An Inscribed Silver Spoon from Ichtratzheim (Bas-Rhin)

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. An Inscribed Silver Spoon from Ichtratzheim (Bas-Rhin) Svante Fischer, Martin Hannes Graf, Carole Fossurier, Madeleine Châtelet, Jean Soulat


Background
In the summer and autumn of 2011 Carole Fossurier of the Institut nationale de recherches d'archéologie preventive (INRAP) led the excavation of some 55 inhumation burials in a cemetery on the site known as "Niederfeld" in Ichtratzheim (Bas-Rhin), currently a suburb 15 km south of Strasbourg in Alsace, France (see Fig 1). The excavation revealed that the cemetery has a burial sequence that extends from the late 6th century until the early 10th century (Fossurier 2013). During the excavation of grave 108, a silver spoon was found in situ between the legs of the buried individual. Runes were soon discovered on both sides of the blade. At the request of Fischer, Fossurier brought the spoon to the Musée d'Archéologie Nationale (MAN) prior to conservation. Together with Soulat, Fischer could then conduct a proper runological autopsy during the XXXIIe Journées internationales d'Archéologie mérovingienne in September 2011.
The autopsy was performed with a digital monocular microscope, kindly dispatched to MAN by the Runology Forum of the Department of Scandinavian Languages of Uppsala University. After the autopsy, the IRF dispatched the digital documentation to Graf of the Schweizerdeutsches Wörterbuch in Zürich in order for him to interpret inscription C. A year later, Fischer, Graf and Soulat presented the runic inscription at the XXXIIIe Journées Internationales d'Archéologie mérovingienne in Strasbourg in September 2012 (Fischer et al. 2012), while the INRAP exhibited the objects in conjunction with the inauguration of the conference. However, during the conservation of the spoon in 2012, corrosion was removed from the surface, revealing further branches on some runes on the backside of the spoon (inscription C) while also causing additional damage to some of the more corroded areas on the front side (inscription B). It was thus deemed necessary to perform a second autopsy in order to compare the inscriptions before and after the conservation. Fischer and Graf carried out the second autopsy with the assistance of Châtelet at the INRAP offices in Strasbourg in July 2013.

The Context: Ichtratzheim
The place name Ichtratzheim belongs to the -heim type. It is indicative of the Germanic settlement along the left bank of the Upper Rhine that followed in the wake of the Barbarian invasions of the early 5th century (Hoeper 1997). Ichtratzheim is located west of the Ill River, and thus situated on loess type soil, next to and just above the "Ried", a lower area prone to seasonal flooding due to its proximity to the Rhine River. The floods made these lands fertile pastures and well suited for the elevation of livestock in a biotope that also harbored a river food chain, offering a varied diet.
The area appears to have been settled at a relatively late date in the Post-Roman era, in the late 6th century. This conclusion is based on the stratigraphic chronology of the Roman period ditches surrounding the Niederfeld cemetery, that were no longer maintained after the late 4th century. Thus, centuries after these had fallen into disrepair, the Niederfeld row-grave cemetery was established in Ichtratzheim (Fossurier 2013, 37-39).
The combination of Carbon-14 dating, chronology of grave goods and burial types of the 54 subsequent graves show continuous use into the 10th century (Fossurier 2013, 84, Fig. 148-151).
The onset of the cemetery may well have been related to considerable demographic growth that caused the local landowners along the left bank of the Upper Rhine to create new farmsteads with their own burial grounds. The remnants of a small building and the successive agglomeration of later graves around the earliest grave 108 in the same east-west direction may indicate Christianization of the cemetery. It is not inconceivable that those who buried the woman in grave 108 had something to do with this process, as her grave is not only the richest but also the oldest within the excavated part of the cemetery, serving as a focal point while never being plundered.
Grave 108 -Structure, Gender, Grave Goods and Social Status Grave 108 was a west-east oriented (262°) rectangular chamber grave measuring c. 2.7×1.8 m (see Figs 2-3). It belongs to the first burial sequence of the excavated part of the cemetery. It was located at an attractive focal point of the burial ground and subsequent burials came to surround it. The buried person in grave 108 was positioned on the back along the northern wall of the chamber with the head lying in the west. The skeletal remains are considered to be those of an adult female according to the gender characteristics of the burial goods and the anthropological examination in situ.
Unfortunately, the chamber grave had been disturbed in its northwestern corner by an intersecting shaft that subsequently caused the destruction of much of the upper body of the buried and many of the grave goods in this part of the grave. It also led to many objects being dislocated from their original positions, notably glass beads that had been around the neck and fragments of one of the two square headed brooches that had been on the hip.
There were several preserved grave goods on and around the body. These include a silver spoon of the general type usually attributed to Byzantine workshops of the 6 th and 7 th centuries (such as the two examples known from Sutton Hoo, cf. Bruce-Mitford 1975, 1978, 1983Kitzinger 1992;Drauschke 2007), two gilt silver square headed brooches of Kühn's type 32 Schwarzrheindorf (Kühn 1940, 275-278;Siegmund 1998, 54), a massive silver kolben arm ring of Wührer's type A.2.1 (Wührer 2000, 16 f.), a gold finger ring of Hadjadj's type 3b (Hadjadj 2008, 56) with important parallels in the Cologne Cathedral burials, a silver buckle and appliqués for a châtelaine, to which one can also associate a rock crystal pendeloque with silver fittings, and the pierced discoid foot of a 5th century Roman translucent glass vessel, possibly re-employed as a spindle whorl while attached to the chatelaine (see There was also a wooden stave bucket with copper alloy hoops located in the west-northwestern part of the grave, and a copper alloy basin in the southeastern corner. In and around these two containers, there were an unusually wide variety of foodstuffs (Fossurier 2013, 102, fig. 93). Especially south of the copper alloy basin there was a substantial distribution of many bones from different animals, notably an entire suckling pig (Sus domesticus), ribs of a beaver (Castor fiber), a gutted and deboned pike (Esox lucius), and two chickens (Gallus domesticus). By contrast, the hooped bucket in the West was found to contain some 3,000 fragments of frog and toad bones (Bufo calamita, Bufo bufo, Rana sp.). The food offerings are quite in accordance with the location of Ichtratzheim on the border of the "Ried" biotope.
The outstanding assembly within the chamber grave indicates a high social status, which would easily qualify the burial as pertaining to the top ranks of Christlein's Qualitätsgruppe C (Christlein 1973). It is certainly among the richest known Merovingian Period female graves along the left bank of the Upper Rhine (Fossurier 2013), and graves containing runic inscriptions in general , Tab. 7-10).

The Silver Spoon
The silver spoon measures 234 mm in length, the blade is 75 mm, the handle 159 mm. It weighs 50.7 g. It is easily classified as an instrumentum domesticum (cochlearia) and the general distribution pattern as grave goods in the Merovingian Period is rather well known (Kitzinger 1940, Dannheimer 1965, Kaske 1967, Hauser 1992, Drauschke 2007. The spoon itself is therefore not unique in anyway but must be regarded as an integral part of a general phenomenon of acculturation and transformation from the world of the Late Roman Empire to the Germanic successor kingdoms of the Early Middle Ages. Nor is this the first runic silver spoon to be discovered in a Merovingian Period burial context. A collection of grave goods from the row grave cemetery of Oberflacht, Kr. Tuttlingen, Baden-Württemberg, includes a silver sieve spoon with a runic inscription (Schiek 1992, Düwel 2002c. The Oberflacht runic inscription is difficult to decipher, though, and neither the inscription nor the decontextualized object offer any additional information for the study of the Ichtratzheim spoon. Typologically, the Ichtratzheim spoon has some, but not all, traits that pertain to Hauser's subgroup 4.2 (Hauser 1992, 32). A major obstacle to a more precise typological attribution is the fact the spoon has a hexagonal handle, a feature missing within group 4 but instead typical of Hauser's group 7. Hauser argues that group 4 in general derives from Byzantine workshops in the Near East and dates to the late 6 th and early 7 th century, whereas Drauschke (2007) points out that Roman silver spoons were produced in workshops in Italy and the Western Balkans in the 6 th century. Hauser's subgroup 4.2 fits well into this earlier 6 th century horizon, given the distribution pattern (grave finds from Thuringia, and stray finds from Dalmatia and Italy), ornaments and the lower weight within the subgroup. An earlier 6th century date and a production much closer to Ichtratzheim can therefore not be excluded but is by no means certain.
As for the interpretation of the silver spoons in graves, it was previously argued that the presence of this type of spoon as grave goods was a certain sign of Christian belief (Dannheimer 1965, 78). This is no longer considered as certain (Drauschke 2007, 57). Still, it is not inconceivable that the spoons were given as tokens of remembrance of rites of passage such as baptisms, communions, etc.
The Inscriptions: Transliteration, normalization and translation There are three different inscriptions (A-C) and two engravings (1-2) on the spoon (see Figs 6-15, Plate I). The two engravings are representations of a vine-leaf and a cross, located on each side of the rectangular block between the blade and the handle. Inscription A is located on top of this block and was obviously part of the original design of a spoon. Inscription B is located inside the concave surface of the blade while inscription C is located on the reverse convex surface.
An initial observation is that there was plenty of room for the runic carver to use after the Latin inscription A and the two engravings 1-2 were in place. The oval shaped spoon blade has a maximum width of 33 mm, and extends some 75 mm from the handle. But the carver exercised considerable discipline and produced two very symmetrical inscriptions that rarely extended over 12 mm in height and 24 mm in length. In this respect, the runic inscriptions B-C look very different from the crowded inscriptions on contemporary brooches while bearing a certain resemblance to the inscription on wood from Neudingen-Baar grave 168 (cf. Waldispühl 2013, 289).
The overall impression of inscriptions B-C is that they were carved at the same time with only one instrument employed by a single individual. All runes have been carved top down and in a left-right direction. The carver appears to have been experienced and while perhaps making the occasional unnecessary mistake in haste, nevertheless rapidly corrected faults with relative ease. This suggests an unusual familiarity with runic carving.
There is no evidence of wear or scratches that have worn down inscriptions B and C. These are undoubtedly recent additions to an older object, in contrast to the very worn inscriptions on sword pommels such as Saint-Dizier and Grenay (Fischer et al. 2008, Truc et al. 2009). The individual runes are generally in pristine condition. It cannot, however, be determined if they were applied to the surface immediately before the burial or if they were simply well preserved because the spoon was kept safe in a pouch or purse protecting it from wear. The latter hypothesis seems very plausible, however, given the spoon's position between the legs of the buried woman where they could have been kept inside an organic material container of sorts in relation to the châtelaine next to her left side.
Engraving 1 Vine-leaf. Height: 10 mm. Maximum width: 6 mm. This symbol is probably related to the function of the spoon as a libation tool tied to the Eucharist and the liturgical use of wine. Engraving 2 Seriffed cross. Height: 8 mm. Width: 6.5 mm. This symbol is probably related to the liturgical use of the spoon. The cross is by far the deepest engraving and has been carved with a different tool than the one used for engraving 1 and inscription A.

Inscription A
Total length: 29 mm. There are eight characters, all clearly legible, of which seven are letters.

+ MATTEVS 1 8
Latin: + Matteus « + The Evangelist Matthew ». This inscription qualifies the object within the ideological realm of Christianity, cf. the two spoons in the princely boat grave of Sutton Hoo (Bruce-Mitford 1975, 1978, 1983. This is the oldest inscription on the spoon and it is contemporary with the engravings D-E. The depth of the incised grooves has facilitated subsequent niello ormentation. In conclusion, the engraving pattern of the Latin characters, except the S, looks very standardized perhaps due to the employment of an engraver in the silver workshop itself. The Latin letters are very symmetrical when compared to the more haphazard Greek letters on the two Sutton Hoo spoons.

+
The equilateral cross measures 3 mm, followed by a distance of 2 mm, this is normal except by the optic illusion of a larger space in between the two letters 4 and 5. This suggests that the addition of the cross was there to emphasize symmetry.
A The letter has the typical Merovingian Period pocket or chevron.
T T The two identical letters have the same distance between as do the other letters in inscription A, but the graphematic structure creates an optic illusion of a greater distance in between the two.

E
The letter has a slightly shorter uppermost branch, as in the case of the Chéhéry inscription (cf. Fischer and Lémant 2003).

U
The Letter is atypical in that it resembles an upside down u-rune. Rather one would have expected the shaped V or U or the current shape in reverse.

S
The reversed S is by no mean infrequent in this period as in the case of the Chéhéry inscription (cf. Fischer and Lémant 2003). However, this one has an asymmetrical shape when compared with the other letters, suggesting that the carver was more familiar the carving of the other letters.

Inscription B
Total length: 24 mm, average length of branches: 3mm, average height of staffs: 12 mm. There are six characters, all clearly legible except nr. 2. lapela 1 6 Pre-Old High German: lapela 'spoon'. This inscription defines the object itself. We can thus classify the item as a 'talking object' (cf. Fischer 2005, 63-64;Schmidt et al. 2011 159f.;Graf 2011, 232-235). l The main staff has a height of 12 mm. The branch protrudes perfectly in a 45° angle from the staff some 3 mm. It has been carved from left to right.
a This part of the spoon is no longer preserved in its entirety after the 2012 conservation, but was clearly visible in 2011. The main staff was slightly bent due to the concave surface. The upper branch bends slightly from the top but has been carved in left-right direction; the lower branch protrudes perfectly in a 40° angle from the staff some 3 mm.

p
The main staff is slightly tilted backward, which makes it appear taller than the other staffs in inscription B. Meanwhile the branches are shorter than usual, 2 mm. The presence of a second descending branch at the top can be explained as a possible mistake, where the carver proceeded to carve the second branch first, only to realize that it would meet a first branch at an incorrect position. Thus the carver corrected the error by proceeding more gently with a thinner branch from the top that met a second branch. This is not surprising as p was an unusual rune and infrequently carved. The otherwise well-experienced carver thus ran into some difficulties. The two lower branches have been carved with more force (especially the second one) and in some haste, since the branches cross each other. This is the oldest known example of p in a lexically meaningful runic inscription, as the 5 th century Hogganvik stone from Norway with the sequence apapa is not meaningful in the general sense (Knirk 2011), and the Anglo-Saxon 'p a d a' coin legend does not appear until the 7 th century (Blackburn 1991, Page 1999. e The rune is composed of two elements, two mirrored staffs with branches, which are separated at the bottom end of the staffs by a 3 mm wide space. First, there is a 12 mm high staff with a 4 mm descending branch to the right, second a 3 mm ascending branch connecting to an 11 mm high main staff to its right. The tiny but very visible 0.8 mm distance between the first, descending and the second, ascending branch may serve to vindicate the reading of the first sequence on the Pforzen belt-buckle as e=lahu (Düwel 1997), especially since the bind-rune e=l is by no means infrequent (although mainly on bracteates, cf. Nowak 2003), whereas the reading ltahu proposed by Looijenga (1997) and Nedoma (2004) remains questionable, especially the representation of lt = ilt. But in the case of Pforzen, this is admittedly a matter of interpretation, not reading, and has actually nothing to do with the reading of e in the case of Ichtratzheim. l The main staff is slightly bent, 11.5 mm. The branch protrudes from the staff some 2.5 mm. a The rune is composed of four elements in that a carver's error has produced a small, additional main staff. This rune is the least accomplished as the main staffs are in an asymmetrical angle. From the top, 5 mm long shorter staff, there are two 4 mm long branches descending, although the upper does not meet the staff. The second main staff is 10 mm long but begins in front of the short staff, some 2.5 mm down.
Plates I-II Drawings by Svante Fischer and Jean Soulat.
As a-apocope already starts in the 5 th century, regional and dialectal differences allow us to assume asynchronous developments. An alternative explanation with the form in an oblique case allows only the dative (< *lapilai > OHG leffile/-a) or accusative singular (< *lapilan > OHG Leffil) since other cases (especially in the plural) could not be accounted for in terms of sound changes Inscription C Total length: 22 mm, average length of branches: 3mm, average height of staffs: 12 mm. There are five characters, all clearly legible except nr. 4. abuda 1 5 a This rune was only partly visible in 2011, as the top of the staff and the two branches were covered by corrosion until the 2012 conservation. The main staff is 12 mm long and extends from the ridge to the edge. This has served as a norm for the ensuing runes, because no other staffs extend further down beyond the first one although there would have been plenty of room had the carver wished to proceed in such a more simple fashion. The upper branch is 3.5 mm and extends from the staff. The lower is 2.5 mm long. None of them meet the staff.
b Continental type b composed by one staff and two separate pockets. Main staff is 12 mm. Distance between pockets: 3.5 mm. The branches of the upper pocket measure 2.5 and 3 mm respectively. The lower pocket: 3 and 4 mm although the branches do not intersect.
u Classic shape. Main staff: 11.5 mm. The lower part of the branch extends 5 mm to the right and continues slightly below the main staff. d Classic shape. This rune is no longer preserved in its entirety after conservation. The rune is a single shape with four components. The first main staff measured approx. 12 mm. The second main staff is much longer, 17 mm. The first branch measures 17 mm, the second 18 mm. Two upper corners have branches and staffs intersecting, the lower right does not. The lower left one was not present during the 2011 autopsy.

a
The carver began to carve at the ridge with some force only to successively let the groove go all the more thinner down towards the edge. Length: 20 mm. The upper branch measures 3.5 mm but stands free, while the lower is perfectly cut from the staff in a 45° angle.

The Sequence abuda
The sequence abuda can, for the time being, not be associated with any known Gmc. (in this case [pre-] OHG) lexeme. Nor is it possible to link it to a proper name, as any potentially underlying basis is lacking. The identification with a proper name viz. an anthroponym seems however still the likeliest explanation given the statistical evidence of text type studies. Theoretically and structurally the syllable allows for a typical dithematic name, the segmentation of which is more complicated, however. A second element -buda would lead to Gmc. *buđō 'messenger' (zero-grade derivative or feminine motivation of Gmc. *beuđanan), the only other attestation of which is found in the name Arsiboda (arsiboda, Bezenye ["B"]-brooch (cf. Nedoma 2004, 207)). Although we cannot completely rule out this possibility -it would give proof of a non-aumlauted form (cf. OHG boto) -, one must admit that the thereby supposed old age of the form makes the interpretation highly unlikely. Furthermore an a-stem would not fit well with the presumed first element. From a superficial perspective, one might see some similarities with ubada in the Bad Ems inscription, which is, however, commonly interpreted as a writing error for gibada or the like (cf. Waldispühl 2013, 256;Nedoma 2004, 371).
Under the conditions of haplographic writing and the lectio difficilior one may assume a stem *amb-, which is possibly a secondary formation from *and-b through assimilation (cf. Förstemann 1900, 99), resulting in multiple possibilities, viz. *anđa 'opposite' (Prep.), *anđi 'and' (Conj.), anđjaz 'end' (noun, m.), *anđōn 'breath' or *anđō [only ON ǫnd] 'portal' (noun, f.). The reading of b is actually supported by a reconstructed m, which would have contributed to the preservation of b. The outcome of this analysis, the form **anđ-buđa, shows regular loss of the linking vowel after a multimoraic syllable (compare Leubwini < *leuƀ[a]winiz, Nordendorf-brooch, cf. Nedoma 2004, 363). The semantic meaning of such a name must remain uncertain, as the various outlined problems show.
An alternative segmentation abud-a does not lead to a satisfying solution either, as a two-stemmed formation is out of the question in this case, and the final -a can, if Germanic at all, thus only be the case ending of a weak feminine short name. Assuming a Latin ā-stem and again the rule that nasals are not written before homorganic consonants, leads to *abund-, derived from the verb abundare 'overflow', tentatively interpreted as a Latin salvation name with a meaning like 'overflowing in virtues'. The nasal cluster -nd-would moreover prevent breaking before a. A Latin female personal name and theonym Abundantia is attested. Therefore a short form *Abunda cannot be excluded from our considerations, especially in an in all likelihood Christian milieu such as Ichtratzheim. In addition, we also find male saints' and salvation names, such as Abundius and Abundantius until the Early Middle Ages, although more seldom than other names of the same type. In the still highly romanceinfluenced area of Raetia, names of the type Abundius can also be found from the High Middle Ages onwards that have left traces in various family names (cf. Huber 1986, 341).

Conclusion
The key question is how to interpret the assembly of the three inscriptions. If we accept that A denotes the Evangelist Matthew in Latin and that this inscription was an integral part of the spoon from the very onset, we are also forced to accept that this may well have been understood at the time for the addition of inscriptions B and C.
There must have been plenty of other Latin inscriptions to look at in Merovingian Period Alsace and there is no reason why such a prestigious textcarrier tied to Christianity could not have been correctly interpreted. But this hypothesis makes the interpretation of inscriptions B and C somewhat more complicated. If B simply identifies the object in a Germanic language it need not stand in conflict with the Latin personal name in A. But B and C are likely to be related, having been carved at the same time by one carver only. Inscription C is likely to be a personal name. So what to do with C? Is abuda to be conceived of as a mere mark of ownership, or can there be a subjective wish behind this name associated with the spoon? Or does the existence of two nomina (a nomen proprium C, and a nomen instrumenti B) even imply a genitive phrase expressing a possessive relation where C owns B?
The class of 'talking objects', objects with inscriptions that refer to themselves, has not yet been defined as such, although it has been dealt with in various recent publications (cf. Graf 2011, 232-235;Fischer 2005, 63-64;Düwel 2002a). The object in question is the second such inscription referring to a spoon (the first one being the medieval Narsarsuaq, Kunartoq (Greenland) wooden spoon, reading sbon (emendation from sboa; cf. Düwel 2002a, 280). Otherwise, combs and brooches are the most attested type of 'talking object' in the runic corpus, although we also find bones, a plane, a stool, a brick etc. It is certain that these objects (a) 'demanded' this kind of self-reference in some way and (b) the connection of inscription and bearer must have added to the value of the object, which was deemed more important than the sole use of the inscription and/or object.
Similar to the use of brooches and (belt-) buckles as amulets (cf. Schopphoff 2009, 66f.;Düwel 2002b, 34), the spoon had a function in the context of sacral and liturgical acts, alluding to a merit or ideological quality beyond its technical use. It must be pointed out that the enhancement of the spoon was probably due to the mentioning of its name (and thus its existence) as well as its visible, material presence; a sort of 'ban' of the material must be disregarded as an explanation, as this is a rather Roman concept. A general evaluation of the use of spoons in early medieval Europe material culture is however still lacking. For the time being, the two naming inscriptions and the self-reflecting 'spoon'-inscription allow us to assume the same "namen-und schriftmagische Kraft" for the Ichtratzheim spoon as recently proposed by Schmidt et al. 2011, 164 for the Frienstedt comb.