Kovač, Mario (2003). Ubojite oštrice – Hladno oružje na području Hrvatske od IX. do kraja XVIII. stoljeća, Gornja Stubica.
Kovács, László (1994-1995). A Kárpát-medence kétélü kardjai a 10. század 2. feléböl. Adattár. In: Communicationes Archaeologicae Hungariae 1994–95, 153–189.
Mesterházy, Károly (1987). Korai avar részleges lovastemetkezések Ártándról és Biharkeresztesről. In: Folia Archaeologica XXXVIII, 219-245.
Miks, Christian (2009). Ein römisches Schwert mit Ringknaufgriff aus dem Rhein bei Mainz. In: Mainzer Archäologische Zeitschrift 8, 129–165.
Milenković 1992 = Марина, Миленковић (1992). Случајни налаз мача из села Лучице код Пожаревца. In: VIMINACIVM – Зборник Радова Народног Музеја 7, 57-60.
Moilanen, Mikko (2015). Marks of fire, value and faith: swords with ferrous inlays in Finland during the late Iron Age (ca. 700-1200 AD), Turku.
Mráv, Zsolt (2006). Paradeschild, Ringknaufschwert und Lanzen aus einem römerzeitlichen Wagengrab in Budaörs. Die Waffengräber der lokalen Elite in Pannonien. In: Archaeologiai Értesítő 131, 33-73.
Oakeshott, Ewart (1964). The Sword in the Age of Chivalry, London – Melbourne.
Oța, Silviu (2008). Orizonturi funerare din Banatul istoric (secolele X-XIV), Sibiu.
Peković 2006 = Пековић, Мирко (2006). Археолошка збирка Војног музеја у Београду – Archaeological Collection of the Military Museum in Belgrade, Београд.
Peković 2015＝Пековић, Мирко (2015). Збирка праисторије, антике и средњег века Војног музеја у Београду, Београд.
Petersen, Jan (1919). De Norske Vikingesverd: En Typologisk-Kronologisk Studie Over Vikingetidens Vaaben, Kristiania.
Petri, Ingo (2017). Vier VLFBERHT-Schwerter aus der Sammlung des Museums für Vor- und Frühgeschichte, Staatliche Museen zu Berlin. In: Acta Praehistorica et Archaeologica 49, 137–187.
Pinter, Karl Zeno (2007). Spada și sabia medievală în Transilvania și Banat (secolele IX – XIV), Sibiu.
Rabovyanov, Deyan (2011). Early Medieval Sword Guards from Bulgaria. In: Archaeologia Bulgarica XV, 2, 73-86.
Radić, Mladen (1991). Mač tipa „X“ iz privatne zbirke u Osijeku. In: Osječki zbornik 21, 77-85.
Radišić, Milica (2022). Double-Edged Swords and Polearms of Western Origin in the Territory of Present-day Serbia: an Overview and Reinterpretation. In: FUROR NORMANNORUM Research on 9-11th Century Double-Edged Swords: Archaeological and Metallurgical Approaches, Budapest, September 19–23th 2022, 43-45.
Radišić, M. O. – Bilogrivić, Goran (2024). Pommelless Swords with Rhomboid Crossguards in the Early Middle Ages – a Research Contribution from the Southern Carpathian Basin. In: Starinar LXXIV/2024, 265-282.
Rejholcová, Mária (1976). Pohrebisko z 10. a 11. storočia v Hurbanove-Bohatej. In: Slovenská Archeológia XXIV-1, 191-234.
Srejović et al. 1983 = Срејовић, Драгослав et al. (1983). Гамзиград. Касноантички царски дворац, Београд.
Stojić, Goran (2022). Живот на простору Виминацијума од 5. до 12. века. In: Vivere in Vrbe Viminacivm Tom 1, 429-471.
Szentpéteri, József (2002). Archäologische Denkmäler der Awarenzeit in Mitteleuropa, Band III, Budapest.
Szőke, B. M. (1992). Karolingerzeitliche Gräberfelder I-II von Garabonc-Ófalu. In: Antaeus 21, 41-203.
Škrivanić 1957 = Шкриванић, Г. А. (1957). Оружје у средњовековној Србији, Босни и Дубровнику, Београд.
Tomsons, Artūrs (2018). Zobeni Latvijas teritorijā no 7. līdz 16. gadsimtam, Rīga.
Turčan, Vladimír (1997). Stredoveký meč vybagrovaný z Dunaja. In: Zborník Slovenského Národného Múzea XCI – Archeológia 7, 79-83.
Vinski, Zdenko (1977-1978). Novi ranokarolinški nalazi u Jugoslaviji. In: Vjesnik Arheološkog muzeja u Zagrebu X-XI, 143-208.
Vinski, Zdenko (1983). Razmatranja o poslijekarolinškim mačevima 10. i 11. stoljeća u Jugoslaviji. In: Starohrvatska prosvjeta: glasilo Hrvatskoga starinarskog družtva u Kninu, ser. 3, 13, 7-64.
Viskupič, Michal (2023a). Early medieval swords from the territory of Romania (8th-11th century). In: Project Forlǫg – Reenactment and science [online]. [2026-01-22]. Available from: https://sagy.vikingove.cz/en/early-medieval-swords-from-the-territory-of-romania-9th-11th-century/
Viskupič, Michal (2023b). Early medieval double-edged swords from the territory of Bulgaria (800 – 1100). In: Project Forlǫg – Reenactment and science [online]. [2026-01-22]. Available from: https://sagy.vikingove.cz/en/early-medieval-double-edged-swords-from-the-territory-of-bulgaria-800-1100/
Yotov, Valeri (2011). Byzantine Time Swords (10th –11th Centuries) in Romania. In: Ţiplic, I. M. (ed.). Militaria Mediaevalia in Central and South Eastern Europe Sibiu, October 14th – 17th, 2010, Sibiu, 35-45.
Yotov, Valeri (2022). The find of Sfântu Gheorghe, Covasna County (1943): questions, that might be raised. In: Yotov, Valeri et al. (eds.). Swords in Byzantium, Varna, 189-198.
Zlatkov 2014 = Златков, Методи (2014). Меч от Градешница – Sword from Gradeshnitsa. In: Цар Самуил (✝ 1014) в битка за България, Българска академия на науките – Национален археологически институт с музей, София, 2014 – Tsar Samuil (✝ 1014) in battle for Bulgaria, Bulgarian Academy of Sciences – National Institute of Archaeology with Museum, Sofia, 2014, Sofia, 142-143.
Žákovský, Petr – Bárta, Patrick – Hošek, Jiří (2023). Středověké meče v českých zemích. Meč jako technologický, kulturně historický a archeologický fenomén, Brno.
Żak, Jan (1960). Problem pochodzenia mieczów tzw. “wikińskich” na ziemiach zachodniosłowiańskich, głównie polskich. In: Archeologia Polski, Tom 4, Zeszyt 2, 297–344.