Inventory of five Swedish Late Bronze Age ‘scrap hoards’

This dataset contains information about metal objects and fragments of metal objects from five Swedish hoard finds from the Late Bronze Age. The main purpose of this data collection was to create a basis for a study of fragmented bronze objects in the so-called ‘scrap hoards’ from this period. The research focus in the dataset is on describing the incomplete objects in these depositions, and to determine/estimate to what degree the fragmentation is due to prehistoric actions (broken during the Bronze Age) or if it might be recent. The questions in focus for this study was which object types were fragmented versus not fragmented, and, to what degree the original object type could be recognized from the fragments. This study is presented in a scientific paper in English. This pilot study is part of a larger project run by Anna Sörman, studying the circulation, use and deposition of fragmented bronze objects, based on studies in north-western France and southern Scandinavia.

The dataset gathers information about the contents of these hoards which have previously been published by Andreas Oldeberg (1927, 1928, 1929, 1934), and the images and details about the finds available in the inventory catalogue (online) of the Swedish History Museum. In one case (the Härnevi hoard), the finds have also been studied first-hand by Anna Sörman, in the storage of the Swedish History Museum. As the focus of the project is on the metalwork objects, the few finds of other materials present in some of these hoards (stone, ceramics, organic materials) have not been included in the dataset.

Two source critical factors should be particularly highlighted regarding the quality of the data. Firstly, these hoards, found between years 1853-1926, have all been collected by private individuals. They have been found during agricultural labour and various groundworks. This means that they should not be expected to be complete, as for example small fragments are likely to have been overlooked. Secondly, the information in the dataset is mainly from secondary sources (with Härnevi as the only exception), which means that renewed primary studies of the material – with special focus on breaks and fragments – would probably lead to certain revisions. Finally it should also be pointed out that, in some cases, it has not been possible to evaluate the probable age of the fracture. This might be due to a lack of information in the secondary sources (published articles and inventory catalogue), or that the object and the patina of the break is too ambigious to be determined. These cases of uncertainty are shown in the dataset under the column “Old break(s)” which is given as either "y" (yes), "n" (no), "y?" (yes?), "n?" (no?), or "?" (indeterminable).

The information in the dataset is structured under the following columns:
1. No = Serial number in the table
2. Study no = Unique number for each object in this study, featuring the number of the hoard (1-5) followed by the number of the item in the hoard, where fragments from the same object are designated by the same number, followed by a serial sub-number. Study no "1.3.1" and "1.3.2." are thus two pieces of the third object listed from hoard 1 (Bräckan). For objects which have been broken in modern times, the fragments have not been given individual sub-numbers.
3. Object type = Functional category of the listed object, such as "Sword" or "Socketed axehead". Objects whose function/type has not been possible to determine are listed as "Unidentified".
4. Complete (yes/no) = Defines the object as complete or incomplete.
5. Old break(s) = Defines if the break(s) on incomplete objects are judged as ancient or recent, based on information in the documentation (often notes about patina, or accidental breaks at the time of discovery etc.)
6. >50% = An estimation of the proportion of the full object (more than 50% or not) represented by an incomplete object piece. >50% y/n is only filled in for objects with ancient breaks. It has not always been possible to estimate/determine and this information it is therefore not consistently provided.
7. Other damages = Notes of any other signs of damage on the object, such as "Bent" or "Crushed".
8. Comments about the fragmentation = Additional, descriptive information about break/fracture(s).
9. Likely to be deliberately fragmented = Interpretative classification based on an assessment of the compiled information about the incomplete objects (patina on break, fragment piece, other damages etc.). Note that not all objects that has "y" at Ancient break has "x" under the column Likely to be deliberately fragmented. Some cases have been excluded, either because they have been judged as too uncertain, or because older breaks can be interpreted as use-damage rather than intentional fragmentation.
10. Site = The place-name of the find site of the hoard to which the item belongs.
11. Parish = The name of the parish where the hoard find was made.
12. Museum no = Inventory number under which the object is stored.
13. Find year = The year of discovery for the hoard to which the item belongs.
14. Comments about date/type = Any further information about the dating or typological determination of the specific object.
15. General comment = Any further comments about the item.
16. Weight (g.) = The weight (in grams) of individual items (only for the Härnevi hoard). Show less..