Northern Sword Types of the First Millennium

What I give you here is an extremely abbreviated account of the "science" of classifying swords found (more or less) in the "North" in the (more or less) first half of the first millennium. Typically, each class has several subclasses, and even the swords in one subclass can look quite different.
The classification is based on a lot of features that are not always immediately apparent. They are also not always clear since "the book" on occasion states plain nonsense ¹⁾.
Unfortunately the book also refers a lot to (non-existent) "carburization" and makes many unsubstantiated claims with regard to the positive influence of pattern welding on mechanical properties like better "elasticity" or "vibrations damping".

The swords get younger in time as you go down the lists. More precise data to the dating of the swords are given farther down. The general system always used for giving periods from about 0 AD to 500 AD is shown below. It will be used throughout.


The time periods in the 1st half of the 1st millennium

12 groups with altogether 60 subgroups and variants to subgroups are needed to classify a grand total of 418 swords! In other words: While there some are common denominators for groups of swords, any given sword is rather unique. More! Two swords that look alike enough to put them in the same group / subgroup may be totally different with respect to their compositions. One might be a randomly piled together piece of inhomogeneous iron / steel, while the other one is a master piece of highly complex pattern welding.

Anyway, here are the 12 major groups. For each type there is a map showing where the swords were found.
I have distilled the comments from long texts and edited the maps to some extent. I might be a bit off on occasion. One reason for this is that the data given in the maps do not always match those in the text.
Whenever I refer to "piling" etc., I interpreted what I read (and I might be wrong).
Most of these swords were definitely or quite likely made in the Roman empire and "exported" to the Barbaricum.
It appears that the metallographic examinations were mostly made in the "East", e.g. in Poland, using the swords found there. It is thus possible that the simple make and bad quality often found just testifies that the guys there made "cheap" copies of the more complex Roman swords. They look like the original (and it is the look that defines the groups here) but have not much in common with the structure of the original. That is just my feeling, however, and I might be completely wrong.
The number in the right-hand corner of the following tables gives the number of known swords of the respective type.

Canterbury-Kopki			1st - 2nd century; B2 - C1a	25
Light to medium light Roman "export" spatha for the "Barbaricum". For horseback use; 72 cm - 88 cm. 2 subtypes; 2 variants Mostly found: 15 in the Przeworsk culture ²⁾ (Poland) area; 10 around the border area of Empire. 14 swords analyzed. 8 pattern welded, 2 "primitive", probably random piling. About 2/3 of these swords were made with a complex pattern welding, including chevron patterns. A few have incrustations and stamps. Roman origin is likely for most but not all.


Canterbury-Kopki swords; map of finds

Buch-Podlodóv			C1a; some C1b; B2	18
Roman spatha variant; long with broad blades. Very light - medium light. 3 subtypes; 3 variants. 12 from Przeworsk culture (Poland) 6 elsewhere, around Empire border. 8 swords analyzed. 7 with pattern welding, 1 "primitive". Complex pattern welding, including chevron patterns. Mostly from Roman smithies. Incrustations; typically 2 - 4 fullers


Buch-Podlodow swords; map of finds

Lachmirowice-Apa			Mostly C1a, B2, a few C1b	47
Medium weight late Roman spatha type. 3 subtypes; 3 variants. 39 Przeworsk culture (Poland), 3 in Danish bogs, 5 elsewhere. 21 swords analyzed; all from the East. 10 primitive (including two with incrustations and thus most likely Roman), 5 pattern welded, rest piling by face-welding. All kinds of cross-sections; with and without fullers.


Lachmirowice-Apa swords; map of finds

Folkeslunda-Zaspy			Mostly C1b	23
Medium weight spatha. 2 subtypes; 1 variant 2 in Danish bogs; rest in the "East". 21 swords analyzed; all "primitive" except the 2 pattern-welded ones from Danish bogs. Simple cross-sections; typically no fullers.


Folkeslunda-Zaspy swords; map of finds

Lauriacum-Hromówka			Mostly C1b or a bit later	11 (?)
Medium length, quite broad. Medium weight 2 subtypes; 1 variant All over; 1 in Danish bogs. Some metallurgical analysis, pretty much all swords are pattern welded. Typically many fullers; some with asymmetric cross-section. Many with incrustations or other adornments


Lauriacum-Hromówka swords; map of finds

Woerden-Bjaers			All C1b	36
Late Roman medium (to heavy) weight spatha. Slashing type 2 subtypes; 2 variants 25 from Danish bogs, rest all over (the East) 2 (eastern ones) analyzed. 1 primitive, 1 pattern welded. Most of the rest obviously pattern welded including very complex patterns All kinds of cross-sections, typically with fullers, some asymmetric. Often stamps and incrustations.


Woerden-Bjaers swords; map of finds

Vimose-Illerup. Long, pointed. Mostly on the light side.			Almost all C1b	102
Longish narrow blades ("rapier like"), middle weight. 6 subtypes; 4 variants. Mostly Danish bogs, otherwise Scandinavia and Barbaricum. 8 analyzed. Primitive piling, good piling and pattern welding occur. More than 50 % with recognizable compelx pattern welding. Quench hardening in some cases.


Vimose-Illerup swords; map of finds

Nydam Kragehul			Mostly C2	58
Similar to Woerden-Bjaers; lighter. 3 subtypes; 2 variants Mostly Nydam. Otherwise Sveden and the East. 8 analyzed (mostly form the East). 2 with primitive piling, rest complex pattern welding. Cross-section mostly symmetric; with and without (at most 2) fullers. Some with incrustations and / or stamps.


Nydam-Kragehul swords; map of finds

Snipstad			C2 (6) - D (3)	8
Similar to Lauriacum-Hromówka. Medium weight; long. 1 subtype; 1 variant. No metallurgical analysis yet (2006). Mostly from Esbjol;l. Asymmetric blades; 2 - 6 fullers per side. Often complex pattern welding visible.


Snipstad swords; map of finds

Voien-Hedelisker			C2 - D1; even D2	13
Akin to Vimose-Illerup. Long, pointed. Mostly on the light side. 2 subtypes; 1 variants. 1 (Eastern) sword analyzed: uniform homogeneous steel (!) Mostly from Danish bogs (Nydam, Esbjoel, Illerup); a few from Norway; the East. No fullers. Most swords made by more complex piling or (simple (?)) pattern welding.


Voien-Hedelisker; map of finds

Esbjol;l-Sarry			Mostly D	106
Late Roman, Migration spatha; pointed, rather heavy. Forerunner of heavy Merovingian swords. 4 subtypes; 6 variants. 11 swords analyzed: Simple and complex piling; some (simple) pattern welding. Some swords are made form rather uniform steel (!) Mostly Danish bogs (70%); Rest from Norway; Przeworsk culture, Empire border Symmetric (six faces) no fullers. No stamps or incrustations


Esbøl-Sarry; map of finds

Osterburken-Vrasselt			C3 - D1	16
Long, broad, unadorned and heavy - the sword for the non-nonsense guy? 2 subtypes; 2 variants. No metallurgical analysis yet (2006) Towards the South: all over in bog. Symmetric; often simple (lentil) + shallow broad fuller. No stamps or incrustations. At least 2 swords have patterns ; most are heavily corroded


Osterburken-Vrasselt; map of finds

Here is a direct comparison of these sword types. I picked one sword of each group, more or less at random, from the selections given in the book. Note that there is a lot of variance within just one group. Pick other examples and the picture would look quite different


The 12 basic sword types Large picture
Source: Illerup Ådal; Vol. 11, 12

Here is a full map for Canterbury-Kopki swords with some areas / cultures higlighted ( by me).


Where Canterbury-Kopki sword were found Large picture
Source: Based on a figure in Illerup Ådal; Vol. 11, 12

The dating of the swords relies mostly on contextual evidence, e.g. what kind of hilt was found with a sword. The fashion in hilt shapes and material changed; here is a kind of temporal map:


Typical sword hilts for certain periods Large picture Real ones
Source: Based on a figure in Illerup Ådal; Vol. 11, 12

This link shows how the fashion in hilts developed after about 400 AD.

Most of the 418 swords discussed here have been dated to one of the periods given above or to, for example "B2 or C1a", with precise percentages given. Below is my attempt to turn many pages of prose into a graph. The color gives the type, the length of the lines the percentage assigned to a certain time slot. The position of the lines on the time scale is more or less arbitrary within the proper time slot(s).


Sword types and time periods Large picture
Source: Based on a figure in Illerup Ådal; Vol. 11, 12

¹⁾	One example First, there is not a single blade with the "tailored" look. Second, the text says: K1: Older blade with strong tapering. K2: Newer blade with tapering. ????
²⁾	Do we want to know anything about cultures with names like that? Well - here are a few data form Wikipedia: The Przeworsk culture is part of an Iron Age archaeological complex that dates from the 3rd century BC to the 5th century AD. It was located in what is now central and southern Poland. It takes its name from the village near the town Przeworsk where the first artifacts were found. The main characteristic feature of the Przeworsk culture are burials. These were mostly cremations, with occasional inhumation. Warrior burials are notable, which often include horse-gear and spurs (and swords). Some burials are exceptionally rich, overshadowing the graves of Germanic groups further west, especially after 400 AD. Pottery and metalwork are often rich and show a great variety. The culture declined in the late 5th century, coinciding with the invasion of the Huns. Other factors may have included the social crisis that occurred as a result of the collapse of the Roman world and the trade contacts it maintained with peoples beyond its borders.