THE SUCIDAVA TYPE OF BUCKLES ANDTHE RELATIONS BETWEEN THE LATE ROMAN

EMPIRE AND THE BARBARIANSIN THE 6IhCENTURY

BY

ALEXANDRU MADGEARU

Joachim Werner gave the name of this type of buckles because several pieces have been found at Celei-Sucidava when he published its paper (in 1955)'. These buckles have a sheld-shaped plate and are adorned bypierced work with cruciform or crescent motifs or with a human stylized face. Joachim Werner establishedtheir dating in second half of the 6th century. Dezs6 Csalldny wrote, few years after, a study on these buckles,focused especially on the pieces^with human face'. The number of buckles increased very much during the lastthree decades. Dan Gh. Tlodor3 and Syna lJenzr'a made two typologies in 1991 and 1992.ln 1992 ValentinVarsik also published a study about several types of buckles including Sucidava, but he did not proposed atypology'. In thg same year the buckles found in the Lower Danubian area were put together into a repertorybv Uwe Fiedlero.

The typologies drawn up by S. Uenze and Dan Gh. Teodor are very important for the study of theseobjects, but we consider possible a more accurate classification. We remind here that S. Uenze classified theSucidava buckles into five types:

a) - with a simple cruciform decor;b) - a cross with all the arms rounded;c) - a cross with the down arm rounded;d) - with tangent circles;e) - with a human stylized face.The classification nrade by D. Gh. Teodor is based on the sanre criterion of the ornament, but is more

coherent:a) - without ornarcnqb) - *ith a siryle cross;c) - with a cross md a crescenqd) - with a cr(Fs, a crcsoent md two latqal perforations;e) - vith c-lrscents, circb seg@ts md lateral perforations;O - witr ahrrm fre:g,t - with various ornanrents.

' I. Werner. in Kolner Jabbuch ftr Vor- und Frathges-chichte, I, 1955. p. 39-40, 45.

' D. Csalldny, nAca Antiqa ASH, 10,1962"1-3,p.5|77.' D. Gh. Tdor, Arheologia lvbtfuvei, 14, I 91, p. I I &125.o S. Uenze, Die sfitaftilen Befestigungen wn &dovec

(Bulgarien). Ergebnisse der deutsch-bulgarisch4sterreichischenAugrabwgen ( I 9 3 4 - I 9 3 7 ), Milnchen, 1992, p. I 8+ I 87, 59&599.

Arheologia Moldovei, XXI, 1998, p.217 - 222

5 V. Varsik, inSlovArch,40,1992, l, p.78,80.u U. Fiedler, Sndien zu Gn)berfeldern des 6. bis' 9.

Jahrhunderts an der unteren Donan, Bonn, 1992, p. 7l-73. Wealso made a repertory in our book (Al. Madgearu, ContinuitateSi discontinuitate culturald la Dunirea de Jos tn secolele VII-I?/{ Bucuregti, 1997), p. 39.

ALEXANDRU MADGEARU

we consider that the shield-shaped buckles without decoration do not belong to the Sucidava type

because this type is defined by a certain decoration. It is true that these buckles (for instance, the piece found in

the B 5g grave of g".o;tj are related to the Sucidava buckles and they could be even regarded as protot)*pes'but the typology could not include them'

on the other hand, if we take into consideration the criterion'of the oqlmeltatiron, than we mst obserw

that the Sucidava buckles are defined by combinations of three basic motifs: the cross, the crcsccnt d fu

human face. The most frequent is the cross. This is the reason why we propose the following typologl:

I. Buckles with cruciform pierced work:a) - with a straight cross, without crescent;b) - with a straight cross and with crescent;c) - with a rounded cross, without crescent;dj - with a cross with one or more rounded arms and with a crescent;ej - with a cross, a crescent and lateral perforations'IL Buckles without crucifurm pierced work:a) - with two vertical crescents;b) - with a crescent or an oval, without cross but with perforations;c) - with a human face;d) - with a very stylized and distorted human face'The advantage of our typology ir ttte-lroupinq o! all pieces ryittt ".-".tiorm decor into

a single class'

Several researches atTirmed that the suciJaia uockies rrad a ctrristian significlionj. It is obvious that thecraftsmen who produced these buckles in ttr" Ro-uo-Byzantine workshops considered the cmciform

ornament

a Christian symbol. In the same time, we could not be iure that all the men who bore them do this because of

their symbolic value. It is probable that some of these buckles appertained to T"1 who were not Christians and

who used them only because their functio".-ny tni"r"uron, *" d-o not agree the interpretation of the Sucidava

buckles found north of the Danube as Christian testimonies''In the pr"uioos, ,"p"noai"t, some belt accessories with similar ornament were also included' Because

these are other objects, with different function, we decided to make different catalogues, based on the sametypological criteria.

we classified all the buckles found in the Middle and Lower Danubian area about which we have enough

information. 1'f," t"11itory .toai"O includes also Transylvania and Moldavia as adjacent regions' The wholespreading area of the Sucidava buckles is very utle 1Cree"e, Minor Asia, crimea, Italy, spain, France and

even Britain). Most of them are concentrated in thJDanubian regon,s, especially on the lirnes'lt was alreadyobserved that these buckles are specific for the Danubian limesfr. The total number of the buckles we have

studied is I14. F.o- th"r", 25 weiefouna in scytnia Minor, 40 in Moesia Secunda, 13 in Dacia Ripensis, 5 in

Moesia prima, 12 in the North-DanuUian LriOieheads, 7 in Pannonia, 6 in Banat and near the confluenceMureg-Tis4 3 in Transylvania, 3 in Moldarria. fre add also 14 belt accessories related to the Sucidava buckles:8 from Scythi4 4 from Dacia Ripensis,2 from Orgova'

In this paper, we are studying onty *tn" aspects implied by the rys9ar9! of.these objects' It is only apreliminary inquiry. The first urp".i *u, iurt stated: the above presented classification' This was made taking

into consideration also ihe pieces founO"outsiJe the area studied in this papel. one could observe that the

number of varians tr not iery big. fhis means that these objec! wgre piodyced. into a small number ofworkshops. It is probable that somJof th"se wor[shops were loc"ated at constantinople. only in this way couldbe explained their rp."uaing everywhere in the ioman-Byzantine Empire and outside' Of course, otherworkshops existed in tfre L6wer Danubian provinces Jvleie many piecgs Yer: found'

The archaeologicalresearches did not discover anyone but it is knfwn a workshop,,at Caridin Grad, where some shield-shaped belt

accessories were produced (a type akin with the Sucidava one)"'The second problem dit"*r"o here is the chronology of tttr various variants of buckles of Sucidava type'

We consider that our tyfology could lighten the dating because the descendence of a type from another

t A. Petre, La romanitd en Scythie Miwure (f - VIfsiicles de notre ire). Recherches archiolqiques, Bucharest,1987, p. 68, pl. 122 bis/I90 b.

t I. Wemer, op. cit., p.40; D. Gh. Teodor, op. cit-'p.85.e See for instance D. Gh. Teodor,op. cit.,p.E5.

to V. Varsik, op. cit., P.89.rr B. Bavant, in Cariiin Grad II. Le quartier sud-ouest

de la ville haute, ed. par B. Bavaut, V. Kondi6, J' M' Spieser'Belgrade-Rome, 1990, P. 220'224.

THE SUCIDAVA TYPE OF BUCKLES

becomes more clear. For instance, it seems obvious that the human face buckles developed from thebuckles with perforations placed lateral to the cross (type I e). The perforations become the eyes of theface (type II c). The I e type evolved from the buckles with a less complicate decor (only with a cross orwith cross and crescent).

We propose the following evolution: the I a type (straight cross) is the most ancienl From this firsr qpederived the variants with cross and with crescent (I b, I d, I e). The type I c (with rounded crmses) could becreated in the same time with I a. From I d derived II c and from this one, II d. The typ6 tI a and II b erohedfrom the types decorated with cross and crescent.

D. Ciall6ny believed that the pieces adorned with a human face are earlier than the cnrciforms::. Ourtypology proves the contrary. Uwe Fiedler expressed the same opinion about the supposition of Csallinl. buthe considers that a cronology of all the variants is impossible''.

An important observation is that the II b type is an intermediary form between the Sucidava buckles andthe P6cs buckles. One could observe how the latter are evolving form the II b type. At the genesis of the Pecstype participated perhaps also the II d type. This one, II d, survived in the 7th century. In the 67 grave from thecemetery of Gy6d, the Sucidava buckle of II b type is associated with a P6cs buckle and with another of Git6rtype. Although the whole cemetery is generally dated in the second half of the 7th centuryro, this grave couldbe earlier, from the first or the second third of the century.

An absolute chronology of the various types could not be precisely established now, but we have sorneguiCe-marks. The single certain fact is the appearance of the variants with human faces (II c and II d) beforethe '80ies of the 6th century. Such pieces were found in closed contextes, dated before 580-590. At Sadovec-Golernanovo Kale was discovered a II c buckle (objet B 40). This fortress was destroyed around 586 orperhaps in 595-59615. Another evidence is given by a II d buckle from a treasure hidden into a water basin rnScupi, Macedonia. This town (moved from present Skoplj.e to Vodno after the earthquake of 518) vas ravagedby the Slavic invasion of 586 when the hoard was hidden'o. A buckle decorated with a human stylised face isrecorded in the fortress Kamen Brijag which was also dastroyed by the Slavic inroads in the '70ies or in thefirst years of Maurikios' reign". Finally, another evidence is brought by a belt accessory with human face fromHistria. It was discovered in the pavement of the road B from the central sector (not on the street but in thegrave). The street level.belongs to the reconstruction nnde after the fire of 593 (in the same pavement was alsofound a coin from 589)'o. The belt accessory should be dated prior to 593.

Therefore, the models with human face appeared before the '80ies. This is a very important evidence forour next statements.

The production of the cruciform buckles continued after the appearance of the human face variants. Insome cases these objets were discovered together, like at Beroe (B 45 grave)re. The cruciform bucklessurvived until the 7th century. A piece was found at Tropaeum, on the VI B level (dated in the first decadesof the 7th century;2o.

If we consider that the human face variants are later, than the location of this cathegory of buckles couldindicate chronological differences in the spreading of the Sucidava type of buckles.

We have no clear evidence about the beginning of the production of the Sucidava buckles. It ispossible that were first produced before the middle of the 6th century. A piece of I e type was found in thefortress Mokranjske Stene in Dacia Ripensis, into a site where all the coins are dated before the middle ofthe 6th centurv''.

It must be paid attention to the fact that the Sucidava buckles (like other types from the 6th-7th centuries)were pieces of the Roman-Byzantine military equipment. The use of the buckles and belts was established bv

r2J. Csalliiny, op. cit.,p.62. 17 A. Salkin, D, Toptanov, in Dobrudla. Etudes ethno-13 U. Fiedler, op. cit., p. 73. culturelles, Sofiq 1987, p. lZ-ll.raA.Kiss, CemeteriesoftheAvarPeriod(567-829),in 18 I.Stoian,M.Sdmpetru, inMateriale,g. 1g70,p. t89-Hungary 2. Avar Cemeteries in County Baranya, Budapest41977. p. 41, Taf. lxl67 14.

15 S. Uenze, op. cit., p. 599, nr. 59. See p.119,292-294

190, frg.9l4..

reA. Petre, op. cit., p. 69,p\. 126.20 I. Bogdan-Cataniciu, Al. Bamea, in Tropaeum

and4l7 forthedestructionofthefortress. Traiani,vol.l,Cetatea.Bucuregti, 1979,p. 192,frg. 174110.'1 .'6 I. Mikuleik, Staro Slapje so okoluite, Skopje, 1982, p. 2l M. Sretenovil,inCahiers des portes de Fer,2. 19g4.

2t9

5l-52, fig. 26. See also V. Varsik, op. cit., p.79. p.229-230, frg.21615.

220 ALEXANDRU MADGEARU

military regulations. The Pseudo-Maurikios Strategikon contains about belts2t. The militarl character of theSucidava buckles is proved by their spreading especially in the fortresses located on the /inre-s t'r rnside theprovinces.

In this case, the existence of some of pieces outside the Empire must be explained-There are not too many Sucidava 6uckles in Barbaricum, but their existence should be taken tni

consideration. We know 19 pieces discovered in the regions north of the Danube:1) Alba-lulia, Alba County, Romania: from the 6th century settlement located in the former Roman camp; type.

I b ; R . R. He i te t , S t tVn,37 ,1986,3 ,Z3g-240, t ig .211;D. Gh. Teodor , op . c i t . , p ' 12 l , f ig . 112;Y '

settlement. house 25; type I a; R. Popovici,Gh. Teodor, op. c i t . , 1991, p. 118, f ig. 113; Y '

3) Bratei, Sibiu County, Romania: from the Gepidic cemetery nr.

Varsik, op. cit., 1992.2) Bornis, Neam! County, Romania: fi'om a 6th century

Arheologia Moldovei,l2. 1988, p,249-251, frg. 112; D.Varsik, op. cit., 1992,p.90, nr. 5.

l2l , f ig.2l l .4) Diineqii, Vaslui County, Romania: from an inhumation grave. W-E oriented, tvpe I d: D. Gh. Teodor, op'

cit., p. 121, fig.2l7;Y . Varsik, op. ci!., p. 90, nr. 14.5) Gydd-Mdriahiegy, Baranya County,'Hu igiry: from an early Avar cemetery, grave 67; type II b; A. Kiss, op'

ci t . , 19' /7,p.41, Taf. IXl6714; V. Varsik, op. c i t . , p.90, nr ' 17'6) Hddmtezdvdiirhely-Kishomok,tsongr6d County, Il,ing9.Vj from a Gepidic cemetery, grave.65; type I b; I'

B6na, A l'atrbi du Moyen Ag, dApidrs et iombards dans le bassin des Carpathes, Budapest, 1976, p'115-116. f ig. 8; V. Varsik. op. c i t . , p. 90, nr. 18.

7) Jdnoshida-Triikirpuszta, Szolnbk County, Hungary: from an early Avar cemetery, grave 55 (female); type lb ; D . C s a l l 6 n y , o p . c i t . , 1 9 6 2 , p . 5 6 , n r . 9 , T a f . 2 l 8 ; Y . V a r s i k , o p - c i l . , p . 9 1 , n r . 2 1 -

8) Keszthely-Fendirpuizta I', Zala'bounty, Hungary: from a destroyed cemetery; type II b; K' Sigi, "ActerAnt iquaAl I { ' ,9 ,1961,3-4 ,p .343,Taf .XY l I I l2 ; V . Vars ik , op . c i t . , p .9 l ,n r '24 '

9) Keszthity-Fendkpuizta 2, Zala'County, Hungary: from a destroyed cemetery;type II b; K. S5gi, op. cit',Taf -XVII I /3; V. Varsik, op. c i t . , p.91,nr.25.

10) Ktdrafatv4, CsongrSd iounty,^Hungary: from an early Avar cemetery, grave 25; type I e; D. Csall|'ny, op.ci t . , p. 56, nr. 8, Taf. 2l7:Y.Varsik, op. c i t . , p. 9l , nr. 26.

11) Magy:arcsandd-bc;kdny, Csongritd County, Hungary: from a Gepidic cemetery. grayg_4; type_Il.b; D'' CJattany. Archaologische Denkmciler der Gepiden int Mittel donaubecken (154-568 v'2./, Budapest,1961,p.141, Taf. CLIXI6; V. Varsik, op. c i t . , p. 91, nr.41.

12) Molvin,'Banai, Yugoslavia: from an eariy Avai cemetery, grave 60 (female); type II b; D. Csall6nY, oP.ci t . ,1962:56, nr. l0;V. Varsik, op. c i t . , p.91, nr.45.

13) Noslac, Alba County Romania: f.o* a bepidic cemetery, grave l0; tyPg I b; M. Rusu, Dacia, N.S.,6,' 1962,p.279,f ig. i i25;D. Gh. Teodor, op. c i t . ,p.121,f ig. l /10;V. varsik, op. c- i t . , p- 91, nr. 47. ^, .

14) Pecica', Arad Couirty, Ronrania: from a deitroyedGepidic iemetery; type I d; D' Csall6ny, op'_cit',.79_61, p'744,Taf .CCXI I I [ t l :O.Gh.Teodor ,op .c i i . ,p .12 l , f ig .2 l3 ;V .Vars ik , op .c i t . , p .91 ,nr ' 57 i 'M. Barbu."Ziridava"

, 19-20, 1996, p. 9l-93.15) Szentes-Nagyhegt, Csongr6d County, Hungary: from a Gepidic cemetery, grave 29; type I d; D. Csall6ny,

op . c i t . ,1961, p . 50-51 , Ta f . XXV/ I3 ; V . Vars ik , op . c i t ' , p .91 ,nr .112.16) Szoreg 1, Csongr6d Couity, Hungary: from a GepidiCcemetery, grave XI; type I b; Csall6ny, op. cit., 1961,

p. 148-149, Taf. CLXXXYil I I2; V. Varsik, op. c i t . , p. 91, nr. 113'l\ S)oreg 2, Csongr6d County, Hungary: from a Gepidic cemetery, grave 103; type I a; Csall6ny, op. cit.,

1961, p. 165, Taf. CI-XXX/4; V. Varsik, op. c i t . , p.91, nr. 114.1S) $tefonc'el Mare, Bacau Counr,,-. Romania: from a 6th century settlement, house 17;type I d; I. Mitrea, C.' -Eminovici,

in Cronica cerietdrilor arheologice 1983-1992,A XXXI-a sesiune nalionald de rapoartearheologie, Bucuregti, 1997, p. 105-106, nr. 70.

19) Tatabdnyl, Kom6rom County,'Hungary: from a Longobardic cemetery; unknown type (not published); V.Varsik, op. c i t . , p. 91. nr. 115.Therefore, eight'pieces were found in Gepidic cemeteries and other four in early Avar graves. The Qmlle

graves from J6nosh*ida-T6tkdrpuszta and Mokrin could be ascribed to Gepidic women. The buckles from Gy6d

?r Maurici i , in Strategicon, X. 8. 1.3 (ed. H. Mihiescu,Buchares t . 1970, p . 315) .

THE SUCIDAVA TYPE OF BUCKLES

5 K. Horedt, Siebenbiirgen im Frilhmittel alter, Bonn, 27 A. Diaconescu. C. Opreanu, in Anuarul Institutului de19E6. p. 31. Istorie Si Arheologie, Cluj-Napoca,29,1989, p. 581-582.

:' G. Coman, Statornicie, continuitate. Repertoriul 28 I. B6na, op. cit.,p. 29.arfuologrc alludelului Vaslui,Bucure$ti, 1980, p. 100-103. 2n See A. Madgearu, in Balkan Studies, 33, 1992,2,

:i F. E. wozniak, in Ball

?22

3r See I. Mitre4 in SCIVA, 30, 1979, 2, p. 145-162;idem, in Pontica,2S-29, 1995-1996, p. 22841 for the latestfindings.

" A. Madgearu, in SCIVA, 44, 1993,2, p. l7l-183; V.Ibler. in Arheoloiki Vestnik, Ljubljana 43,1992, p. 135-148.

33 V. M. Butnariu, in 8SNR, 77'79, 1983-1985 (1986).nr. l3l-133, p.220,nr.81 (M[ndstirea); Gh. Poenaru, in Dacia,N.S., 37, 1993, p. 316, nr. 7l (Giegti ' from a hoard withunknown composition).

3n R. R. Heitel, op. cit., p.239-240.

ALEXANDRU MADGEARU

Davideni and Baclu-Curtea Domneasc[13|. The buckles found at Keszthely-Fendkpuszta (type I b) are alsoquite different from the pieces known in the Empire. It is known that at Kesthely and in.the,l.TPy:f"usurvived until the 7th century several Roman workshops for producing bronze and gold objects". The P6cstype buckles (also named Kisahely-P6cs) are deriving from ihe Sucidava type II b and are a creation of aPannonian workshops.

The buckles aie the belt accessories with human face (II c and II d) were discovered onll on the telrrtLrnof the Roman-Byzantine Empire, especially in Dacia Ripensis, Moesia Secunda and Sclthia. fh:ilabsence inBarbaricum is iignificant. We think^ thai this was due to the intemrptiol or. to tre diminishing of thepenetration of the"buckles, in the period when appeared the types II c and tr d. This fact could haPpen becausethe Byzantine-Avar wars of 573-602.

The 6thcentury gold Byzantine coins (solidi'l discovered north of the Danube (in Wallachia, Moldavia,and also in TransylvanTa and

-Banat) are issued especially under Justinian. Pieces from Justin II and Tiberius II

are known only from Mlnistirea and Giegti (bottrin Dimbovila County, in Central Wallachia)33.Radu R. Heitel wrote in his paper aboui ttre findings of Alba Iulia that the intemrption of the spreading of

the Suciclava buckles was one of ttr" first consequends of the Avar inroads; he dated this fact around 582when Baian conquered Sirmium3a. This opinion istonfirmed by all the North-Danubian discoveries.

The pieces found in later sites, datei in the last quarter oi'the 6th century and in the 7th century could belocal artifacts (type II b in the workshops of the Keszethely) or objects remained in use for a long time'

A task foi the future researches will be to establish a more iccurate chronolory of the Gepidic and Avargraves where the Suvidava buckles were found.

The careful inquiry ofthe Sucidava buckles are belt accessories could bring various data about theByzantine-Barbarian reiations and about the 6th-7th century civilization in the peripheral area of theRoman-Byzantine Empire.