Zbirka / Series E-Monographiae Instituti Archaeologici Sloveniae 14 Uredniki zbirke / Editors of the series Jana Horvat, Benjamin Štular, Anton Velušček Benjamin Štular GRAVE ORIENTATION IN THE MIDDLE AGES A Case Study from Bled Island Reviewed by Stefan Eichert, Jiří Macháček Editor and DTP Mateja Belak Translation Edisa Lozić, Benjamin Štular Publisher Založba ZRC Represented by Aleš Pogačnik Issued by ZRC SAZU, Inštitut za arheologijo Represented by Anton Velušček Published with the support of ZRC SAZU Represented by Oto Luthar Front cover design Benjamin Štular Front cover photo Mystic sunrise at lake Bled with duck (https://commons.wikimedia.org/wiki/File:Mystic_sun rise_at_lake_bled_with_duck_2013.jpg, Dreamy Pixel, CC BY 4.0 , via Wikimedia Commons) Ljubljana 2022 The first e-edition is freely available in e-form (pdf) under the Creative Commons 4.0 CC-BY-NC-SA. https://doi.org/10.3986/9789610506331. This book is a substantial revision of chapters 3−8 of the book Srednjeveški Blejski otok v arheoloških virih / Medieval archaeology of Bled Island, Benjamin Štular (ed.), Opera Instituti Archaeologici Sloveniae 42, 2020, Založba ZRC, https://doi.org/10.3986/9789610502609. Kataložni zapis o publikaciji (CIP) pripravili v Narodni in univerzitetni knjižnici v Ljubljani COBISS.SI-ID 127207427 ISBN 978-961-05-0633-1 (PDF) The author acknowledges the financial support from the Slovenian Research Agency (research core funding Nos. P6-0064 (B) and J6-9450). GRAVE ORIENTATION IN THE MIDDLE AGES A Case Study from Bled Island Benjamin Štular LJUBLJANA 2022 CONTENTS 1 Introduction ......................................................................................................................................................................... 7 1. 1 Introduction to the second edition ........................................................................................................................ 7 1.2 Why the second edition ............................................................................................................................................ 7 1.3 Foreword to the first edition .................................................................................................................................... 8 1.4 Description of the first edition ................................................................................................................................ 9 2 Data: documentation from the 1962 to 1964 excavation ............................................................................................. 11 2.1 Introduction ............................................................................................................................................................ 11 2.2 Plan drawings of horizontal surfaces ................................................................................................................... 11 2.3 Section (Profile) drawings ..................................................................................................................................... 12 2.4 Artefacts (small finds) ............................................................................................................................................ 12 2.5 The value of documentation for stratigraphic analysis ...................................................................................... 15 2.6 Conclusions ............................................................................................................................................................. 15 3 Methodology ...................................................................................................................................................................... 17 3.1 Introduction ............................................................................................................................................................ 17 3.2 Georeferencing ........................................................................................................................................................ 17 3.3 Geomorphology of Bled Island ............................................................................................................................. 21 3.4 Stratigraphic analysis .............................................................................................................................................. 24 4 Site stratigraphy ................................................................................................................................................................. 29 4.1 Introduction ............................................................................................................................................................ 29 4.2 Phase 1 ...................................................................................................................................................................... 30 4.3 Phase 2 ...................................................................................................................................................................... 32 4.4 Phase 3 ...................................................................................................................................................................... 35 4.5 Phase 4 ...................................................................................................................................................................... 36 4.6 Phase 5 ...................................................................................................................................................................... 36 4.7 Prehistory ................................................................................................................................................................. 37 4.8 Preliminary absolute chronology ......................................................................................................................... 40 5 Building analysis ................................................................................................................................................................ 41 5.1 Introduction ............................................................................................................................................................ 41 5.2 Ground plans ........................................................................................................................................................... 41 5.3 Interpretation .......................................................................................................................................................... 45 6 Orientation of the Medieval churches ............................................................................................................................ 47 6.1 State-of-the-art ........................................................................................................................................................ 47 6.2 Bled Island church .................................................................................................................................................. 49 7 Orientation of the graves .................................................................................................................................................. 51 7.1 State-of-the-art ........................................................................................................................................................ 51 7.2 Methodology ........................................................................................................................................................... 53 7.3 Methods used in the Bled Island case study ........................................................................................................ 55 7.4 Bled Island case study ............................................................................................................................................. 56 7.5 Mortuary landscape ................................................................................................................................................ 61 7.6 Conclusions ............................................................................................................................................................. 61 8 Spatial organisation of the cemetery area ...................................................................................................................... 65 8.1 Introduction ............................................................................................................................................................ 65 8.2 Orientation axes of the Early Medieval cemetery .............................................................................................. 65 9 Spatial analysis of grave attributes ................................................................................................................................... 69 9.1 Introduction ............................................................................................................................................................ 69 9.2 Pits ............................................................................................................................................................................ 69 9.3 Backfil s .................................................................................................................................................................... 70 9.4 Reburials .................................................................................................................................................................. 72 9.5 Arm posture ............................................................................................................................................................ 73 9.6 Conclusions ............................................................................................................................................................. 74 10 Absolute chronology of the Bled Island cemetery ...................................................................................................... 75 10.1 Introduction .......................................................................................................................................................... 75 10.2 Early Medieval cemetery ..................................................................................................................................... 76 10.3 High Medieval cemetery ...................................................................................................................................... 86 10.4 Churches ................................................................................................................................................................ 86 11 Interpretation of the Bled Island in the Middle Ages ................................................................................................. 89 11.1 Written sources ..................................................................................................................................................... 89 11.2 Early Medieval cemetery ..................................................................................................................................... 91 11.3 High Medieval guardians ..................................................................................................................................... 94 11.4 High Medieval church cemetery ........................................................................................................................ 95 11.5 Late Medieval and Post-Medieval church graves ............................................................................................. 98 11.6 Conclusion ............................................................................................................................................................. 99 12. Sources and bibliography ............................................................................................................................................ 101 1 INTRODUCTION 1. 1 INTRODUCTION following both pre-Christian and Christian rites. This TO THE SECOND EDITION grave is one of the rare archaeological finds that testify almost exactly to the point in time (in reality weeks The aim of this book is to demonstrate the im- or perhaps months) when the population of the small portance of grave orientation for the archaeology of region of Bled was (at least ostensibly) Christianised. Medieval death and burial. In doing this, I want to Archaeology was able to capture one of the very first emphasize two key points. First, grave orientation was reactions of a community to the new religion in one an important, and in some cases the most important, of the most intimate moments for any community, the conduit for the symbolic meaning of burials in the Mid- interment of a beloved member. And this insight was dle Ages. Second, for an archaeologist to reconstruct only possible because the grave orientation was analysed this meaning, the analysis must be conducted with in detail and studied in the context of the cemetery, the methodological rigour, and inferences must take the church, and the zeitgeist. broader context into account. I hope that this exciting example will be enough to To this end, the book addresses three interrelated encourage you to read the book, which is divided into issues. First, the state of scholarship on grave orientation three parts: Methodology, Analysis and Interpretation. studies; both in this case study and in the literature, the The methodology ( Chaps. 2 and 3) will be of interest topic is intertwined with church orientation. Second, the to archaeologists who are themselves struggling with methodology for analysing grave orientation; since the legacy data. The analysis of stratigraphy ( Chap. 4) is only case study is based on old excavations, special attention relevant to this particular case study, although some is given to the reuse of the legacy data. Third, the Bled innovative solutions are presented on how to extract Island case study. stratigraphic information from non-stratigraphic exca- Bled Island is an islet on Lake Bled (Slovenia). It has vations. The building analysis ( Chap. 5) of the church a special meaning for today’s Slovenian population and is an integral part of the context of the cemetery. The it held a special meaning for the local populations since analysis continues with the core chapters of this book prehistory. It is thus not surprising that the two Medieval dealing with the orientation of Medieval churches and cemeteries and the church on the islet are imbued with graves ( Chaps. 6 and 7) and the closely related spatial symbolism. And much of this symbolism was expressed organisation of the Early Medieval cemetery ( Chap. 8). through the orientation of the individual graves, grave This is followed by the spatial analysis of grave attributes groups, and the church building. ( Chap. 9) and the closely related absolute chronology The embedded symbolism is best il ustrated by the ( Chap. 10). All the information gathered is then dis- example. In the first decade of the 11th century, a single cussed in the archaeological interpretation (Chap. 11). interment occurred on Bled Island. The deceased woman The catalogue of the graves is available in the first was placed in the grave with jewellery according to the edition (Bitenc, Knific 2020b), but the plates with the centuries-old burial rites of her pre-Christian ancestors. drawings of the grave goods are reproduced for con- However, her burial differed from all previous ones in venience. one single feature: The grave was oriented parallel to the Christian church-to-be. More precisely, the church building must have been at least partial y built, other- 1.2 WHY THE SECOND EDITION wise the grave could not have been aligned with it. But the church could not yet been dedicated, otherwise the This book is the second edition of the core chapters “pagan” burial in the consecrated churchyard would of the book “Srednejveški Blejski otok v arheoloških not have been allowed. Thus, this was a biritual burial, virih / Medieval archaeology of Bled Island” ( Chaps. 3 7 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND to 8). More or less. Thus, a few words need to be written since 2020 that I have been able to use to noticeably about why I felt that this text needed to be republished improve some of the conclusions. after only two years. I hope that in its present form the book will raise The first edition was an edited volume resulting the awareness of how important grave orientation was from several research projects. The focus of the projects in some Medieval contexts. involved was to investigate the Early Medieval past of the Bled Island and, in particular, the possibility that there was an Early Medieval sanctuary there. Like many 1.3 FOREWORD such edited volumes, the book tried to be too much for TO THE FIRST EDITION too many people. It strove for a concise analysis of the Medieval cemetery, the Early Medieval artefacts, and the The book you are reading was written within the Early Medieval mythical landscape. These three goals framework of three research projects. The idea was are, on the one hand, too diverse for the interest of most conceived and the analysis of archaeological data was readers but, on the other hand, very specific to the place, performed within the scope of the seminal research the Bled Island. Moreover, the book was presented as a project Sanctuaries. Blejski otok (Bled Island), an island book about the Bled Island and was thus of immediate on the Lake Bled, seemed one of the most promising sites importance only for those who were already familiar for finding evidence of an Early Medieval sanctuary. It with Bled, archaeological y or otherwise. turned out, however, that a comprehensive and in-depth The number of potential readers for that book was analysis of the site as a whole would be needed. The work thus very smal . As we undeniably live in an age where was therefore finished within the framework of research more scientific texts are written on a given subject than projects, entitled The settlement of the Southeastern a single researcher can read, there has always been a Alpine region in the Early Middle Ages and Inventory, danger that individual subjects in the book will be over- analysis and evaluation of the primary and secondary looked by the scientific community. One such subject, sources of Slovene researchers on ‘the old faith in the region deeply buried in the chapter titled Morphometric analysis of Soča river’. Cemetery analysis took place within the of the site, was grave orientation. programme Archaeological research. This was a part of my contribution to that book. I Such a book would normally be expected to mainly focused on the spatial analysis of the cemetery, contain a consolidated archaeological and anthropo- which is surprisingly often overlooked in modern logical analysis, an analysis of written sources, and a research of Medieval cemeteries. The most important confrontation of the findings in the conclusion. But, result that came from that analysis, in my opinion, was when it comes to Bled Island in the Middle Ages, this the grave orientation. Grave orientation is a venerable is not possible. The most comprehensive archaeological topic in Early Medieval archaeology, but despite some sources are those from the 10th century, and the events very convincing results it is currently not considered im- of the 11th and 12th centuries cannot be identified with portant. Most modern studies may mention it briefly, but much accuracy. Due to the excavation method, there rarely build their conclusions on it. However, the Bled is no contextualised archaeological data from the 13th Island case study demonstrates that, with appropriate century onwards. When it comes to written sources, methodological rigour, grave orientation can provide the situation is reversed. The earliest direct written crucial insights for an understanding of a Medieval source for Bled Island dates to 1185 and is followed cemetery. by two brief mentions in the 13th century. Only from For this reason I have decided to republish my text the 14th century onwards do written sources become in the form of a monograph book. This edition offers informative enough to enable the creation of a continu- only modest additional scholarly insights compared ous interpretation. to the first edition, but the text has been stylistical y There is a second, perhaps even more significant completely rewritten. In the process, I found that the difference between the written and archaeological original structure of the text, which followed the scheme sources for the Early and High Middle Ages. The vast of method-result-discussion-conclusion, was not suit- majority of the former are biased writings produced by able. Too many strands of inquiry were opened up at the the extremely narrow and isolated social class of monks same time for a reader to be comfortable with. For this and priests. The latter are exactly the opposite: they reason, the structure of the book was changed with two were unwittingly – or at least without the knowledge objectives in mind. First, to reinforce the importance of that one day they would be readable – produced by all grave orientation. Second, to minimise the number of people. The consequences are well known: medieval threads opened at the same time. historiography – at least when it comes to the study of I have also taken this opportunity to significantly the discussed area – focuses on the study of a sequence improve the references. In addition to adding new bib- of legal acts ( cf. Chapter 8.1.1). Archaeology focuses liographic units, there have also been some publications on the study of material culture ( cf. Chapter 2), on the 8 1 INTRODUCTION stories of individuals and individual communities within focused on the archaeology of individuals and commu- the context of long-term processes ( cf. Chapter 8.1.2). nities. The very poor state of preservation of the bone Here, we have stumbled upon an unexpected di- archive meant that anthropological analysis could not chotomy between the archaeology of material culture per be included on equal footing with archaeological data in se and the archaeology of individuals and communities. the process of interpreting the site (Leben Seljak). This This volume contains an outstanding example of the is followed by a broader perspective that places the new first approach, written by T. Knific and P. Bitenc, and an findings in the context of the archaeological landscape example of the second approach, written by me. During of the Bled micro-region (Pleterski). Important for the the process of the creation of this book, I have come to book are the results of the excavations near the vil age a realisation that might be of a broader significance for of Bodešče (Modrijan). archaeology: there is no better and worse method; even Two key scientific questions are posed in the book: less so a right and wrong method, as it is taught by some Was there a pre-Christian sanctuary on Bled Island? university programmes. What we are seeing are two What was the chronology of the churches? distinct scientific fields of archaeology that ask differ- It can be established that the supra-local object of ent questions of the same archaeological record about pre-Christian worship on the Bled Island was a spring. the same people from the past. More questions result in Judging by analogies from written sources, the spring more answers and our knowledge of the studied past is was surrounded by a grove or at least lay in the shade therefore enriched. of a tree. Ritual y connected to the spring was a nearby The above realisation had a key effect on the struc- place marked with a rock. There is no direct archaeo- ture of this book. Individual analyses were conducted logical data on when this situation arose, but it can be in parallel with, and independently from, each other. assumed with a fair degree of certainty that this part The original separation of tasks into a cemetery analysis of the Bled Island had a special meaning already in (Knific, Bitenc) and a stratigraphic and building analysis prehistoric times. (Štular) proved infeasible. The intertwined archaeologi- At the beginning of the 10th century, a small com- cal record forced everybody to investigate everything. munity began to bury their dead on the island. The origi- Our intent was to merge the two final products into a nal organization of the cemetery symbolical y connected homogeneous whole. Soon, however, it was found that it with the spring that was worshipped on the island and duplications were the exception rather than the rule, with the central place of the Early Medieval mortuary and the advantages of two different approaches were landscape, the Višelnica bonfire site. The graves of the immense. Both the cemetery and the structures analysis first generation were careful y placed near a fireplace, are therefore presented intact. Natural y, both exploit where the fire burned for at least part of the burials. the same catalogue of graves, plates with artefacts, and The following three generations buried their dead ac- reproductions of original documents. cording to changed rules: The graves were arranged in The very poor state of preservation of the bone rows, some were oriented towards the cardinal east and archive meant that an anthropological analysis could testified to considerable astronomical knowledge. The not be included in the process of interpretation of last grave in this cemetery was dug in the first decade the site on an equal footing with archaeological data of the 11th century and its orientation corresponded to (Leben Seljak). that of the first church. This is followed by taking a broader perspective, In the first decade of the 11th century, probably where the new findings are placed in the context of after the formal acquisition of the property by the Bish- the archaeological landscape of the Bled micro-region opric of Brixen in 1004, a smal wooden church was built (Pleterski). on the island, dedicated to St Mary of the Assumption. A Important for the book are the findings of the ex- grave under the threshold, the guardian of the entrance, cavations near the vil age of Bodešče (Modrijan). provided symbolic protection for the church. Despite the apparent modesty of the church, its construction incor- porated some cutting-edge architectural and astronomi- 1.4 DESCRIPTION cal knowledge for the time. After only a few decades, the OF THE FIRST EDITION original wooden church was demolished and replaced by a slightly larger and more beautiful building. This church The book Srednejveški Blejski otok v arheoloških was also protected by the guardian of the entrance. To virih / Medieval Archaeology of Bled Island presents the west of the church, in the shadow of the sunrise on the results of a state-of-the-art archaeological analysis the day of the Assumption, a very small congregation of data from archaeological excavations carried out in of mixed gender and age began to bury their dead. The 1962 to 1964. community was orthodox and conservative, as reflected Polona Bitenc and Timotej Knific analysed the in the strict observance of the same rules for Christian archaeology of material culture, while Benjamin Štular burials for almost two centuries. These characteristics, 9 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND together with the exceptional position of the cemetery or at least bishops), a nonadult was buried. The continu- on the island, place this community in the circle of peo- ity with the old church is reflected in the burial of the ple who were part of the administrative and defensive third guardian of the entrance and the other burials to apparatus of the Bishops of Brixen. the west of the church. After almost a hundred years, perhaps in the sec- The archaeological data agree well with the his- ond decade of the 12th century, the wooden church was torians’ interpretations, which refer to the deed of demolished and the construction of a new stone church donation of 1004 and the consecration of a church in began. This was a relatively ambitious Romanesque 1142. However, it should be noted that a high degree of architecture that completely changed the appearance of agreement between archaeological and historical inter- the island of Bled: an island with a small church near a pretations does not make any of them more solid. Since spring in a grove became a church on an island. Imme- these interpretations are based on completely different diately east of the altar of the new church, in the holiest sources, they remain separate, and each is only as solid place in the Christian world (usual y reserved for saints as the arguments that support it. 10 2 DATA: DOCUMENTATION* FROM THE 1962 TO 1964 EXCAVATION 2.1 INTRODUCTION 2.2 PLAN DRAWINGS OF HORIZONTAL SURFACES In the following, I will briefly present the method of archaeological excavations that occurred on Bled The main focus of the documentation was on Island in 1962 and 1964. A more detailed history of the textual descriptions of the surfaces of arbitrary levels excavation was written by Bitenc and Knific (2020a). and plan drawings of the same. The textual descriptions For modern analysis, the most important informa- were recorded in the archaeological field diary in which tion is that the so-called planum or arbitrary excavation the head of the excavations described and interpreted technique was used (Šribar 1974, 7−8). Using this meth- the archaeological record (which he did not excavate od, the archaeological record is excavated in arbitrary himself). According to contemporary practice, the di- levels or spits rather than feature by feature in the reverse ary was supplemented by occasional photographs and stratigraphic sequence. The choice of this method was sketches of the situation. in line with the archaeological practices of the time, that In contrast to the above mentioned unprecise exca- is, before the introduction of single context recording vation parameters, the plan drawings were recorded with (Harris, Ottaway 1976; Barker 1977) and stratigraphic a high degree of accuracy. They were drawn on a scale excavation (Harris 1979). The excavations on the Bled of 1:20 or 1:50, which was a methodological advance- Island were in fact among the first in Slovenia (and ment for Slovenian (and Yugoslavian) archaeology at the the then Yugoslavia) to introduce the planum method, time (see the excavations of the Župna cerkev cemetery which was thus a methodological novelty at the time in Kranj in 1953 and 1964−1966 for the standards at ( cf. Šribar 1974, 8). the time: Štular, Belak 2012; Štular, Belak 2013). Scale The quality of the planum excavation depends on and orientation are not noted on the drawings. For my the following parameters: The depth of arbitrary levels, analysis, only the copies of the original drawings were the number of control sections (i.e. profiles), and ac- made available. The copies contained the following data: curacy in the xyz recording of artefacts. Context numbers, descriptions of the stratigraphic con- Each of the listed parameters has a direct influence texts in the margins, absolute elevation above sea level, on the quality, but also − in inverse proportion - on the drawings of at least two corners of the church building, speed of excavation. Digging arbitrary levels 0.1−0.2 m and (only in excavation area IP 1/1) drawings of the thick, making control sections every 4 or 5 m, and col- measurement grid. lecting artefacts within 4 or 5 m quadrants later became The quality of these drawings and their purpose can an established practice in Slovenian archaeology ( cf. best be il ustrated by the example of Planum 1 ( Fig. 1). It Šribar 1969, passim; Evis, Hanson, Cheetham 2016, 178). documents the situation before the top soil was broken Unfortunately, these parameters were far less precise and it lists grass vegetation species and shows contours in the excavations on Bled Island. An arbitrary level at 0.05 m intervals. Mapping grass vegetation species thickness was between 0.4 m and 2 m (0.2 m inside the is surely not important for archaeology and the spac- church), there were no control sections and the xy co- ing between contours is misleading, as such accuracy ordinates of the artefacts were not recorded. This means would require at least 20 elevation measurements per that the way the method was applied was all but precise. square metre. No such measurements are marked in this plan drawing, while for other plan drawings about 0.01 elevations per square metre were measured. The actual measurements were thus about 2000 times less accurate * All artefacts and documentation are kept in the Natio- than the precision of the plan drawing leads us to believe. nal Museum of Slovenia (NMS). That the precision of the contours is not based on the 11 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 1: Bled Island, trench 1/1 before 1962 excavation; the plan drawing is oriented according to excavation grid (data source: NMS archives AO Rn 222/1). accuracy of the measurements is also apparent from the present church building, the stratigraphy recorded the shape of the contours, which should be much more in the sections cannot be matched with the stratigraphy “jagged” ( cf. Fig. 13). Methodological y, the dichotomy recorded on the plan drawings due to discrepancies be-between the presumed accuracy of the recording, on tween the textual descriptions and the drawings. Inside the one hand, and the excavation of 2 m thick arbitrary the present church building, sections were placed along levels, on the other, is simply inexplicable. It seems that the edges of the trenches that abutted the wal s; thus, the the accurate recording of the archaeological record was sections are adjacent and paral el to the wal s. In the upper not the only, perhaps not even the most important goal. part, these sections only record the standing wall and its The authors were obviously trying to leave an impression foundation, in the lower part the stratigraphy parallel to on the onlookers. the wal s. These sections are thus meaningless as far as The method chosen for the documentation had the stratigraphic relationships between the archaeological an important effect on the quality of the information. strata and the standing wal s are concerned; this would Unfortunately, the elevation measurements recorded require sections perpendicular to the wal s ( Fig. 3). on the plan drawings are almost without exception the In summary, the existing section drawings are of elevations of arbitrary levels. They merely recorded how limited value for stratigraphic analysis. The methodo- accurate the horizontal excavation of each level was. The logical mistakes made by the excavators are due to the only informative measurements are the elevations of the inexperience that accompanied the introduction of a skeletons in the graves. new methodology. In summary, the existing plan drawings are of limited value for stratigraphic analysis. 2.4 ARTEFACTS (SMALL FINDS) 2.3 SECTION (PROFILE) DRAWINGS Artefacts found outside the graves were not sys- tematical y recorded. During the excavation of the test The analysed documentation contains 35 section trenches in 1962, artefacts within the trenches were (profile) drawings which are supplemented with descrip- recorded only approximately according to arbitrary tions of the recorded stratigraphic contexts. However, levels, for example, planum 4/5 (Šribar 1966, 1−19). In these are not strategical y placed control sections but 1963 and 1964, no xyz information was recorded for the profiles of the edges of the trenches ( Fig. 2). North of artefacts found outside the graves. For the artefacts from 12 2 DATA: DOCUMENTATION FROM THE 1962 TO 1964 EXCAVATION Fig. 2: Bled Island, location of sample sections (red) on the plan of all trenches of the 1962−1964 excavation; the plan drawing is oriented according to excavation grid (data source: NMS archives AO Rn 235). Fig. 3: Bled Island, section drawings of the narthex trench; top – section drawings, bottom − location of sections. Not to scale (data source: NMS archives AO Rn 222/29). 13 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 4: Bled Island, idealised sections; above − the cemetery north of the present church, below − the interior of the present church. The scale between height and length is distorted. the excavation area north of the present church, the xy spatial information either. Due to the very small quan- information is given at the grid quadrant level, but the tity of artefacts, which contrasts with the descriptions z information is not available (Šribar 1966, 56−58, 88). in the field diaries, it can be assumed that the artefacts The location of artefacts from the interior of the church were subjected to a strict selection process during the is recorded descriptively, usual y without giving any z excavation. This corresponds to the common practice information (for example, in the corner between the two in archaeological excavations in Slovenia until the apses and the southern side altar; Šribar 1966, 119−123, mid-1980s, where artefacts dated as later than the Early 139−140, 155−156). The situation is similar in the area Middle Ages were discarded, except for a few “beautiful” south of the church (Šribar 1966, 178−179; Šribar 1966b). pieces (e.g., Štular 2009, 18−19). The inventory cards for the artefacts now kept in In summary, the artefacts found outside the graves the National Museum of Slovenia do not contain any are of limited value for modern archaeological analysis. 14 2 DATA: DOCUMENTATION FROM THE 1962 TO 1964 EXCAVATION 2.5 THE VALUE OF DOCUMENTATION and 3 is therefore 594; 79 or 13% of which were recorded FOR STRATIGRAPHIC ANALYSIS during excavation. Thus, for the archaeological site on the Bled Island, it must be assumed that about 87% of the For the stratigraphic analysis of any arbitrary stratigraphic relationships and 50% of the stratigraphic excavation, it is crucial to estimate the proportion of contexts were not recorded. stratigraphic relationships recorded. This result is comparable to the results of a study In the method of arbitrary excavation in planums, that measured the efficiency of modern forensic arbi- neither (1) the stratigraphic relationships between strati- trary excavation. This study used 0.1 m thick levels and graphic units recorded within the same planum nor (2) recorded 51% of the actual stratigraphic relationships the stratigraphic units excavated simultaneously within (Evis, Hanson, Cheetham 2016). The difference between a single planum (spit) are recorded. 13 and 51% of the relationships recorded is due to thin- For sites with uniform, mostly horizontal contexts, ner levels. A more important difference is that in the the first deficiency can be partial y remedied by the con- above study, experienced archaeologists systematical y trol sections. As described above, this was unfortunately searched for stratigraphic relationships and recorded the not the case for the Bled Island site. Additional y, the excavations at the same time the layers were removed. stratigraphy of the site is extremely uneven and undulat- None of the excavators on Bled Island except the site ing in places. In such cases, even systematical y recorded manager was an archaeologist and the stratigraphic control sections could only mitigate the shortcomings relationships were not recorded while the layers were of the planum technique to a limited extent. being removed, but only after the plan drawing was At sites with predominantly horizontal strata, the completed. Thus, it seems that my estimate of 13% of the second deficiency is less pronounced when the thick- stratigraphic relationships actual y recorded is realistic ness of the arbitrary levels is close to the thickness of the compared to this study. stratigraphic layers. At the Bled Island site, the thickness Although 13% may seem small share, the strati- of the levels was at least twice the average thickness of graphic relationships that were detected are “real” and the strata. Thus, the pavements inside the church were can be used for archaeological analysis as such. up to 0.05 m thick and the levels 0.1 m or more thick. In the area with the highest grave frequency, the horizontal levels were up to 0.4 m thick and the depth of the grave 2.6 CONCLUSIONS pits was between 0.1 and 0.2 m. We have quantified this loss of data for arbitrary Overal , it can be said that the quantity and quality Levels 2 and 3 within the church nave of the Bled Island of the data recorded during the archaeological excava- site. Based on the super-position of Levels 2 and 3, 79 tions on Bled Island between 1962 and 1964 do not meet stratigraphic relationships could be reconstructed from today’s standards. But this is understandable since the the available records. However, there were 137 units in excavations occurred more than half a century ago. The the plan drawing of Levels 2 and 81 in Level 3 for which biggest shortcoming was certainly the great thickness of the stratigraphic relationships could not be reconstructed. the arbitrary levels: A thickness of, say, 0.2 m outside and Therefore, only 79 of the 297 stratigraphic relationships 0.1 m inside the church, which would be in line with the were recorded. Furthermore, the reconstructed section standards of the time, would allow a much better inter- shows that due to the thickness of the soil removed within pretation of the archaeological record. Otherwise, I can a single level, more than half of the archaeological record only feel sorry for the archaeologists of the time: Without was removed without being recorded ( Fig. 4). So, it can be the basic tool of modern archaeology, the stratigraphic assumed that another 297 stratigraphic relationships have excavation method, they were at a loss in the face of been lost. The total estimated number of stratigraphic re- such complex site with complicated geomorphological lationships in the archaeological record between Levels 2 and stratigraphic conditions. 15 3 METHODOLOGY 3.1 INTRODUCTION in archival excavation additional steps are required so that the archival data (for example plan drawings Interpreting archaeological excavations is most oriented to the excavation grid) can be georeferenced efficient when it is done immediately after the excava- and transformed into spatial information (GIS database tions. The more time that elapses between excavation positioned in a modern coordinate system). and interpretation, the greater the risk of loss of infor- The fourth and last step is no different from the ar- mation. Two other causes of information loss are the chaeological interpretation of non-archival excavations. absence of excavators in the interpretation process and The following three subchapters describe the tools when the interpreter has no practical experience with and methods used in the analysis of the archival excava- the method used in the excavation being interpreted. tions on Bled Island. When all three conditions are met the excavation archive transforms from information to data1 and one is then dealing with an archival excavation. In archaeological 3.2 GEOREFERENCING practice, the change from “normal” (i.e., non-archival) to archival excavation analysis is often reflected in the The starting point for the spatial analysis was the use of the term excavation archive instead of excavation consolidation of the spatial data in a single geodatabase. documentation. This was implemented in the GIS environment2, in the A modern interpretation of archival excavations coordinate system D96/ TM. First, the excavation data can be roughly divided into four steps ( cf. Štular 2008; were georeferenced. During the excavation 68 plan Pleterski 2008, especial y 27−40; Štular 2009, especial y drawings were created in different relative coordinate 39−46): Digitisation (converting field diaries, draw- systems, which were designed ad hoc for each excavation ings, photographs, etc. into digital data), datafication area individual y. During the post-excavation analysis (transferring digital data into information, for example in the 1970s, an amalgamated plan drawing was created plan drawing into GIS geodatabase), data analysis, and by merging the plan drawings into yet another quadrant archaeological interpretation. grid, which covers the entire excavation area ( Fig. 2). The first two steps are typical activities for creating Photographs of coloured in copies of plan draw- a digital archive. Compared to non-archival excavations, ings were acquired for my analysis (I was not allowed these are additional steps in which the basic methods to scan the originals because of their historic value). of historiography are applied (Grafenauer 1960, for Sixteen photographs of plan drawings of graves were example, gives a concise description of a method that acquired in full resolution; the remaining photographs is still relevant for my purpose). had only been available to me in a lossy compression In the third step, the tools to analyse information format and in reduced resolution. The numbers and from archival excavations differ from standard meth- letters in these photos were poorly legible or not legible ods. The standard tools need to be complemented by at al . The exact method by which the drawings were additional procedures aimed at “translating” archival photographed was not known to me and I did not have data into information suitable for modern tools. This access to the metadata. The only information available process can be il ustrated by the example of spatial data: was that high-quality equipment had been used and the In modern excavations, spatial data are already georefer- distortion of the camera lens had been corrected in the enced in a modern absolute coordinate system. However, Adobe Photoshop software. Under these circumstances, it would not have been advantageous for me to further 1 Data are the facts or details from which information is derived. Individual data are rarely useful in themselves. For process the data for distortion or noise removal. data to become information, it must be placed in a context. 2 ArcGIS 10.5, ArcMap module. 17 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 5: Bled Island, plan drawings of the graves north of the present church, georeferenced with the grid reconstruction method. The residual error of georeferencing can be seen when comparing the modern plan of the present church (red) with two plans from the 1962−1964 excavation (black, grey); the plan drawing is oriented to the north. To georeference plan drawings in the absolute in the original drawings increased with the distance coordinate system, the method previously developed from each starting point of the grid set up in the field. for the archaeological excavations of the Župna cerkev Furthermore, this error does not occur in the grid drawn cemetery in Kranj was first tested. That method is in the plan drawings of Excavation area 1/1. This was based on a reconstructed grid of quadrants on which because the aforementioned grid was laid out with stakes individual archaeological drawings are georeferenced only in the excavation area north of the church, but the (Pleterski, Štular, Belak 2016, 25−26). However, it did stakes were not retained and were replaced by strings not provide satisfactory results with the Bled Island ( Fig. 6). The laying out of the grid with stakes was later data and had resulted in errors of the order of meters abandoned altogether. Thus, a homemade measuring ( Fig. 5). Analysis of the results revealed that the error device ( Fig. 7) was used to take measurements for plan 18 3 METHODOLOGY Fig. 6: Bled Island, a view of the trench north of the present church; after the re- moval of the grid stakes the grid was marked with strings (source: NMS archives OA film No. 5839). Fig. 7: Bled Island, view of the trench north of the present church; during the plan drawing, measurements were taken with a self-made measuring device ( source: NMS archives OA film No. 8531). 19 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND drawings. An improvised protractor was used in the narrow church space where the architectural elements served as the basis for the measurements (Bitenc, Knific 2020a, Fig. 1.22). From what has been said and taking into account the methods of recording plan drawings at that time (Šribar 1969; Šribar 1974; Šribar 1976; Šribar 1977; Šribar 1981), it can be deduced that there were system- atic errors in all measurement methods used during the excavation and that the error increased with the distance from the starting (zero) point of the measurement grid. Since the same kind of systematic errors also occurred in maps from the 19th century, the plan drawings were georeferenced in GIS using the method developed for georeferencing such maps (Štular 2010; cf. Conolly, Lake 2006, 86−89): Each plan drawing was georeferenced by its location relative to the church building. This was possible first because of the fortunate circumstance that the excavations were carried out within and in the Fig. 8: The raster pyramid method. immediate vicinity of the church to which no structural changes had been made since the excavations; and sec- ond, because the rules of good practice were observed in the field and each plan drawing contains at least one necessary. A visual inspection of the drawings showed church wall with at least two corners. that another rule of good practice was followed in the Since a sufficiently detailed geodetic plan of the field: Fixed points in space, in this case church wal s, church that could be integrated into the modern co- were measured only once and copied into later plan ordinate system (D96/TM) was not available, I used drawings. Thus, for each excavation area, the first of the the method of raster pyramids, in which small-scale plan drawings − the master − was rubbersheeted onto information is gradual y extrapolated to increasingly the church; the others were rubbersheeted to the master. detailed scales ( Fig. 8; cf. Internet source 1). From the The procedure described for georeferencing plan draw-smallest to the largest scale, the following data were ings in the absolute coordinate system was optimised in used: Processed and visualised airborne LiDAR data such a way that the internal proportions of each plan (method according to Štular, Lozić 2016; processed were distorted as little as possible, and at the same time by E. Lozić), digital orthophoto image DOF025, geo- an optimal overall accuracy was achieved. detic plan of the church from 2009, and archaeological All 68 plan drawings were georeferenced using drawings. the method described above. Working in the GIS envi- Airborne LiDAR data was used to create a rough 3D ronment offers advantages such as the combination of model of the exterior of the church. A digital orthophoto plans from different excavation areas, smooth transi- image is four times more accurate, but is only suitable tions between different plan drawings, and working in for the south wall due to the oblique perspective. These the absolute coordinate system. This makes labelling of data were used to georeference the geodetic plan of the excavation areas and quadrants redundant. church (Berk, Boldin 2017) in the modern coordinate Next, all wal s, paved walking surfaces, and pits system (D96/ TM). The latter was the background tem- were vectorised for spatial analysis. The vectorisation plate onto which the plan drawings were georeferenced was carried out in a two-dimensional (2D) GIS envi- using the rubbersheet method, in which plan drawings ronment at a scale of 1:20. In this way, the precision of were moved, rotated, and scaled (but not skewed or the original plan drawings was maintained. The main distort)3 as required to fit the map by selecting a series purpose of the vectorisation was to enable stratigraphic of source-destination point pairs. analysis in a 2D GIS environment. In this method, each drawing is georeferenced Elevation data was included as an attribute so that independently of the others. However, for archaeologi- each context (wal , pavement, pit) could be displayed cal analysis of spatial documentation, the accuracy of in a 3D space. However, the individual contexts were the correspondences between the individual drawings displayed horizontal y because in most cases only one is more important than the absolute georeferencing er- elevation per context was recorded. The result is thus not ror of the individual drawings. Thus, a further step was a true 3D model, for which about 1000 times more eleva- 3 ArcGIS 10.5 software package, the Georeferencing tool, tion measurements would be required. The simulation of transformation of the 1st order, i.e. affine. the elements in 3D space nevertheless gives an idea of the 20 3 METHODOLOGY grave pit burnt layer wall Fig. 9: Bled Island, axonometric projection of selected stratigraphic contexts: graves, pits, burnt layers and wal s. Height differences are shown in 10-fold scale. relationships between the individual contexts in space, areas. Second, the information from the recorded sec- including those excavated in different years ( Fig. 9). tions which testified that the occupation or occupation For the purpose of spatial analysis of the cemetery, surface, at the time of the cemetery north of the present each grave was treated as a point. For those graves church was about 1 m above the bedrock ( Fig. 4). where the position of the head could be reconstructed, 136 elevation measurements of the bedrock were the point was positioned at the joint between head and recorded in the plan drawings ( Fig. 11). Unfortunately, neck. Poorly preserved graves were mapped to a point only 63 of those are direct measurements of the elevation located in the geometric centre of the skeletal remains. of the bedrock. I was able to obtain data for a further If the state of preservation of the skeleton permit- 50 indirect measurements and 23 reconstructed ones. ted, the cartographic azimuth was measured to the The former include measurements of soil layers in the nearest degree. Cartographic azimuth is the deviation immediate vicinity of the bedrock, while the latter in- from north in the national coordinate system (D96 in clude measurements of skeletons lying directly on the the Mercator projection) and should not be confused bedrock. These can deviate from the actual situation with the astronomical azimuth (see Chap. 7). by up to 0.1 m, but this is negligible given the required scale of the final result. In the next step, I used the thus obtained 136 3.3 GEOMORPHOLOGY measurements to interpolate the digital surface model OF BLED ISLAND within the excavated area. The best interpolation method in such cases is kriging (Chaplot et al. 2006). Due to the The geomorphology of the Bled Island has changed small number of measurements, the value for each cell considerably in recent centuries. The main feature of to- was calculated based on all measurements. day’s islet is a Baroque church surrounded by a viewing Next, the information on the present-day surface platform, with a staircase leading down to the southern outside the area of the church platform was obtained with shore ( Fig. 10). For archaeological analysis, it is thus of the analysis of the airborne LiDAR data (for the method utmost importance to obtain all possible information see: Štular, Lozić 2016; analysis: E. Lozić; Fig. 12). about the geomorphology of the islet before these changes. By comparing the two sources of data, the estimate So, I developed a method to fuse the recorded that the occupation level was about 1 m above bedrock elevation data for the bedrock beneath the Baroque was found to be satisfactory. This means that after in- buildings with airborne LiDAR data of the surrounding tegrating the data from both sources and adding one area. I relied on two pieces of information. First, the re- metre to the reconstructed bedrock surface, the transi- construction of the bedrock surface within the excavated tion between the areas of each data source was smooth. 21 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 10: Bled Island, view from the shore to the western part of the island (photo A. Pleterski). The final step was to merge the data from both areas into a unified point cloud and then create a digital terrain model (DTM) with the base cell size of 1 m for the entire islet using the Kriging interpolation method ( Fig. 13). We believe that this reconstruction adequately reflects the geomorphology of the island before the Late Medieval and Post-Medieval building works. The high- est point of the island was where the bell tower stands today. The present staircase that leads to the church from the south side of the island was built on the site of an earlier corridor that provided the easiest access to the top. The only relatively flat area on the entire islet was a small patch in the northern part of the present church Fig. 12: Bled Island, point cloud of ground points of the unbuilt nave and immediately north of it. area (by E. Lozić, method according to Štular, Lozić 2016; source: airborna laser scanning of Slovenia, eVode; http://gis. arso.gov.si/evode; D96/ TM coordinate system). 22 3 METHODOLOGY direct measurement indirect measurement reconstructed measurement excavation trench existing wall not excavated or destroyed Fig. 11: Bled Island, location of above-sea-level measurements of the bedrock. Fig. 13: Bled Island, geomorphological reconstruction of the surface before medieval building activity. Axonometric 3D view of the digital surface model; x, y, and z axes are in the D96/ TM coordinate system. 23 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND 3.4 STRATIGRAPHIC ANALYSIS Stratigraphic analysis of the data is the only way to find the answer, even if the arbitrary method was used The most important archaeological question in the excavations. The method I used was simple: to concerning Bled Island was the relationship between identify as many stratigraphic relationships as possible. the cemetery and the church buildings. Obtaining a This was done by correlating those plan drawings that are satisfactory answer to this question is complicated, in superposition; these were then used to create a Har- on the one hand, by the relatively poor archaeological ris matrix. For example, if the paved occupation surface documentation described above and, on the other, by termed Pavement 23 in Planum 11 partial y overlaps with the complex building history. While answering such Pavement 10 in (subsequently drawn and thus lower) questions is routine in modern stratigraphic excavations, Planum 12, the former is above the latter. However, the the situation is different when analysing more then half stratigraphic relationship between, for example, adjacent a century old data ( cf. Štular 2008). Pavement 23 and the Layer 15 in Planum 11 ( Fig. 14) is Fig. 14: Bled Island, present interior of the church. Overlay of plan drawings 11 (grey) and 12 (black); the plan drawing is oriented to the north (data sources: NMS archives AO Rn 222/11 and NMS archives AO Rn 222/12). 24 3 METHODOLOGY Fig. 15: Bled Island, Harris matrix; entire matrix above, excerpt below. 25 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 16: Bled Island, Harris matrix of selected stratigraphic groups with stratigraphic phases. lost forever. Moreover, all the pavements and layers within ( cf. Herzog 2004, 10). The forced fragmentation of the the arbitrary 0.5-m thick levels, that were not recorded in original data into individual stratigraphic contexts (cut, the plan drawings, were lost. fil , skeleton), which would inevitably be followed by During the excavations, the individual contexts integration into a stratigraphic group due to the lack of on each new plan drawing were numbered, beginning data, would not yield any new information. Thus, each with 1. To distinguish the contexts between the draw- grave retained the label assigned to it by the excavators, ings, I additional y assigned each of them the number of and the same labels are used in the catalogue of the the respective drawing. For example, the pavements de- graves, for example, g001 for Grave 1. If the documen- scribed above were numbered 11−23 for No. 23 on plan tation (usual y plan drawings) indicated that a single drawing 11 and 12−10 for No. 10 on plan drawing 12. label was used for several graves, letters of the alphabet The wal s were not always numbered during exca- were appended to the label (for example, g082a for the vation. I have therefore labelled them with the number skeleton and g082b to g082e for the skul s stacked at the of the plan drawing and a consecutive number of the edge of the grave pit). wal , for example, “02_wall1” for wall No. 1 in plan The result was 688 stratigraphic elements (strati- drawing No. 2. graphic contexts or stratigraphic groups). These included The archaeological record of each grave consists 128 graves, 347 layers and 213 cuts, wal s, pavements, of several stratigraphic contexts, at least the cut and and occupation levels. For these stratigraphic elements, at least one backfil . Nevertheless, in the available data 905 unique stratigraphic relationships were recognised. no distinction is made between these contexts: cuts The stratigraphic relationships were analysed are rarely described, and different fil s in a single grave with Stratify software.4 The advantages of this software are never described. So, each grave was recorded as a include, apart from the automation of matrix creation, stratigraphic group and entered as such in the matrix 4 Stratify 1.5, www.stratify.org. 26 3 METHODOLOGY the constant checking of the correctness of the relation- titative analysis based on mathematical algorithms is not ships, the consideration of additional data (absolute possible. Moreover, the finds – except for those from the date, location in space, depth, etc.), and the possibility of graves – could not be assigned to individual stratigraphic integrating stratigraphic units into stratigraphic groups contexts. This means that the quantitative analysis of the (Herzog 2004). A major disadvantage is that the output stratigraphy would not bring any new information, but of the final Harris matrix is a bitmap image that is all only additional data. In other words, quantitative analysis but impossible to read or use in any kind of analysis, would only increase the information noise. automated or manual ( Fig. 15). I thus decided to undertake a qualitative analysis Methodologically, stratigraphic analysis can be of the stratigraphy. The focus was on the graves and described as quantitative analysis: all relationships are architectural elements (wal s, pavements, postholes) taken into account, and the position of each stratigraphic that are in direct stratigraphic contact with the central element is analysed in relation to all others. But the church building. Based on their location in space and particular circumstances of the site in question required the known stratigraphic relationships between them, the a different approach. The fact that only an estimated graves were divided into seven stratigraphic groups, that 13% of all stratigraphic relationships were recorded (see is, groupings of stratigraphic groups ( Fig. 16). Similarly, above) is crucial here. Such a loss of data cannot simply the architectural elements were divided into five phases be ignored. If 87% of the information is missing, a quan- of construction. 27 4 SITE STRATIGRAPHY 4.1 INTRODUCTION Group 2 is in the area of the so-called church shed, where approximately ten graves are arranged in a single In the original interpretation of the Bled Island row, most of them with several reburials. site in the 1970s, five groups of graves were defined on Group 3 consists of approximately ten graves lo- the basis of their position in the cemetery. They were cated in a natural depression northeast of the church. described as follows (Šribar 1972, 390−391; Fig. 17). Group 4 is approximately twenty graves in the Group 1 is located north of the church, where there church (in the nave, in the small vestry, and in the corner are approximately 60 graves in six irregular rows. between the south wall of the nave and the east wall of the side chapel), including Grave 72. Fig. 17: Bled Island, location of Šribar cemetery groups 1 to 5. 29 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 18: Bled Island, location of the cemetery groups G1 to G7. Group 5 consists of the remaining graves inside the Phase 1 are the graves earlier than the earliest church church nave and the presbytery. building ( Fig. 18: G1, G2). Phase 2 are the two earliest Based on the analysis of the grave goods, these church buildings and the contemporary graves ( Fig. groups were defined chronologically relative to the 18: G3). Phase 3 are the third church building and the church architecture as follows. The interment in Group 1 contemporary graves ( Fig. 18: G4, G5). Phase 4 are the possibly began before the first church was built and it fourth church building and the contemporary graves ( Fig. certainly occurred during its use. Groups 4 and 5 are 18: G6). Phase 5 are the fifth (and still standing) church also contemporaneous to the first church. During the building and the contemporary graves ( Fig. 18: G7). use of the second church, the interment in Groups 1 and Each stratigraphic phase is described in more 5 continued. Group 3 could not be defined chronologi- detail below. cal y (Šribar 1972, 390−391). Šribar’s interpretation was based on the spatial distribution of the graves, sometimes called horizontal 4.2 PHASE 1 stratigraphy. However, my interpretation was based on stratigraphic analysis ( Fig. 16). Stratigraphic proximity Phase 1 comprises the graves that are stratigraphi-was the main criterion for belonging to a particular cal y earlier than the northern face of the still standing group and position in relation to the church only sec- church building ( Fig. 19). ondary. Most of the graves are positioned northwest of the Based on this, I defined five stratigraphic phases present church building ( Fig. 18: G1). They are sepa- (from 1 to 5) and seven cemetery groups ( Fig. 18). The rated from the others by an empty area, are in mutual cemetery groups were defined as follows. stratigraphic contact and are covered by three relatively 30 4 SITE STRATIGRAPHY Fig. 19: Bled Island, plan of Phase 1. uniform layers of soil (stratigraphic contexts 03−07, numerous prehistoric finds were discovered during the 03−09, 03−10; 03−12; 03−16, 03−17, 03−18, 03−19). excavation of the burnt clay of the fireplace ( cf. Bitenc, The stratigraphical y earliest element of this phase Knific 2020a, 25−40). However, this interpretation is a fireplace in the centre of this burial group. The fire- was based on a common mistake: The fireplace’s burnt place was described by the excavators as a layer of burnt clay was understood as the backfill of a pit and thus clay surrounded on the western side by a semicircular stratigraphical y belonging to the fireplace. In reality, arrangement of stones; prominent among them was a the prehistoric finds were embedded in a layer (the solitary stone, which was there at the time the fires were backfill of Hollow A) on top of which the fireplace was burning, as evidenced by the soil of a lighter shade, built. The heat of the fire burnt the underlying soil, thus i.e., less burnt, beneath and around it. The excavators creating the impression of a burnt “backfil ”. In other assumed the fireplace to be a prehistoric feature, since words, from the perspective of the layer containing the 31 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND metric dating of the skeletal remains from the Grave 72. The method used was radioactive carbon dating, and the uncalibrated result is 1115 ± 30 BP. The result was calibrated with OxCal v4.2.3 software (Bronk Ramsey, Lee 2013) using the IntCal13 calibration curve (Reimer et al. 2013). The calibrated dates within 1 standard deviation of the mean (σ1) are between years 895 and 930 CE (34.6% probability) and between years 939 and 972 CE (33.6% probability). The calibrated dates within 2 standard deviations of the mean (σ2) are between the years 779 and 789 CE (1.2% probability) and between the years 869 and 1013 (94.2% probability) CE. In ar- chaeology, the most frequently cited is the range of σ1 values, which is 934 ± 38 CE. The graves from Phase 1 therefore belong to the Early Medieval cemetery on the Bled Island, which is earlier than the first church ( Fig. 18: G1 and G2). 4.3 PHASE 2 Phase 2 ( Fig. 22) is the period of the two earliest churches (see Chap. 5) and two contemporary graves. Stratigraphical y, the most important elements of this phase are the two church apses ( Fig. 21). On the one hand, the same location and the same construction technique of the two apses indicates the continuity, on the other hand, the two apses are stratigraphical y clearly separated from each other. As noted in the field diary: Fig. 20: Illustration of the processes on the fireplace from before the start of construction of the second apse, a 10 cm prehistory until excavation: a − activities on the prehistoric thick layer of clay had been laid over the foundations of occupation or ground surface; b − archaeologization, i.e. the first apse (...); the western end of the arch (...) of the the creation of an archaeological record of the prehistoric apse is partial y covered with a pavement (...) belonging activities; c − fireplace on the Early Medieval occupation sur- to the apse above (Šribar 1966, 197). face; d − archaeologization of the Early Medieval processes; The sequence of these two very similar buildings e − archaeological excavation of the burnt layer containing with apses is confirmed by the stratigraphy of Graves prehistoric pottery sherds. 67 and 71 (Šribar 1966, 116−117). The analysis of the ground plan and the field diary shows that the interment of Grave 71 occurred first; then the grave was opened, the body remains removed, and the emptied grave pit prehistoric finds the burnt “backfil ” is an interface or backfilled. The pit for Grave 67 was then dug about half interfacies created in a post-depositional process. Thus, a metre further south, creating a cut in the northern the pottery found in the burnt “backfil ” belongs to the part of Grave 71, which had already been emptied and prehistoric Hollow A and not to the Early Medieval backfilled by this time. Then a mortar pavement was fireplace ( Fig. 20). That the fireplace was used before laid above both graves. The position of the two graves or during the interment of some graves (Nos. 40, 41, is particularly revealing: the earlier grave No. 71 was 46) is attested by fragments of burnt clay found in the immediately in front of the earlier church building, backfill of these graves. the later grave No. 67 immediately in front of the later Stratigraphical y contemporaneously to the above church building, and both were in the most likely posi- are graves in the immediate vicinity of the northeast tion of the respective entrances ( Fig. 18: G3; different corner of the present church ( Fig. 18: G2) and Grave 72. Šribar 1972, 391). The latter lies in an isolated position and it predates the The stratigraphic discontinuity on the one hand first church as it is positioned under its apse ( Fig. 21; and the clear elements of architectural and sepulchral Šribar 1966, 118; cf. Šribar 1966, 199). continuity on the other justify the division of Phase 2 The absolute dates for Phase 1 were determined by into Subphases 2a and 2b. The terminus post quem for the jewellery in the graves (see Chap. 10) and by radio- Phase 2a is provided by the C14 dating of the Grave 72, 32 4 SITE STRATIGRAPHY [cross-section through both apses and the pit for Grave 72] [I) A quadrangle of stone bonded with a poor mixture of lime. Bed made of uneven mortar and 18 cm thick layer of clay underneath (2 types of stones)] [II) Top apse on wet-tamped clay (stones 0.25-0.35 to 0.1-0.3 cm. Height up to 0.20. Height in profile 20-30 cm. A. Black soil (base for II.)] [III) Second apse in black layer (burnt?) B. Below III Layer of humus.] [IV) Two rows of stones above the edge of the pit for Grave 72 (In brownish-red greasy clay).] Fig. 21: Bled Island, apses of the two latest church buildings. Top left: Plan drawing. Top right: section drawing of the two apses and Grave 72 in length. Bottom: Section drawing of the two apses and Grave 72 in width (data sources: NMS archive AO Rn 222/11, Rn 222/12, Rn 222/13, Rn 222/18). 33 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 22: Bled Island, plan of Phase 2a (F2a) and 2b (F2b). which is certainly after 779 CE, but most probably in the zontal y separating the two earliest apses (Šribar 1966, second half of the 10th century. The dating of Phase 2b 108, 121). The pottery in question has been lost and to roughly the end of the Early Middle Ages is supported therefore cannot be dated precisely. not only by the architecture, but also by two fragments Only for the sake of clarity of the text will this phase of Early Medieval pottery discovered in the layer hori- and its graves be referred to as pre-Romanesque. 34 4 SITE STRATIGRAPHY Fig. 23: Bled Island, plan of Phase 3. 4.4 PHASE 3 ( cf. Fig. 19), thus there are two possible interpretations. The first is that the graves are contemporaneous with the Phase 3 ( Fig. 23) is stratigraphical y more complex third church, because the legs of the deceased extend and probably lasted longer than Phases 1 and 2. Its begin- exactly to its west wal . The second possibility is that ning is marked by the construction of the third church the interment of this group started in Phase 2b, which building (see Chap. 5). The archaeological record of this is suggested by the spatial position of the graves (see phase consists of the remains of the horseshoe-shaped Chap. 9). Since at least the latest graves of this group all apse (stratigraphic contexts 79−8, 79−9, 79−10), mortar belong to Phase 3, the group is considered in the context pavements (stratigraphic contexts 12−04, 12−05, 12−06, of Phase 3. These graves will be referred to as the High 12−10, 12−49, 12−50) and a protrusion from the wall Medieval cemetery on the Bled Island. in the southwest corner of the present church nave Additional y, Phase 3 comprises another group of ( Fig. 32; cf. Fig. 24). graves ( Fig. 18: G5). These graves are stratigraphical y At least two groups of graves belong to Phase 3. connected to the horseshoe-shaped apse, but are some- The stratigraphical y earlier one is the group of graves what less homogeneous as a group. Some are located in in the present western church portico ( Fig. 18: G4). It is the southeast part of the church, others just outside to a stratigraphical y homogeneous group that lies under the south of the apse. two layers of soil (stratigraphic contexts 24−01, 24−03) The approximate dating of this phase is made and was overlain by the west wall of the church nave in possible by the connection with the contemporary the subsequent Phase 4. architecture, which was built in the Romanesque style The stratigraphic relationship of the graves to the (see Chap. 5). contemporary church building has not been recorded 35 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 24: Bled Island, plan of Phases 4 and 5: 1 − Phase 4 grave; 2 − Phase 5 grave; 3 − wall protrusion marking where the western wall of Phase 3 stood; 4 − existing wall; 5 − unexcavated or destroyed. Only for the sake of clarity of the text this phase is predate the construction of the Gothic architecture, as referred to as Romanesque and its graves as the High they would have been destroyed during levelling. Medieval cemetery. Only for the sake of clarity of the text this phase wil be referred to as Gothic and the graves as Late Medieval. 4.5 PHASE 4 4.6 PHASE 5 Phase 4 ( Fig. 24) comprises 13 graves to the south of the church building ( Fig. 18: G6). Most of them Phase 5 ( Fig. 24) is the period of Baroque archi- were tightly packed against the wal , in a narrow space tecture that characterises the present church building. between the church building and the bell tower. The Stratigraphical y, it includes two adjacent tombs in the shallow graves were partial y cut into the bedrock and northwest part of the nave, the position of which was only thinly covered with soil. So, there is no direct marked in the Baroque pavement of the church ( Fig. stratigraphic contact with other grave groups or archi- 18: G7). tectural remains. Only for the sake of clarity of the text this phase The area where the graves are located was levelled will be referred to as Baroque. to the bedrock during the construction of the Gothic church ( Fig. 13). Thus, the graves in this group cannot 36 4 SITE STRATIGRAPHY Fig. 25: Bled Island, location of prehistoric finds: 1 − existing wall; 2 − unexcavated or destroyed; 3 − stone tool; 4 − outside excavation trench; 5 − pit or posthole; 6 − number of finds per m2. 4.7 PREHISTORY the subject of this contribution. In the following, only the prehistoric features that were in the past mistaken Scarce prehistoric archaeological remains were also for Medieval will be described in detail. excavated on the Bled Island site. Mostly these were ar- tefacts discovered in secondary positions, for example in The first to be described is the pit positioned under the backfil s of the graves from Phase 1. In some places, the presbytery of the present church. To understand it, especial y in the presbytery of the present church, the stratigraphic Phases 1, 2, and 3 must be considered. documentation mentions a prehistoric “cultural layer”. The pit consists of a rectangular cut, three backfil s, However, based on the descriptions of the layers, it can a recut, and its backfill ( Fig. 26: P4−1, 2, 3, and 4). The be assumed that these were only artefacts in secondary four backfil s were described in the field diary as follows. position within the Medieval contexts. Prehistoric finds First, dark brown clay mixed with charcoal and clay daub. can thus not form a stratigraphic phase. Second, hard clay mixed with numerous fragments of Regardless, some inferences can be made. Analy- charcoal and clay daub (a hearth). Third, a patch of hard sis of the distribution of the prehistoric finds shows a clay. And fourth, black-brown clay and a possible site of a concentration in the area of the present presbytery of rectangular posthol e (note by B. Š.: the description refers the church and a second concentration to the north of to the backfilling of the recut as described in the margin the present church ( Fig. 25). The prehistoric artefacts of plan drawing Rn 221/27). certainly deserve a more detailed analysis, but are not 37 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND The pit was cut into the bedrock and covered by a Phase 3 levelling: a layer of reddish-brown clay (Šribar 1966, 107) and light brown humus mixed with crushed stones and fine sand (Šribar 1966, 100, Nos. 6 and 7). The pit and its backfil s have no direct stratigraphic con- tact with the Phase 1 graves, nor with the architectural remains of Phase 2. As far as can be surmised from the descriptions of the layers, the first two of the above de- scribed backfil s were formed in separate events. Each event followed the burning of a wattle and daub build- ing typical of the prehistoric period in close proximity. Perhaps it can be inferred that the surrounding area was cleared after a fire, but not in the sense of levelling the ruins for a new building, as the amount of charcoal and burnt clay is insufficient. The third backfill appears to have been created during the consolidation of the pit floor, while the recut and its backfill are a post hole, i.e., archaeological evidence for a rectangular post or stone. A thorough analysis of the plan drawings provides further details about the stratigraphic position of the pit. The starting point is the information − indirectly attested several times in the documentation (Šribar 1966, pas- sim) − that the church pavement was horizontal and that any differences in level were negligible. In a plan draw- ing of arbitrary level 3 (Planum 3), the aforementioned levelling of reddish-brown clay is shown as a continuous Fig. 26: Bled Island, pit next to the two earliest apses, plan layer that completely covers the pit. However, a little drawing of arbitrary levels 3 (P3), 4 (P4), 5 (P5) and 6 (P6). further west, the surface of the clay at the western end of On P3, the green stone (No. 30) and the layer above the pit the grave pit [72] is at the level of its upper edge (Šribar (No. 24) are marked. On P4 the pit backfil s (Nos. 1−3) and 1966, 106). This means that Planum 3 was excavated at the posthole backfill (No. 4) are marked (data sources: NMS archives AO Rn 222/12−15). a level roughly corresponding to the occupation level, Fig. 27: Bled Island, stratigraphy of the area of the pit and the three latest apses. The measurements above sea level are to scale, all other elements are only il ustrated: 1 − Phase 1, occupation surface on the pit‘s backfill; 2a, 2b − Phases 2a and 2b, wal s of the pre-Romanesque apses with the respective occupation surfaces; 3 − Phase 3, Romanesque apse and contemporary occupation surface. 38 4 SITE STRATIGRAPHY Fig. 28: Bled Island, view of the pit next to the two earliest apses documented on arbitrary level 3 (looking south; source: NMS archive AO film No. sv4712). i.e., the occupation surface that was walked on at the use. It is a rectangular stone cut from green tuff and is time of the interment of Grave 72. This observation is about 0.3 m long, 0.2 m wide, and 0.25 m high (source: also confirmed by the analysis of the section drawing: Archive NMS AO Rn 222/12). This relatively easy-to-cut The top of the pit was recorded as being about 0.2 m green tuff (local y known as “groh”) was a particularly lower than the top of Grave 72 ( Fig. 27). Thus, it can be popular material for architectural elements in the area assumed with a fair degree of certainty that Grave 72 of today’s western Gorenjska from the end of the Middle and pit with its backfil s were all covered with a layer of Ages until modern times (Avguštin 1971). According to clay that represented the occupation surface at the time. the description, the prehistoric polished axe mentioned The pit and its backfill are therefore stratigraphical y above seems to have been made of the same material earlier than Grave 72. (Bitenc, Knific 2020b: Pl. 13: 11). Green tuff was also The time at which the pit was dug can only be in- used local y as a material for gravestones in the Roman ferred indirectly. Immediately next to the pit, numerous period ( lupa 3726).5 prehistoric finds were recorded on the burnt surface: At To summarize the stratigraphic information. The least 28 fragments of pottery, more than 61 fragments of cut piece of green tuff lay (1) on top of the backfill of burnt clay, two stone tools, a fragment of a river pebble the pit, (2) embedded in a levelling accumulation of (NMS inv. No. S1922 a−c) and a fragment of a greenish- reddish-brown clay (which covered the pit and was the grey stone axe with a finely polished surface, measuring occupation surface when Grave 72 was dug), and (3) 5 x 6.2 cm (Šribar 1966, 121−122). These finds indicate a beneath the pavement of Phase 3. The stratigraphic (1) possible prehistoric dating of the pit, which is indirectly terminus post quem for the placement of the stone is thus supported by the presence of clay daub in the backfil . the backfilling of the pit prior to Phase 1; (2) the terminus A cut stone ( Fig. 28) lying on top of the backfilled ad quem is the creation of the occupation surface used at pit may shed some light on the duration of the pit’s 5 I thank E. Lozić for the information. 39 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND the time of the interment of Grave 72; (3) the terminus ing repeated activity; a (first) post or stone as a spatial ante quem is the construction of the building in Phase 3. marker; an intentional y placed cut stone of green tuff The terminus ante quem rules out the possibility that the as the (second) spatial marker; an exceptional Early green stone is an infiltrated High- or Late Medieval ar- Medieval grave (No. 72); the apses of the two earliest chitectural element, while the terminus ad quem proves church buildings. Taken together, this evidence suggests that the stone was protruding from the ground at the that continuity in the special meaning of this place is time of the interment of Grave 72 and also later, at the all but certain. time of the construction of the earliest church. Of crucial importance is the position of the stone above the final backfill of the pit and its embedding in 4.8 PRELIMINARY the levelling accumulation. This means that the people ABSOLUTE CHRONOLOGY who backfilled the pit deliberately placed the stone in its position. (Much) later, when the occupation surface A preliminary chronology of the phases is thus as was levelled, further deliberate care was taken for the follows. stone to remain visible by ensuring that it protruded about 0.2 m above the ground. The most convincing Phase 1: interpretation of this deliberate care is that the stone The Early Middle Ages, characterised by artefacts; was used as a spatial marker for the location of the pit. the terminus ad quem is C14 σ1 895−972 CE. Thus, the exact purpose and date of the pit and stone are unknown. The above does not provide direct Phase 2: evidence that the people who dug Grave 72 and the The Early Middle Ages, characterised by Pre- people who built the first church − both of whom used Romanesque architecture; the terminus post quem the occupation surface from which the green cut stone is C14 σ1 895−972 CE. protruded − were aware of the location and significance of the pit. However, considering the rarity of cut stones Phase 3: in the Early Middle Ages, the stone must have been a The High Middle Ages, characterised by Roman- remarkable feature in the landscape of Bled Island. As esque architecture. such, it is highly likely that it was indeed understood as a spatial marker (for the term see Fletcher 2004, 136, Phase 4: note 1; Novaković 2001, 221) for the pit and/or its mean- The Late Middle Ages, characterised by Gothic ing. This interpretation is supported by the posthole architecture. (see above, recut), which was most likely a remnant of an earlier spatial marker or perhaps even the original Phase 5: position of the cut stone. The Post-Medieval period, characterised by Ba- So, the following archaeological evidence emerged roque architecture. from this area of no more than a few square metres: a concentration of prehistoric finds, including a polished A more detailed chronology is presented in the axe made of green tuff; a pit with several backfil s indicat- following sections (see Chap. 10). 40 5 BUILDING ANALYSIS 5.1 INTRODUCTION wal . Although Šribar (1966, 198) lists at least four other postholes measuring 15 x 20 cm at a distance of 2 m Until the Late Middle Ages, the only building on from the shoulder of the apse, they do not appear in the the Bled Island was the Christian church, and in this plan drawings or in any other documentation. In this chapter I have reconstructed its building history. particular case, the plan drawings were considered to Unfortunately, archaeological analysis of the exist- be the more reliable data source than the field diaries. ing building (for example, analysis of materials, building Primarily on the basis of the remains of the apse techniques, connections between architectural elements) as documented in the plan drawings, the construc- was not carried out during the construction works in tion technique can be imagined as similar to that of the church. Thus, the scientific description, assessment, the church built in the middle of the 10th century in and dating of the standing building structures was not present-day Prague (Malostranské trg 2), Czech. It was possible. a round ecclesiastical building with a foundation of For this reason, the present analysis was limited to roughly worked clay-bound stones. The superstructure excavation archaeology. The available data were mainly was probably made of wood and plastered with yellow plan drawings and field diaries. As a result, this analysis clay. Analysis of clay plaster revealed that the wall of focused on the earliest phases of the church. the Prague church was smoothed and painted with red A note on terminology: although the interpretation and ochre pigments based on iron oxides (Čiháková of the function of the buildings is not discussed until 2018, 297−298).6 the end of this chapter, the term church has been used The sparse archaeological traces alone are not suf- throughout for the building for the sake of simplicity. ficient for a reconstruction of the ground plan. This is only possible if the continuity of the orientation of the apsidal buildings is taken into account. Such continuity 5.2 GROUND PLANS is testified by several archaeological analogies; when a new church was built over an earlier one in the Medi- Based on the analysis of the relative stratigraphy eval period, the orientation of the earlier building was presented in previous chapter, a plan of the architectural precisely observed (Strmčnik-Gulič 1994; Sagadin 2017; elements for each stratigraphic phase was drawn up. cf. Čaval 2010, 168). Stratigraphical y the earliest building was a church On Bled Island, I based the orientation of the with a semi-elliptical apse, which I termed the first church on the north wall of the present church. This church building ( Fig. 29). It was very poorly preserved, wall has Romanesque foundations and was surveyed in having been largely demolished by later buildings. The situ to have a declination of 78.4 degrees with respect foundation of the apse ( Fig. 21) was described in the to the north (see Chap. 6). Using this information, the field diary as being approximately 70 cm wide, built recorded remains of the apse, and the longitudinal pits of roughly worked stones up to 40 x 60 cm in size, and on the north, south and west sides, I was able to recon- bound with a dark yellow, low-quality mortar (Šribar struct the ground plan ( Fig. 29: b, c). 1966, 105 and 199). However, according to the plan This reconstructed ground plan is asymmetrical, drawings, clay was the binding material of the stones. which was not a common trait for Medieval churches. Of the rest of the building, only small longitudinal pits However, it is supported by the location of Grave 71: A in the bedrock ( cf. Šribar 1966, 113−114) and a small symmetrical building would have to be built over the patch of mortar pavement have survived ( Fig. 24: a; Fig. grave ( Fig. 29: f). Nonetheless, since the reconstructed 25: 13−09, 14−06, 14−08, 14−09, 14−B1). The recorded 6 The author would like to thank J. Rihter for this infor- pits can be interpreted as the foundations of a wooden mation. 41 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 29: Bled Island, Phase 2a, first building with semicircular apse, reconstruction of the ground plan: a − archaeological remains (purple for pits, light brown for mortar pavement, dark brown for wall foundations); b − reconstruction of the outer wal s and the central axis of the apse; c − reconstruction of the ground plan; d − idealised south wall taking into account symmetry; e − reconstruction of the actual (brown) and the idealised (yellow) ground plan; f − length-width proportions of the idealised ground plan. location of the south wall is based on a single pit, the When all the evidence is considered, the most plau- version with a symmetrical ground plan was also consid- sible reconstruction of the first church with apse is the ered. The proportion of the nave constructed in this way one with an asymmetrical nave, which was 7.5 m long would be 1:√2, which was typical for early Romanesque and 4.8 m wide and had an area of 35.9 m2. The entire churches with a semi-elliptical apse (Stopar 2017, 18−33; building was 9 m long. The apse had a semi-elliptical Fig. 29: d, e, f). shape and was 2.8 m wide and 3.5 m long, making for a ratio of 4:5. 42 5 BUILDING ANALYSIS Fig. 30: Bled Island, Phase 2b, second building with semicircular apse, reconstruction of the ground plan: a − archaeological remains (yellow for the mortar pavement, brown for the wall foundation); b − reconstruction of the outer wal s and the central axis of the apse; c − reconstruction of the ground plan; d − reconstruction of the actual (brown for the nave, yellow for the annex) and the idealised (black line) ground plan. Of the second church building ( Fig. 30), the the south side. The latter can be reconstructed if one as- remains of 0.67−0.72 m wide foundations made of sumes that the building was symmetrical. The position roughly worked stones bound with mortar (stratigraphic of Grave 67 directly adjacent to the thus reconstructed contexts: 12−18, 12−25, 13−20, 13−25, 13−27) and south wall confirms the symmetrical a reconstruction mortar pavement (stratigraphic contexts: 13−12, 13−13, ( Fig. 30: b, c). 13−14, 13−19, 13−20; Fig. 30: a) were preserved. It had The nave has the shape of an irregular rectangle a semi-elliptical apse just as the first church, but the with a width of 5.95−6.15 m, a length of 6.75 m, and an binding material is yellowish mortar instead of the clay area of 52.3 m2. In the narrower part, the ratio between (Šribar 1966, 198). On the basis of the scarce remains, length and width deviates only slightly, by 0.1 m, from it could not be determined with certainty whether the the √3:2 ratio typical of the Pre-Romanesque style. building was constructed of roughly worked stones At some later time the nave was extended by 3.1 m bound with mortar or whether it was a wooden build- to the west with a room in the ground plan of an irregular ing with stone foundations. The only indirect evidence trapezoid. The preserved mortar pavements testify that, was the sandstone fragments with remains of frescoes, at least in the last phase of the building’s use, the nave and which were built into the foundation of the next (third) the extension functioned as a single room ( Fig. 30: c). church and can be considered as ruins of this (second) Integrated into the apse described above was a drys- church (Šribar 1966, 138). The frescoes are an indica- tone wall made of large square pieces of limestone and tion that at least parts of the second church were built tuff, up to 60 x 70 cm in size (Šribar 1966, 108 and 197, from sandstone. also the drawing of Planum 3, No. 28). While the strati- On the basis of the preserved wal s, the ground graphic relationship to the apse was not ful y understood plan of the building can be detected on all sides except by the excavators (Šribar 1966, 197), an analysis of the 43 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 31: Bled Island, view of the drystone wall made of larger limestones and square tufa stone as documented on the arbitrary level 3 (view to the south; source: NMS archive AO film No. sv4649). photographs clearly showed that it was built against the the present church. More specifical y, the entire north pre-existing apse and later than the apse. There is no doubt and south wal s and parts of the east wal survive in that this wall and the second apse were demolished to the the present church. The west wall and the apse were same level, i.e. at the same time ( Fig. 31). This means that destroyed during subsequent alterations. The location of at least in the last phase of their use, they were contempo- the apse is evidenced by the preserved mortar pavements raneous and functioned as a whole. There are not enough and a masonry projection in the southwest corner of the elements to reconstruct the ground plan of the potential present nave. The remains of the apse were recorded room; it is not even completely clear whether it was a archaeological y ( Fig. 21; Šribar 1966, 136). building at al . Šribar’s conclusion that it was an extension The alterations after the third church building were to the second church (1966, 110) can thus only be slightly built in the Late Medieval Gothic and Post Medieval improved: There were two wal s that perhaps belonged to Baroque architectural style. These alterations are not pre- a small rectangular room attached to the second church. served in the archaeological record and are thus not the In the next stratigraphic Phase 3, the construction subject of this analysis. I will only mention those that were of a building with a horseshoe-shaped apse occurred. in the past interpreted as Early Medieval (Šribar 1967, 63). I termed it the third church building ( Fig. 32). Its area First, a wall in the shape of a small square was was 3.2 times larger than the second church building that discovered directly under the pavement of Phase 5. It it replaced. This was the most ambitious reconstruction is stratigraphical y later than Phase 2 and earlier than in the architectural history of the Bled Island church, in Phase 5. Due to its position, it can be interpreted as part which the previous building was razed to the ground. of the interior of the church in Phase 4 or Phase 5, i.e., Such a radical action was never repeated again, and most it is of Late Medieval or Post-Medieval date. of the wal s built in this phase still stand incorporated in 44 5 BUILDING ANALYSIS Fig. 32: Bled Island, Phase 3, building with horseshoe-shaped apse, ground plan reconstruction: reconstructed ground plan in purple, mortar pavement in yellow, wall foundations in brown, present church wal s in grey; location of grave No. 118 is marked. Second, the wal s north of the present church have First, apsidal buildings in the region between the been interpreted as part of a Romanesque church with ninth and 11th centuries were without exception Chris- three naves and two apses. However, these wal s are tian church buildings (Stopar 2017, 18−36). Second, at stratigraphical y later than the Gothic and earlier than least from the 8th century onward, Christian churches the Baroque church and can be interpreted as parts of were built in such a way that the altar was located in annexes with secular function. Such annexes were com- the east and the congregation faced east when praying mon in the Late Middle Ages. The evidence thus clearly (Gordon 1971, 213), which is the case with the apsidal refutes the controversial hypothesis of an Early Medieval buildings on Bled Island. Third, in the Southeastern church with three naves and three apses (Šribar 1967; Alpine region (Slovenia, Austria, and Italy), all build- Šribar 1971). ings that have been archaeological y recorded under the foundations of Late Medieval churches and have the same orientation are interpreted as churches (St George 5.3 INTERPRETATION in Legen: Strmčnik-Gulič 1994; the parish church in Kranj: Sagadin 2017; St Martin in Laško: Stopar 2017, Since archaeological record of activities inside 26; St Daniel in Gailtal: Eichert 2012, 22−23; Molzbichl the buildings have not been preserved or recorded, in Spittal: Eichert 2012, 51−55; St Peter at Edling, Spittal: the interpretation of the buildings’ function cannot be Eichert 2012, 62−64; St Martin, Rive d’Arcano: Gleir- taken for granted. Nonetheless, there are four convinc- scher 2006, Abb. 3: 2; St Martin, Zell at Kufstein, and St ing arguments that all buildings with apses served as Martin, Pfongau: Gleirscher 2006, Abb. 6: 2,3). Fourth, Christian churches. there are no documented cases of High Medieval graves 45 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND directly adjacent to a contemporaneous building that The dimensions, the shape of the apses, and the was not a church. regular proportions of the naves of the two churches The apses, the orientation towards the east, the with semi-elliptical apses are typical elements of Pre- continuous use of space, and the contemporary graves Romanesque ecclesiastical architecture (Stopar 2017, are thus reliable arguments to firmly interpret the Bled 18−19). The complete destruction of the two earliest Island buildings with apses as consecutive construction churches strengthens the interpretation that these phases of a Christian church. buildings were at least partial y constructed of wood Based on what a ground plan can reveal about the and that the demolition was due to dilapidation. Per- architecture, it seems that the first church was original y haps the duration of such buildings can be inferred designed according to the principles of Pre-Romanesque from the comparison with a roughly contemporaneous ecclesiastical architecture: As a symmetrical building small wooden church in nearby Lesce. This church was with a semi-elliptical apse and a 1:√2 ratio of the nave. demolished “after a long period of time” ( post multorum However, the implementation had to be adjusted to suit vero cursum temporum) and replaced by a stone church. the rough terrain. Instead, an asymmetrical, mostly In Lesce, this period was estimated to have been six to wooden church was built. seven decades (Bizjak 2012). This church was razed to the ground and replaced The second church on the Bled Island was followed by a building similar in appearance, size, and propor- by the construction of a stone church with a horseshoe- tions, but perhaps at least partial y built of stone. During shaped apse, which can be described as a typical Ro- its lifetime, an extension of the nave was built in the manesque church due to its size and shape. Unlike the form of an irregular trapezoid on the west side. Due to two previous churches, whose construction involved its unusual shape, the extension functional y divided the adjustments to the terrain, this time the builders lev- nave into two asymmetrical parts. A similar ground plan elled the bedrock and made significant changes to the can be observed in the Early Medieval church of St Peter island’s geomorphology. The levelled surface measured (Sclavons, Cordenone, Italy: Gleirscher 2006, Abb. 3:1) at least 74.2 m2. The preserved mortar pavements are and St Augustin (St Margarethen, Austria: Gleirscher an indication of a uniform and well-maintained interior 2006, Abb. 6:1). Perhaps this subdivision had a similar of the church. liturgical significance as the subdivision of the nave by For the sake of clarity only, the two earlier churches stone screens with interlace decoration, for example, in will be referred to as the first and second (Pre-Roman- Molzbichl (Eichert 2012, 51−55; see the literature cited esque) church (of Phase 2), and the latter one as the third there). On the other hand, the extension could also (Romanesque) church (of Phase 3). simply reflect an ever larger congregation that had to be served in the church. 46 6 ORIENTATION OF THE MEDIEVAL CHURCHES 6.1 STATE-OF-THE-ART Unde et quando templum construebant, ori- entem spectabant aequinoctialem, ita ut lineae The orientation or alignment of the Medieval ab ortu ad occidentem missae fierent partes churches has long been the subject of study, but only caeli dextra sinistra aequales; ut qui consuleret recent methodological y advanced studies have yielded atque precaretur rectum aspiceret orientem. 7 a deeper understanding of the subject. I will first briefly present the current state-of-the-art in the topic and then This was still true in the middle of the 13th cen- the case of the Bled Island. tury, as the French bishop Guil ame Durando reports in Like the Greco-Roman temples before them, his summary of a centuries-long tradition of liturgical Christian churches have faced east since at least the commentaries: 8th century. The reasons for this were probably rooted in deep prehistory. The human division of the horizon De Ecclessia, et ejus partibus... Debet quoque into four parts is probably egocentric and has its origin sic fundari, ut caput recte inspiciat versus in the physiognomy of the human body, which knows orientem (...) videlicet versus ortum solis four directions: front, back, left, and right. Initial y, the equinoctialem... 8 positioning based on these four directions may have been based on an orientation point in the landscape, Modern analyses of the orientation of churches but early on this was replaced by astronomical obser- in Slovenia (Čaval 2010; Čaval 2014), Austria (Eichert vations. The division of the horizon into four parts is 2012, 244−251), north Italy (Spinazzè 2014, 2016a, certainly earlier than organised religion, but it gained 2016b), the Iberian Peninsula (Gonzáles-Garcia, Bel- symbolic significance with the advent of the latter. monte 2015; Valcárcel, Palmero 2018), England and Linguistic studies point to the great significance of the Wales (Ali, Cunich 2001; Hinton 2006, 2012; Sassin east in the worldview of prehistoric Indo-Europeans, Allen 2015) confirm the notion that an eastward ori- which undoubtedly has its origins in sun worship. As entation was at least sought, if not always achieved. It a contrast to east, Christianity associates the west with is noteworthy that each of these studies recorded non- satanic darkness, sadness, and death. In religions such random variations within the general pattern of eastward as Christianity, Islam, and Judaism, this division was orientation. For example, the mean orientation of 630 transferred to the conception of the afterlife (e.g., Evans medieval churches in Wales deviates by 8° to the north 1989, 234). from the geodesic true or cardinal east (Sassin Allen The beginnings of modern research on the orien- 2015, 158−159). tation of medieval churches date to the 19th century. The recorded non-random variations or differences At that time, on the basis of an 18th century English in the orientation of the churches were explained in vari- poem, it was hypothesised that medieval churches were oriented according to the point where the sun rose on 7 Orig. XV.4. Translation after Barney et al. 2006: Hen- the day of the church’s patron saint (Hinton 2012, 9−11). ce when they would build a temple they would face toward However, a thorough analysis of the Medieval texts re- the equinoctial sunrise, so that lines laid from the east to the vealed that the stated intention was to align the church west would divide the sky into equal parts on the right and with the rising sun at the equinox (Vogel et al. 1962). In the left, and thus whoever would take counsel or pray would his compilation of medieval sciences, Isidore of Seville look directly east. 8 (+ 636) summarised the procedure undertaken by the Durand, Rationale I.1/8. Translation by the author: Of gromatici in classical Antiquity as follows: the Church, and its parts... It must also be founded in such a way that the head looks correctly towards the east (...), that is, towards the equinoctial sunrise... 47 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 33: Slovenia, histogram of the eastern declinations of the Sun at Romanesque churches; marked in red are -23° (winter solstice), 0° (equinox) and 23° (summer solstice) (data source: Čaval 2010, Fig. 6.5). ous ways. For example, churches were aligned according most common of these dates was 15 August, the day of to the point at which the sun rose or set on the day of the the Assumption of St Mary (Valcárcel, Palmero 2018, church’s patron saint, on Easter or another important 185). So these two studies came to different conclusions, ecclesiastical holiday, or on the day the foundation stone even though they examined the same group of churches. was laid. In addition, differences in orientation could The next comparable study is the analysis of 143 be the result of regional, chronological or monastic dif- large Early Medieval churches in England and Wales, ferences, magnetic declination, altitude above sea level, most of which were founded in the 11th and 12th cen- local topography or regional context (Sassin Allen 2015, turies. Two-fifths of them were found to be aligned with 158−170; see also the literature cited there). The expla- either the rising or setting point of the sun on the day nations can be grouped into four categories that classify of the church patron and a quarter with the rising point the orientation of churches as symbolic, landscape or of the sun at Easter (Ali, Cunich 2001). The latter could topography related, liturgical, and structural (Valcárcel, be explained by the contemporary Catholic liturgy: Palmero 2018, 175−176; see also the literature cited there; Celebrating Easter on the correct date was one of the cf. Čaval 2010). There are pro and con arguments for each key elements of so-called orthopraxy, i.e., the accurate of the explanations listed, and none can explain all exam- performance of rituals (Geary 2001). ples recorded. Thus, it seems best to examine individual Spatial y much closer to the Bled Island is an analy- examples or at least groups of comparable churches in the sis of 181 churches from the Swiss Alps and Tuscany context of landscapes and other relevant domains (e.g., (Italy), built either before or during the Romanesque Pleterski 2014). Methodological y, it is essential to note period, i.e., between the 10th and 12th centuries. Most that astronomical orientations have only recently been of them were found to be aligned with the rising point analysed (Čaval 2010; Pleterski 2014; Gonzáles-Garcia, of the sun on the day of the church patron or with the Belmonte 2015; Spinazzè 2016b; Valcárcel, Palmero 2018). rising point of the sun on one of the four Marian feast While analyses of calibrated magnetic declinations are of days (Spinazzè 2016b). comparable methodological quality (Ali, Cunich 2001), al Most relevant for Bled Island is the study of the other studies analysed cartographic orientations and thus astronomical orientation of Romanesque churches in their results are not suitable for an astronomical analysis Slovenia. The result shows a wide range of alignments, of where on the horizon the sun sets and rises. either grouped around significant days in the solar and In the absence of review studies, it seems useful to lunar cycles (for example, solstices and equinoxes) or on briefly present existing studies that are relevant to Bled the dates of important liturgical feasts of pre-Christian Island. The first is an analysis of 167 churches built on origin. This includes the Assumption of St Mary, which the Iberian Peninsula before 1086. Most of them were has determined the orientation of eight churches. Only oriented according to the canonical y established date a few churches are aligned according to the day of their of the equinox: 25 March according to the Julian calen- patron saint (Čaval 2010; Fig. 33). dar. No orientations according to the day of the church On the basis of the comparative studies described, patron were found in the study (Gonzáles-Garcia, two conclusions can be drawn that are directly relevant Belmonte 2015). A subsequent study was limited to the to the Bled Island. First, all studies dealing with astro- selection to 49 pre-Romanesque churches, mostly built nomical orientation have found a high proportion of in the 10th century in the so-cal ed Mozarabic style. The astronomically oriented churches that focus on sig- results of this study found the prevalence of astronomi- nificant liturgical dates. Second, in the Alpine regions cal orientation, but this time largely based on the rising and perhaps also in England, a certain proportion of point of the sun on an important Christian holiday. The churches were oriented according to the feast day of their 48 6 ORIENTATION OF THE MEDIEVAL CHURCHES patron saint. Methodological y rigorous modern stud- Archeoastronomy cannot determine whether the ies have thus confirmed the predominant astronomical builders of this particular church were guided by the orientation of churches built between the 10th and 12th liturgical feast day, by the patron saint, or consciously centuries, which had escaped earlier researchers. by both. Nevertheless, while the distinction between these options is important to modern researchers, it probably made no difference to medieval builders and 6.2 BLED ISLAND CHURCH church users. The analysis of the architectural remains on Bled The orientation of the Church of the Assumption of Island shows that the orientation of the church buildings St Mary on the on Bled Island has been analysed using did not change after the first construction. The earliest state of the art methodology by Čaval (2010, 176−178, church was astronomical y oriented ( Fig. 29) and later and 309−314). The eastern astronomic orientation of the buildings simply followed this orientation. In other present church on Bled Island was measured as 76° 43’ words, only the Early Medieval pre-Romanesque first for the northern façade and 76° 34’ for the southern church was astronomically oriented. However, both façade. The northern façade corresponds to that of the the second church and the High Medieval Romanesque third church building and is therefore parallel to the third church ( Fig. 32) were oriented in accordance with orientation of the first and second church buildings. the earlier architecture. This explains why the Ascension The average declination of the sunrise was calculated Day orientation is an exception in the Slovenian study of at 11° 38’. Romanesque churches, while it is the rule in the study This declination of the sunrise coincides with the of Iberian and Italian Pre-Romanesque churches. The sunrise on 16 August according to the Julian calendar, astronomical orientation of a church according to the which corresponds to the Assumption of St Mary on 15 rising point of the sun on the day of the Assumption of August within a margin of error of one day. This means St Mary is a feature of Early Medieval churches built in that the church on the Bled Island is aligned with the the pre-Romanesque style. When new churches were sun’s rising point on the day of the Assumption. Since built in the Romanesque style in the High Middle Ages, the church on Bled Island is dedicated to the Assumption this rule was no longer observed. Nevertheless, where a of St Mary (see Chap. 11.1), the orientation corresponds Romanesque church directly replaced an Early Medieval to the feast day of the patron saint. church, the alignment was maintained. 49 7 ORIENTATION OF THE GRAVES 7.1 STATE-OF-THE-ART and The orientation of Medieval graves, like the Incedebamus autem tacentes,... contra ortum orientation of Medieval churches, has often been the solis solstitialem...10 subject of archaeological research. Indeed, alignment is an extremely significant aspect, especial y in religions as wel as where interment is the most important rite (Parker Pearson 1999). Medieval graves are usual y oriented in . .quasi contra ortum solis brumalem me ducere an east-west direction, with the head on the west side, coepit.11 i.e., facing east (e.g., Knific 2004; Brundke 2013, 32 with examples and literature from north-east Bavaria; This means that the 8th century Angles were not Nowotny 2018). This orientation must have had similar only able to determine the cardinal directions, but were or perhaps even the same reasons as the orientation of also constantly aware of the apparent movement of the contemporary churches, but the exact answer is not yet sunrise along the skyline. This was used for orientation known. According to the aforementioned 13th century in the landscape in much the same way as the wind rose French bishop Guil ame Durando, the deceased were (compass rose) was used for navigation at sea. aligned in this way so that they could be resurrected fac- Bede was used as a roughly contemporary analogy ing the place of Christ’s resurrection. However, this was to the cemetery on the Bled Island, but this knowledge probably only a rationalisation of the existing condition was present and widespread much earlier and was (Evans 1989, 234). discovered independently by different cultures (e.g., In Christian practice, this rule was often applied Ministr 2007). by interring parallel to the adjacent church (Gordon Thus, the orientation of graves in Medieval Europe 1971, 211−217; cf. Čaval 2010, 37−58 and the literature was certainly no accident. The intention of the mourn- cited there). However, Medieval graves in cemeteries ers − especial y those responsible for digging the grave − without a church also faced east. That there was sufficient was to orient the body towards the east, in accordance technical knowledge in the Early Middle Ages for the with the established ritual and most likely also with their astronomical determination of the true or cardinal east religious beliefs. It is no coincidence that the Medieval is testified, for instance, by Bede the Venerable (+ 735). meaning of the word “orient” was the direction towards As analyses of his work show, it was common knowledge the east. Hence, modern archaeological descriptions, for at the time that the true south was determined by the example the east orientation or the orientation of the shadow at the daily culmination point, i.e., when the head towards the west, may be artefactual y correct, but shadows are shortest (Hoare, Sweet 2000, 166; cf. Sas- they do not convey the correct meaning. Regardless, due sin Allen 2015, 154). In Bede’s time, people were able to the established terminology, I will use the terms “east to determine the exact direction of the sun’s rising at orientation” or (as a synonym) “orientation towards the the winter and summer solstices at any time of the year: 10 Bed. Venerab. Histor. lib. 5 cap. 12. Translation Miller . . ab euroaustro, id est, ab alto brumalis 1999, 189 (there incorrectly labelled as cap. XIII): We proce- exortus... 9 eded in silence,..., towards the north-east quarter of the hea- vens, where the sun rises at midsummer. 11 Bed. Venerab. Histor. lib. 5 cap. 12. Translation Miller 9 Bed. Venerab. Histor. lib. 4 cap. 3. Translation Miller 1999, 191 (there incorrectly labelled as cap. XIII): ...began 1999, 119: ...from the south-east quarter of the sky, that is, to lead me south east, to the quarter where the sun rises in from the highest point of the sun’s path... winter. 51 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND only just beginning (Achino et al. 2019). The development in neighbouring areas was similar to that in Slovenia, but it is not my intention to provide an in-depth history of research in the wider region. The solar azimuth hypothesis is relatively easy to test. Because of the Earth‘s axial tilt of 23.5°, the Sun’s azimuth on Earth changes with respect to the cardinal or true geographic east throughout the year (e.g., Strahler 2011, 54−55). As a consequence, the location of sunrise appears to move along the horizon; it is north of the cardinal east in summer and south of it in winter. The exact location depends on the geographical longitude and the horizon. On Bled Island, it is 61.5° ± 0.5° at the summer solstice and 127° ± 0.5° at the winter solstice ( Fig. 34; calculation by A. Pleterski; cf. Škerlj 1952, 112). If the orientation of the individual graves follows the apparent movement of the sunrise on the horizon described above, the hypothesis of solar azimuth is confirmed. For instance, an analysis of the cemeteries of the Native American Tutelo people from Pennsylvania (USA) − who are known from other sources to have buried their dead facing the rising sun on the day of interment − revealed a 90% coincidence (Evans 1989, 234−235). However, using this method, many authors have refuted the solar azimuth hypothesis for Early Medieval graves in the region (Škerlj 1952; Šolle 1959, 376−377; Hanuliak 1984, 109−114 for a review with literature for Slovakia; Herrmann 1989, 186; Brundke 2013, 32−33). The Early Medieval graves were not Fig. 34: Bled Island, azimuth of sunrise in yellow and grave aligned according to the sunrise on the day of interment. orientations in red. Various interpretations were thus sought, which can be divided into four groups as follows. Graves are oriented: east” in the following, even though etymological y they are pleonasms. 1. in the general (random) east direction; Nevertheless, it remains insufficiently clear accord- 2. parallel to the church (in the case of cemeteries near ing to which point or points in the east the mourners ori- churches); ented the grave pits and/or the bodies of the deceased. In 3. astronomical y, i.e., according to the Sun or other the early 1900s, researchers of Early Medieval graves put celestial body; forward the so-called solar azimuth hypothesis (German: 4. geographically, i.e., facing a specific place in the Sonnenazimutthese), which assumed that graves were landscape. aligned according to the sunrise on the day of interment (Niederle 1931, 84−85; Karaman 1940, 10−11). The first Let us take a closer look at each of the proposed such study for the territory of Slovenia was carried out interpretations. by Škerlj (1952), who analysed the Pristava site in Bled, First hypothesis. The orientation of graves is most the Ptuj castle, and three smaller cemeteries. According often explained as a general orientation toward the to him, the earliest graves of the Pristava cemetery in east (recently, e.g., Knific 2004; Brundke 2013, 32 with Bled, belonging to the so-called indigenous Romanised examples and literature for north-east Bavaria; Filipec population, were aligned according to sunrise on the 2016; Rapan Papeša 2016; Vyroubal, Pleše, Novak 2016). day of interment. This study had a notable impact on Some authors explain the deviation from the true east as archaeological practice in the region, inspiring excavators an error on the part of the mourners (Müller 2013, 152; of Early Medieval cemeteries to record the orientation of cf. Rempel 1966, 11), without stating what definition of skeletons in situ (e.g., Berce et al. 2012). However, the data the east the mourners were supposedly unsuccessful y on grave orientation was rarely published and over time aiming for. this practice was neglected with a few exceptions (e.g., Second most common hypothesis is, that the graves Sagadin 1985, 96−97). Modern studies on this topic are were aligned parallel to a contemporary church building 52 7 ORIENTATION OF THE GRAVES (Rajchl 2001, 127−128 for the Czech Republic; Müller in finding the exact point in the east towards which 2013, 152 for Germany; Azinović Bebek, Janeš 2016; the graves were aligned. In the search for this answer, Čimin 2016). Although this hypothesis can be easily two viable hypotheses are left: the astronomical and tested, it was not supported by measurements or other geographical orientations. field data in any of the studies listed, but it is attested in a Medieval poem written around the year 1200 (Fry 1999, 67). 7.2 METHODOLOGY Third hypothesis states that the graves are oriented astronomical y. This hypothesis was proven for a handful The analysis of church orientation has shown that of examples. An obvious astronomical orientation can be the study of the orientation of Early Medieval graves found in an Early Medieval cemetery around the church must also be carried out with similar methodological in the settlement of Břeclav-Pohansko (Czech Republic). rigour. It is somewhat surprising that the methodol- The graves were aligned in three different directions: ogy for measuring the orientation of graves has not towards the summer solstice, towards the sunrise at yet been systematical y considered. Below is a modest the equinox, or towards the southern full moon (Rajchl attempt of that. 2001). The graves of the Velkých Bílovicích cemetery in First, I must briefly discuss terminology. In ar- the Czech Republic had a similar orientation (Měřínský chaeological literature the terms grave, tomb, burial, and 1985, 24). In Slovenia, Early Medieval graves from the interment are often used interchangeably, but in this text site of Brezje above Zreče were aligned with the winter they are used with a specific meaning. solstice, i.e., with the sunrise at Christmas (Kaiser 2004). A grave is the sum of the archaeological y recorded Fourth hypothesis. Studies demonstrating the ori- remains of an interment; it incorporates, among other entation of Early Medieval graves in the direction of a things, the grave pit and its backfill(s), the skeletal re- specific place in the landscape are not abundant, but the mains, the grave marker, and the grave goods. Grave and known examples are supported by data and systematic burial (noun) can be used as synonyms, but the term analyses (Hanuliak 1984; Evans 1989, 244−250; Pleterski burial is ambiguous as it can be either a noun or a verb. 2014, 115−362). A tomb is a grave constructed in such a manner The above interpretations can be evaluated as fol- that the body is placed in an empty space and can be lows. The first two hypotheses are refuted with some con- reopened. Typically, tombs are cut into bedrock or fidence. In my opinion, the most common archaeologi- constructed of non-perishable materials such as stone. cal interpretation of the orientation of graves “towards Interment is the process that begins with the the east” in general is nothing other than mental inertia. preparation of a grave or tomb or other physical object Namely, instead of answering the question to which intended for the disposal of the body. Interment con- point due east the graves were oriented, this interpreta- tinues with the process of the disposal of the body and tion simply sidesteps the issue. However, rebutting the accompanying ritual(s) taking place near the grave/ solar azimuth hypothesis took considerable effort by tomb/other. A typical Catholic interment, for example, several researchers. They managed to refute the inter- includes the digging of the grave pit, the lowering of pretation that for the mourners the east was the point on the body (wrapped in a shroud or placed in a coffin) the horizon where the Sun rose on that particular day. into the grave pit, and the backfilling of the grave pit. As Bede demonstrates so vividly, knowledge about the The term burial ritual can be used as a synonym, but movement of the Sun was much more sophisticated in the term burial (verb) is ambiguous as it can be either a the Early Middle Ages. noun or a verb. The term funeral has a broader meaning The second most common hypothesis is that the and includes, in the catholic example, the mass and the graves are aligned parallel to the church. Although procession from the church to the cemetery. this hypothesis has yet to be confirmed by systematic In the context of orientation, the object of the analyses, it does not seem to be questionable for the discussion must first be defined. Are we interested in Medieval cemeteries with a church (German: Kirchen- the orientation of the body, grave pit, or a grave? A gräberfeld). However, it does not answer the question lively discussion could ensue, and the answer would of the meaning of orientation. As Durando testifies, by be different for different societies and thus different High Medieval times interment took place without (or archaeological contexts. with very few) exceptions in the churchyards, but the However, as is often the case with archaeologi- knowledge of the original meaning of the orientation cal practice, this question is rendered a mute point in had already been lost. the face of the imperfect reality of the archaeological Thus to answer the question of the cultural meaning record. The real question then becomes: Which part of of the eastward orientation, cemeteries without churches the archaeological record best reflects the orientation need to be examined. And the key to understanding of a grave? Since the orientation of graves is primarily the meaning of the orientation of Medieval graves lies studied from the perspective of interment, the orienta- 53 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND tion of the grave pit seems to be the most informative In archaeological practice, this means that dark soil is to reflect or mirror (for the expression, see Klejn 1987, typical mainly for Medieval cemeteries with continuity 41) the mourners’ conscious decision regarding the of burials in the Post-Medieval period. In earlier cem- alignment of the grave. Ideal y, the orientation of the eteries, dark soil is only present in alkaline soils with potential coffin or plank and the orientation of the high clay and moisture content. body would also be known. Following the example of In the cemetery of the Bled Island, the dark soil archaeoastronomy (e.g., Čaval 2010, 113), the average was present but not overwhelming. Nonetheless, among of all three measurements would be a good indication the at least 168 recorded burials, only 22 grave pits were of cultural processes taking place during the interment. identified and only eight of them in total. An even worse Unfortunately, in Medieval cemeteries coffins or ratio was found at the much larger cemetery of Župna burial planks are rarely preserved, if they were used at al . cerkev in Kranj, for example, where dark soil was pre- None were documented in the cemetery discussed here. dominant (Pleterski, Štular, Belak 2016; Pleterski, Štular, Additional y, grave pits are often difficult to detect. Belak 2017; Pleterski et al. 2019). This is especial y the case with Medieval cemeteries Returning to measuring the orientation of the near churches that have been used for centuries or Medieval burials it can be concluded, that reliable ar- even millennia. The reason for this is what is known in chaeological data on coffins and grave pits in Medieval archaeological practice as dark soil. This is a very dark cemeteries are very often lacking. Mostly, only the ori- brown to black, almost buttery soil layer that often incor- entation of skeletal remains can be systematical y meas- porates the entire volume of the archaeological record: ured. But even this method is fraught with difficulties. The stratigraphic layers through which the grave pits To meaningful y measure the orientation of skeletal were dug, the grave backfil s, and the layers covering the remains, the post-depositional processes that occur dur- graves. As a result, the occupation surface and individual ing the decomposition of the body must be taken into grave pits are hardly or not at all recognisable. account (Knüsel, Robb 2016, 661−667). A common issue In cemeteries that have been in continuous use, is the displacements of a body during decomposition in this dark soil is partly the result of repeated reburials, the cavity of a coffin or tomb, which can significantly where earlier graves were often exhumed and the grave affect the position of the remains (Knüsel 2014, 30−35; pits backfilled with the same soil. see the literature cited there). Such displacements can There is another more sinister cause for dark soil: be detected on plan drawings, for example, as detached adipocere, also known as corpse wax. This is a sub- limbs or unnatural bends of the spine. stance produced by the anaerobic bacterial hydrolysis The obstacles do not end there. Several problems of body fat in a corpse as part of the saponification also arise from the measurements themselves and hence process when the fat breaks down into glycerol and three types of orientation measurements must be dis- unsaturated fatty acids. The latter become saturated tinguished: measurements taken in situ, cartographic and penetrate the tissue and soil. The result is a very measurements, and astronomical measurements. persistent biocidal substance that has a wax-like ap- In the past, measurements in situ were often made pearance (Haglund, Sorg 1997, 568; Vranová, Marvo, with a magnetic compass, which is problematic because Rejšek 2015, 1421). Under extreme conditions, such as of the deviation of the magnetic north from geographic in tightly sealed coffins, this leads to mummification of north, i.e., magnetic declination. Another discrepancy the internal organs and muscle tissue (e.g., Papageor- can be caused by the magnetic deviation resulting, for gopoulou et al. 2010). Nevertheless, small amounts of example, from the proximity of iron structures or a adipocere always form in moist, clayey layers (Esteves high content of iron ore (Čaval 2010, 113−114). Both da Silva et al. 2009). In slightly alkaline soils, this sub- types of errors can be rectified if sufficient metadata stance accumulates over time ( cf. Gordon, Buikstra and paradata are available: what was measured, when, 1981) and this is how dark soil is formed. Dark soil with what type of equipment and with what accuracy. can be so saturated with adipocere that sniffer dogs can In archival excavations, these data are usual y missing. detect (Rebmann, Koenig, David 2000, 68, 122−123; Cartographic measurements by measuring skeletal Pintar, Glavaš 2017). From the point of view of ar- remains on plan drawings remains the most common chaeological stratigraphy, this is a post-depositional way to obtain the data. Two types of errors must be process in which the colour and consistency of the considered for this data. grave backfill and surrounding stratigraphic contexts The first is due to the accuracy of the recording in become so uniform that they can no longer be identi- the field. A reduction in accuracy can be observed, for fied during excavation. example, when comparing plans drawn at scales of 1:5, This effect disappears with the disintegration of 1:10 and 1:20. In my experience the accuracy is lowest adipocere, a process that lasts between a few centuries when georeferenced oblique photos are vectorised, and a millennium, and in exceptional cases several which is the prevailing modern method in Slovenia. millennia ( cf. Papageorgopoulou et al. 2010, Table 1). There is no method to eliminate this kind of error, so 54 7 ORIENTATION OF THE GRAVES it is important to take this into account during inter- pretation. The second type comprises the errors that occur when archaeological plans are transferred from relative to absolute coordinate systems, i.e. during georeferenc- ing. This is needed because archaeological documenta- tion is almost always produced in a Cartesian (plan drawings) or cylindrical (measurements with a total station) coordinate system, but the orientation with respect to true geographic north can only be measured in a spherical projection ( Fig. 35). In addition, the transformation is also necessary to transfer the data into a geographical information system (GIS) environ- ment (for example, QGIS). Working in GIS makes it possible to correlate the orientations of different graves within a single large cemetery, or of graves from several cemeteries, or of graves in relation to a selected point in the landscape. The error that occurs in the process of transfor- mation is due to the difference between Cartesian (for example, a quadrant grid in the excavation area) or cy- lindrical coordinate systems (for example, the national D96/Transverse Mercator) on the one hand and geo- graphic coordinate systems with spherical projections (for example, WGS84) on the other ( cf. Šprajc 1991, 11−13). Nevertheless, this error is known and negligible for our purposes if the transformation is carried out in a modern GIS system with the appropriate conversion formulae. The third type of grave orientation, as listed above, is astronomic. Astronomic orientation and declination are a prerequisite for archeoastronomy. They are calcu- lated from the azimuth of the Sun, the height of the ho- rizon, and the geographic latitude. The declination is the azimuth of the point on the horizon at which a particular celestial body, for example, the Sun or the Moon, rises or sets on a particular date. The declination is necessary to determine, for example, on which astronomical date the Sun rose in the direction of a certain burial (Šprajc 1991; Čaval 2010, 112−118 and the literature cited there). Fig. 35: Illustration of azimuth, cone and valve cartographic In summary, there are three types of grave orienta- projection (top); comparison of a spherical geographical coor- tion measurements: in situ, cartographic, and astronomi- dinate grid WGS84 (blue) and a geographical valve coordinate cal. The latter are the most precise and the only ones that grid D48/GK (black) for Slovenia (bottom). allows archaeoasronomical conclusions. However, in the archaeological practice of Medieval archaeology, the most common data that can be systematical y obtained are cartographic measurements of the orientation of 7.3 METHODS USED skeletal remains. Although subject to potential error, IN THE BLED ISLAND CASE STUDY such data can be successful y used for archaeological interpretation, including aspects of astronomy, if a rigid On Bled Island, only the cartographic orientation methodological approach is adopted. I will demonstrate data of the skeletal remains was available. This, then, one such example below. was the data used as the proxy for grave orientation in my study. It was measured on the plan drawings, which were original y drawn on a scale of 1:20. Measurements 55 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND were carried out in the GIS environment.12 First, the ured cartographic orientations with an accuracy of 1° well-preserved skeletal remains of 23 individuals were by subtracting 2° from the measured values. measured, for which the positions of the head and pelvis The calculated astronomical azimuth formed the was known and no traces of post-depositional move- starting point for the calculation of the declinations of ments were detected. The cartographic azimuth was sunrise and sunset. The eastern declination of the church measured for the line drawn between the joint between is thus 11° 38’ (Čaval 2010, 177). The eastern declination the head and the spine ( os axis) and the centre of the pel- of an individual grave (DE) was calculated by adding the vis ( os pubis). As the number of these measurements was value of the astronomical azimuth of the church (AA) too small for a quantitative analysis, the measurements to the subtracted difference between the cartographic of all adequately preserved skeletal remains were also azimuth (AC) and the deviation of the cartographic included. These were examples where the torso with the azimuth from the astronomical azimuth (AC − AA) to spine or the legs with the pelvis were preserved or where the eastern declination of the church (δE): it was possible to reconstruct the location of the head and at least the pelvis, knees, or ankles. Care was taken DE =(AA –(AC–(AC–AA)))+δE. to disregard parts of the skeletal remains that showed obvious signs of the post-depositional disturbances After replacing the values AA (AC − AA) and δE described above. Thus, of the 168 individuals recorded with constant values that apply to this particular case, the in the Bled Island cemetery, the cartographic azimuth final formula for calculating the cartographic azimuth of could be determined for 65 individuals. the graves on the Bled Island into declination is obtained: The azimuth was measured with an accuracy of 1/5°. Although the precision of the measurements was DE =(76–(AC–2))+11. affected by many of the errors mentioned above, the most important factor in this particular case was the The same procedure could be used to calculate the poor state of preservation of the skeletal remains. Based western declinations of the graves (DW): on comparison with similar data from the Župna cerkev cemetery in Kranj, the precision of the measurements DW = (256 – ((AC+180) – 2)) + (–2). was estimated at ± 1°. The results of the measurements are therefore rounded to the nearest whole number. It should be emphasised that the accuracy of the When it comes to archival excavations, a subse- calculated declinations decreases as the values go fur- quent measurement of each grave in situ with modern ther from 76° due to the rugged horizon. Especial y for methods is of course not possible. Thus, astronomical Graves 36, 38, 55, 59, 111, 112, and 113, the calculated azimuths and declinations were calculated from the deviations are approximations only. Nevertheless, the cartographic ones by comparing them with the archaeo- declinations could be calculated accurately from the astronomical measurements of the existing church astronomical azimuths and new measurements of the building. This was possible because all investigated horizon height if necessary ( Fig. 36). graves were located less than 10 m from the present church building. At the same time, the Bled Island is possitioned by between 200 m and 1400 m from the 7.4 BLED ISLAND CASE STUDY shore, which means that the horizon is relatively low and even despite the alpine surroundings. All this means As mentioned above, 65 graves were available for that differences in the calculations of declinations due to orientation analysis in the Bled Island case study. Since the distance between individual graves and the church the number of graves with measured orientations was can be neglected. too small for complex statistical analyses, the method of The eastern astronomical orientation of the present visual comparison between the measured and expected church on the Bled Island was, as mentioned, measured values was used ( Fig. 37). The astronomical azimuths as 76° 43’ for the north façade and 76° 34’ for the south of all graves were found to be between 57° and 173°, façade (Čaval 2010, 177). The cartographic orientation of that is, oriented between the northwest and the south. the same two façades measured in the same GIS environ- However, three graves stand out as statistical outliers. If ment as the grave orientation was 78° 59’ and 78° 49’. This these outliers are excluded, the azimuths are distributed means that the cartographic orientation differed from in the expected east sector between 65° and 111°. It can the astronomical one by 2° 15’ ± 1’. The astronomical therefore be said that the orientation of the graves on orientations were therefore calculated from the meas- the Bled Island is general y to the east. In the following, I will first test the two null hypoth- eses: the solar azimuth hypothesis and the hypothesis of 12 The measurements were conducted using a combinati- on of ArcGIS Desktop 10.5 software for the display and Pixel- orientation towards “the east in general”. Stick software to measure the angle of the computer screen. 56 7 ORIENTATION OF THE GRAVES Grave AA Ac δE δW Grave AA Ac δE δW 3a 67 65 22 9 69 78 76 11 -2 4a 77 75 12 -1 71 78 76 11 -2 5 88 86 1 -12 72 59 57 30 17 6 96 94 -7 -20 76a 80 78 9 -4 7 71 69 18 5 77a 83 81 6 -7 10a 90 88 -1 -14 78 87 85 2 -11 12 95 93 -6 -19 79a 77 75 12 -1 18 88 86 1 -12 80 74 72 15 2 19 83 81 6 -7 81a 86 84 3 -10 20a 96 94 -7 -20 82a 77 75 12 -1 22 96 94 -7 -20 84 80 78 9 -4 24a 86 84 3 -10 87 78 76 11 -2 29 90 88 -1 -14 90a 75 73 14 1 30 92 90 -3 -16 91 77 75 12 -1 32a 91 89 -2 -15 93a 81 79 8 -5 35 95 93 -6 -19 94 83 81 6 -7 36 101 99 -12 -25 95a 78 76 11 -2 38 113 111 -24 -37 98a 75 73 14 1 40 90 88 -1 -14 102 77 75 12 -1 42 91 89 -2 -15 103 76 74 13 0 45 82 80 7 -6 105 76 74 13 0 46 72 70 17 4 106 77 75 12 -1 49 91 89 -2 -15 109b 79 77 10 -3 50 98 96 -9 -22 111 106 104 -17 -30 51 92 90 -3 -16 112 104 102 -15 -28 54 96 94 -7 -20 113 134 132 -45 -58 55 99 97 -10 -23 115a 87 85 2 -11 56 95 93 -6 -19 118 78 76 11 -2 59 100 98 -11 -24 119 175 173 -86 -99 60 91 89 -2 -15 121a 84 82 5 -8 61 88 86 1 -12 121b 84 82 5 -8 65a 91 89 -2 -15 124 77 75 12 -1 67 78 76 11 -2 Fig. 36: Bled Island, grave orientation; values in grey are approximate values only: AA – astronomical azimuth; AC – cartographic azimuth; δE – eastern declination; δW – western declination. First, the solar azimuth hypothesis, according to cluded. Apart from one grave (3a), al others lie between which the orientation of each grave coincides with sun- 61.5° ± 0.5° and 127° ± 0.5° and could therefore have rise on the day of interment. On the Bled Island, the Sun been aligned with the point on the horizon where the rises above the horizon between 61.5° ± 0.5° (summer sun rose on the day of interment. solstice) and 127° ± 0.5° (winter solstice; calculation But a more detailed analysis refutes the solar azimuth by A. Pleterski; cf. Škerlj 1952, 112). The cartographic hypothesis. The expected mortality of people in the Mid- orientations of the graves on the Bled Island are between dle Ages was not evenly distributed throughout the year, 59° and 175°, or between 65° and 111° if outliers are ex- but was highest at the end of winter (e.g., Brown 1983). 57 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 37: Bled Island, diagram of cartographic azimuths of graves (top) and below map of selected groups : 1 − 76° ± 1°; 2 − 89° ± 1°; 3 − 94° ± 1°; 4 − others; 5 − marginal values; 0 − no data. The expected distribution of grave orientations should summer sunrise. In other words, if the solar hypothesis thus show a marked deviation from the easterly direction was correct, most interments on the Bled Island would to the south, where the winter sunrise is on the horizon. have occurred in summer and none towards the end of If mortality would be highest in January and February, winter, which is the direct opposite of expected. Since this for instance, most graves would be oriented between 104° is extremely unlikely and would be unprecedented in the and 121° ( cf. Škerlj 1952, 113). However, the measured existing records, the solar azimuth hypothesis can be ruled values deviate significantly to the north towards the out with very high confidence ( Fig. 34). 58 7 ORIENTATION OF THE GRAVES Fig. 38: Bled Island, map of graves in lines north of the present church: grey − present church building; red − idealised lines of graves with cartographic azimuth 4°; blue arrow − Grave 38 (data source: NMS archives AO Rn 222/4−5). The second nul hypothesis states that the graves are nomical y towards cardinal east within the measurement oriented toward the east in general and that any deviations error of 1°. All (graves 10a, 29, 30, 32a, 40, 42, 49, 51, are the result of the imprecision of the mourners. The 60) but one (Grave 65a) of the graves belong to Early expected values of orientations confirming this hypoth- Medieval cemetery and thus predate the earliest church esis would have a normal or Gaussian distribution with building. a peak at 90°. However, the actual values again deviate At the same time, these graves have a geographical significantly from the expected values, with the peak at orientation: 5.4 km from the Church of the Assumption 76°. Furthermore, the range of values ± 21° is too large of St Mary on Bled Island, in an easterly direction, is the to be explained by measurement error. As we learned Church of the Assumption of St Mary in the present-day from Bede, Early Medieval people were able to navigate town of Lesce. Both churches have the same orientation, themselves in the landscape based on the movement of both churches are dedicated to the same patron saint the Sun with great accuracy. This refutes the hypothesis and both were pilgrimage churches in the High Middle that the graves on Bled Island are oriented towards the Ages. According to sources from the beginning of the east in general and that the deviations are mere errors. 12th century, the church in Lesce was built “long ago” as Having refuted both null hypotheses, the only a proprietary church in honour of St Mary, it attracted two alternative hypotheses are that in the Bled Island pilgrims from near and far and their donations were kept cemetery, the mourners consciously aligned the graves by the feudal lord (Bizjak 2012, 38 with listed sources; either astronomical y or geographical y. cf. Höfler 2016a, 216). In Lesce, next to the medieval The orientation of the graves on Bled Island is clus- church, the existence of a Late Roman Early Christian tered around three declinations ( Fig. 37): the cardinal or church is attested by a poorly preserved mosaic (Božič true geographic east (89° ± 1°), the direction of the rows 1995). This implies the existence of an Early Medieval of graves (94° ± 1°), and the orientation of the church church as wel , but no corroborating archaeological (76° ± 1°). Let us have a look at each one. remains have yet been found. However, it is very likely The first cluster is represented by ten graves with that there was a place of special meaning in Lesce in the an orientation 89° ± 1°. These graves are oriented astro- Early Middle Ages; it is difficult to imagine that there 59 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND could have been an established pilgrimage centre in the an astronomical orientation. So, are these graves astro- middle of the 11th century (see Chap. 11.1) without a nomical y oriented? At least some of the mourners of tradition from the second half of the 10th century. that time are likely to have visited the church on the day The graves on Bled Island with an azimuth of of the Assumption, the feast day of the patron saint of 89° ± 1° can therefore be interpreted as aligned astro- the church. They thus possessed empirical knowledge nomical y with the cardinal east, which has a significant of the astronomical orientation of the church, that is, methodological value: It proves that people burying they could observe the sunrise first hand. However, the their dead on the Bled Island in the Early Middle liturgical significance of this orientation seems to have Ages had astronomical knowledge comparable to that no longer been significant, given the paucity of contem- attested by Bede the Venerable. These graves are also porary Romanesque churches oriented according to the very likely geographical y oriented to the site of the day of the Assumption of Mary. second church of the Assumption in Lesce. There is a The answer to the above question is thus No. At strong indication, but no direct evidence of activity in the level of interpretation of the interment, the High Lesce that would be contemporaneous with the graves Medieval graves were oriented parallel to the church on the island of Bled. and the astronomical orientation is coincidental. Even The second cluster is represented by six graves with if astronomical orientation was the intention of the an orientation of 94° ± 1°. All graves are from the Early burial ritual, in practice it was achieved by following Medieval cemetery (graves 6, 12, 20a, 22, 35, 54). In this the orientation of the church. case, the orientation of the graves is perpendicular to the The astronomical component certainly lost all grave rows. These rows deviate from the north-south di- meaning for post-Medieval graves. After the introduc- rection by about 4° to the east. Therefore, graves with an tion of the Gregorian calendar in the 16th century, the orientation of 94° ± 1° are perpendicular to these rows, sun no longer set on the axis of the church on the day which means they most closely follow the row system. of the Assumption. Nevertheless, no less than two thirds of the graves in the Importantly, the results clearly demonstrate that rows deviate from the 94° alignment by more than ± 3° the graves that are stratigraphical y contemporaneous ( Fig. 38). This suggests that alignment was not a major with the church show a systematic trend of orientation factor in the burial ritual for these graves. Rather, the parallel to the church. This is, to my knowledge, the orientation 94° ± 1° is a side effect of the rows, a case of first empirical evidence that graves in medieval church so-called equifinality (for the expression, see Dincauze cemeteries were aligned parallel to the church. 2000, 31). We do not know to which point the rows This information can also be used as a test of ac- were oriented, but it was not the top of Gradiška Hil , curacy: Half of the graves contemporary to the church as Pleterski had assumed (1996, 173−174). (outliers excluded) are aligned very precisely parallel The third cluster is represented by 14 graves ori- to the church building (± 1°) and three quarters have ented parallel to the church building. Their azimuth is a small deviation (± 3°). The rest of the graves have 76° ± 1° and the declination of the Sun is 11° ± 1°. One predominantly south declination, but their number is grave is from the Early Medieval cemetery, two are Pre- too small to further analyse the phenomenon of south Romanesque church graves, and nine are from the High versus north declination ( Fig. 39). This gives us a meth-Medieval cemetery. The remaining two graves are dated odological y significant indication of the precision of the to Late Medieval Phases 4 and 5. The most numerous and at the same time most important graves with this orientation belong to the High Medieval cemetery ( Fig. 18: G4; see Chap. 10.2). Their grave pits were cut into the bedrock and used for several successive reburials. So, these graves are remi- niscent of tombs that required considerable labour to build. Their precise alignment with the church building demonstrates that the deceased were placed in the tombs with great attention. The tombs were thus the result of a meticulously planned and careful y executed interment with a clear connection to the church building. The latter is direct archaeological evidence that these were Christian burials. As mentioned above, the church building was Fig. 39: Bled Island, graph of cartographic azimuths of graves astronomically aligned with the sunrise on the day coincident with church buildings: red − very precise orien- of the Assumption of St Mary according to the Julian tation (76° ± 1°), yellow − precise orientation (76° ± 3°), calendar, which means that all these graves also have grey − other. 60 7 ORIENTATION OF THE GRAVES orientation of graves in cases where the mourners are cemeteries are located on the line of sight connecting the known to have deliberately aligned a grave with a clear Višelnica midsummer bonfire site and the hill Šmarjetna landmark (in this case a church building): Mourners gora above the town of Kranj (Pleterski 2014, 274−276). on the Bled Island were habitual y able to align graves The listed cemeteries are positioned in the line of sight within ± 3° of the intended orientation. Therefore, all with great precision. The greatest deviation, that of interpretations in which a deviation of more than ± 3° is Bodešče, is only 25 m. This is just 1/3° declination when explained as an error or inaccuracy of the mourners are viewed from Višelnica. Furthermore, the western edge questionable, if not refuted. This is thus further objective of the sixth cemetery, Na Sedlu, is less than 40 m from evidence against the hypothesis of “in general east direc- this line of sight. tion”. At the same time, it is an important methodological Thus, cemeteries were not positioned in the land- guide for future studies on grave orientation. scape randomly. An additional element confirming this Outliers have already been mentioned. Four graves is the fact that four of the listed cemeteries are located (38, 72, 113, 119) have a statistical y significant deviation on a relatively flat plain that offers many alternative from the majority. locations. If the accuracy of the measurements is 1°, the Grave 119 has a cartographic azimuth of 173° ± 1°. statistical probability of four random locations being It is a post-Medieval grave that was interred after the aligned on a flat terrain is only 0.003% ( cf. Pleterski 2014, westernmost extension of the church was built. It is 115−116). At the same time, the viewshed analysis con- aligned perpendicular to the church wall, probably firms that all cemeteries except Rečica are visible from indicating that it was marked by a gravestone set into the Višelnica viewpoint, and on a clear day Šmarjetna the church wal , as was common from the 16th century gora is also visible in the distance ( Fig. 40). onwards (e.g., Rodwell 1989, 177; Mytum 2006, 97). This non-randomness provides two conclusive Grave 113 has a cartographic azimuth of 132° ± 1°. proofs. First, the Bled mortuary landscape was con- It belongs to the group of graves (111, 112, 114, 117) that sciously organized and thus the locations of Early were dug into a rocky hollow. Their orientation is thus Medieval cemeteries had a cultural meaning. Second, the consequence of geomorphological factors and not of the alignment of these locations was first established cultural selection. This group of graves is stratigraphi- and later observed with the naked eye from a single cal y contemporaneous with the Early Medieval graves viewpoint, that at the Višelnica bonfire site. in rows, where orientation was of less importance. The Višelnica bonfire site was thus an integral part Grave 38 has a cartographic azimuth of 111° ± 1°. of Bled’s mortuary landscape. It is the only place of the It is one of the graves in rows and it is apparent that its Bled mortuary landscape that is visible from Bled Island orientation deviates strongly from the neighbouring and Grave 72 is oriented perpendicular to it. Thus, the graves ( Fig. 38). The orientation of this grave supports grave − and symbolical y the entire cemetery − is linked the interpretation that a part of the cemetery was de- to the contemporary mortuary landscape described signed according to the orientation axes, which will be above ( Fig. 41). further discussed bellow (see Chap. 8.2). The last of the outliers is Grave 72, which is also discussed in more detail in the next section. 7.6 CONCLUSIONS Analysis has confirmed that the orientation of 7.5 MORTUARY LANDSCAPE many, but not al , of the graves in the Bled Island cem- etery was a conscious decision. This is perhaps the most The Early Medieval Grave 72 has a cartographic important conclusion for archaeology in general: A azimuth of 57° ± 1° and is unique in every respect. It is methodological y rigorous analysis of Medieval cemeter- isolated from the contemporary graves and is positioned ies can yield cultural information, especial y about burial right next to the green stone. As deliberated above (see rites. Several further conclusions can be drawn both Chap. 4.7), that stone was a spatial marker for a location regarding the Bled Island cemetery and archaeological of special meaning that was already ancient in Early methodology in general. Middle Ages. The orientation of the Grave 72 was not First, the church on Bled Island is astronomical y contingent on the terrain, it was not astronomical, and aligned according to the sunrise on the day of the patron it was not consistent with other features of the Bled saint of the church, St Mary of the Assumption. Most Island cemetery. Rather, it was related to the mortuary interments that occurred after the construction of the landscape. church were aligned parallel to the church. In this case, The mortuary landscape (for the term see e.g., the orientation of the graves is thus also a chronologi- William 2006, 179−214) of the Bled region consists of cal indicator. five contemporary Early Medieval cemeteries: Spodnje Second, some of the graves earlier than the church Gorje, Rečica, Pristava, Želeče and Bodešče. All these face due cardinal east. This proves a high level of as- 61 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 40: The Bled area, Early Medieval archaeological sites and viewshed from the Višelnica bonfire: 1 − Višelnica bonfire; 2 − Bled Island; 3 − cemeteries on the line; 4 − other cemeteries; 5 − Višelnica bonfire-Bled Island line (grave No. 72); 6 − Višelnica bonfire-Šmarjetna gora line above Kranj; 7 − view from the Višelnica bonfire viewshed (authors: E. Lozić, B. Štular). tronomical knowledge, comparable to that of the 8th Four. Extremely important for the interpretation century Angles described by Bede the Venerable. of the Early Medieval cemetery is the orientation of Third, for some of the graves that predate the Grave 72, which is likely one of the earliest. By its geo- church, orientation was not an important part of burial graphical orientation, this grave clearly expresses the rites, because the focus was on arranging the graves in intention of the mourners to associate this grave − and rows. Based on this, a general hypothesis can be put thus likely the entire cemetery − with the contemporary forward: In the so-called cemetery in rows (German: mortuary landscape. Reihengräberfeld) orientation of the graves was not Five, perhaps the most significant is the realisa- important; while the orientation perpendicular to the tion that one should not look for a single hypothesis to rows is predominant, the precision is ± 3° or less ( cf. explain the orientation of all Medieval graves, because Pleterski 2003). even in a single cemetery several different orientation systems may have been used simultaneously. 62 7 ORIENTATION OF THE GRAVES Fig. 41: Bled area, map of Early Medieval cemeteries (mapped from Fig. 40) and Grave 72; grave not to scale, but in correct cartographic orientation. In conclusion, of the five existing hypotheses to orientation − are all represented in the Bled Island explain the orientation of the Medieval graves, I have cemetery and are thus confirmed. While the orientation been able to objectively refute two: the solar azimuth parallel to the church was clearly contemporaneous with hypothesis and the general east hypothesis. the church, both the astronomical and the geographical The other three hypotheses − parallel to the church orientations predate the church. Thus, in addition to the building, astronomical orientation, and geographical symbolic, the orientation of the graves in the Middle Ages also has chronological implications. 63 8 SPATIAL ORGANISATION OF THE CEMETERY AREA 8.1 INTRODUCTION plained (among other things) that these demons lived in the sea, in springs or in forests and that people who The organisation of the cemetery area has been did not know God sacrificed to them or worshipped studied by Early Medieval archaeology since the 1900s. them as gods.14 The subject is closely interwoven with the orientation of Isidore of Seville (+ 636) wrote in the chapter on graves and is often part of the same discussions. Since sacred buildings that ancient people claimed that sanc- the 1990s it has regained traction (e.g., Pleterski 1996; tuaries had purification springs in front of the entrances Pleterski 2002; overview in Pleterski 2020), but it has not and that in his time there were buildings with sacred been widely adopted in Early Medieval archaeology. One springs where worshippers were purified.15 of the most common criticisms of existing studies was Charlemagne (+ 814) ordered everybody to destroy that they did not rely sufficiently on archaeological data. the sacred groves, stones and springs where rituals were In this chapter I present the analysis of the or- held.16 ganisation of the Bled Island cemetery area. Special The list of sins ( penitentiae), which was compiled care was taken to ensure that the interpretation was around the year 1000 and whose alleged author was based exclusively on archaeological data and relevant Archbishop Theodore of Canterbury (+ 669), states historiographical information. that all who offer sacrifices to trees, springs, stones, or enclosures should be punished.17 Bishop Thietmar of Merseburg (+ 1018) summa- 8.2 ORIENTATION AXES rised earlier sources about the Slavs who worshipped a OF THE EARLY MEDIEVAL CEMETERY spring near the Elbe (Germany).18 14 In the Early Medieval cemetery without the church Praeter haec autem multi daemones ex il is qui de caelo ( Fig. 18: G1 and G2) two orientation axes for the spatial expulsi sunt aut in mare aut in fluminibus aut in fontibus aut in silvis praesident, quos similiter homines ignorantes deum organization of the cemetery can be identified. The axes quasi deos colunt et sacrificant il is (De correctione rustico- are defined by the locations of the spring, the fireplace, rum 8). Similar elswhere: Post diluvium ... (a)lii adorabant and the green stone that marks the foot of Grave 72. The solem, alii lunam vel stel as, alii ignem, alii aquam profundam latter two features are attested archaeological y and are vel fontes aquarum, credentes haec omnia non a deo esse facta undoubtedly stratigraphical y contemporaneous with ad usum hominum, sed ipsa ex se orta deos esse (De correcti- the cemetery (see Chaps. 4.2 and 4.7). one rusticorum 6). The role of a small spring of artesian water on Bled 15 Delubra veteres dicebant templa fontes habentia, quibus Island can be interpreted based on medieval written ante ingressum diluebantur; ... Ipsa sunt nunc aedes cum sacris sources in which the pagan or non-Christian veneration fontibus, in quibus fideles regenerati purificantur: ... (Orig. XV.4) of springs13 is mentioned relatively frequently (Dowden 16 Item de arboribus vel petris vel fontibus, ubi aliqui stulti 2000, 39−45; Kajkowski, Kuczkowski 2017, 22). luminaria vel alias observationes faciunt, omnino mandamus, In evidence of the temporal and spatial ubiquity ut iste pessimus usus et Deo execrabilis, ubicumque inveniatur, of this phenomenon in Medieval Europe, I provide an tol atur et distruatur (Capit. I., p. 59 c. 65). extensive list of sources. 17 Siquis ad arbores, vel ad fontes, vel ad lapides sive ad St Martin of Braga (+ 580), a bishop in Gal aecia cancel os, vel ubicunque excepto in ecclesia Dei, votum voverit who was born in the Roman province Pannonia (Hun- aut exsolverit, iii. annos poeniteat (Paen. Theo. XXVII, 18). 18 gary), regarded pre-Christian gods as demons. He ex- Glomuzi est fons, non plus ab Albi quam duo miliaria positus, qui unam de se paludem generans, mira, ut incolae pro vero asserunt oculisque approbatum est a multis, sepe ope- 13 In Latin sources, the term fons, fontis is used. ratur. Cum bona pax est indigenis profutura, suumque [haec] 65 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Bishop Engilmar of Poreč mentioned a healing of Diomedes in the area of the river source of Timavo spring in 1037 in his commentary on the writings of Ar- ( Lacus Timavi): nold about the miracle of St Emmeram of Regensburg.19 In his biography of Bishop Otto of Bamberg, the At the very end of the Adriatic Gulf, there Benedictine monk Herbord (+ 1168) mentions the is a sanctuary of Diomedes cal ed Timavon veneration of a spring in the shade of an oak tree in (Timauon), which is worth mentioning. The Szczecin (Poland).20 sanctuary has a port, a wonderful sacred grove Helmold (+ after 1177) wrote in his Chronicle of and five springs of river water, which immedi- the Slavs that the Slavs do not allow Christians to come to ately flows to the sea as a wide and deep river their sacred springs and lakes because they believe that (after Šašel Kos 1999, 255). the water becomes impure if a Christian approaches it.21 Also, Bishop Vicelinus allegedly observed the worship of The most famous veneration of a spring today is forests, springs and other superstitions among the Slavs the spring at Lourdes, which was declared miraculous along the Elbe.22 Furthermore, he writes in a chapter on by the Catholic Church on 18 January 1862 and im- the conversion of Pribislav that after the consecration mediately became one of the most popular pilgrimage of the church, the Slavs in Oldenburg refrained from destinations in Europe. swearing oaths in forests and near springs and rocks.23 Medieval sources reflect the problems the Church A document from 1331, issued by the Franciscan encountered in eradicating the veneration of springs friar Franciscus de Clugia as a commendation of a that were considered sacred. These and similar pagan participant in a campaign to Kobarid, states that in traditions − which should be understood as an expres- the mountains countless Slavs worshipped a tree and a sion of the multipolarity of medieval culture (Geary spring there as a god.24 2001) − survived at least until the end of the Middle In Bled area, in 1253 and 1348 a lime tree is men- Ages. As indicated by Isidore of Seville, one way to tioned near the St Martin’s church that was a place of address this problem was to build churches near the gathering and a place of signing legal documents (Pleter- venerated springs (Kajkowski, Kuczkowski 2017, 23; ski 2014, 239−240 with source references). cf. Vlasto 1970, Geary 2001, 192; Štular, Hrovatin 2002, The above list contains selected examples from the 45−46). This approach was first recommended by Pope Middle Ages, but the veneration of springs is a broader Gregory I (+ 604) in his guidance for Christianisation phenomenon. The earliest written source in the region is (e.g., Demacopoulos 2008) and was followed into the Strabo’s (+ around 24 BCE) description of the sanctuary 10th century (Geary 2001, 190). In the wider surroundings of Bled Island, several terra non mentitur fructum, idem tritico et avena ac glandine medieval churches are known to have been built on refertus, laetos vicinorum ad se crebro confluentium efficit ani- top of or next to springs, for example, Bamberg Cathe- mos. Quando autem seva bel i tempestas ingruerit, sanguine et dral (Bavaria, Germany), the Church of St Michael in cinere certum futuri exitus indicium premonstrat (Thietmar Olomouc (Moravia, Czech) and the Church of St John I.3). under the Cliff (Bohemia, Czech) (Pleterski 2017, 38). 19 Hic quoque dum Ratisbonne constitutus apud Sanc- tum Emmerammum scripta meae parvitatis videret et legeret The Church often declared such constructions as places atque probaret, die quadam stans iuxta puteum seu fontem, of apparitions of Mary, and these churches regularly per quem recordabatur ibi sanam olim redditam fuisse fontem, became pilgrimage churches (Kajkowski, Kuczkowski vocavit me, et occasionem sumens ex eodem fonte, dixit quod 2017, 22−23). Two such examples in Slovenia are the scripsi; ... (Ann. Car. Sax. II.54; Kos 1911, št. 100). Church of the Nativity of St Mary on the hill of Homec 20 Erat praeterea ibi quercus ingens et frondosa, et fons (Štular, Hrovatin 2002, 54) and the Church of Our Lady subter eam amoenissimus, quam plebs simplex numinis alicui- of Sorrows, original y St Mary of the Assumption, in us inhabitatione sacram aestimans, magna veneratione colebat Leskovec near Krško (Rihter 2018, 7). (Herbord II.32). To sum the written sources, it can be said that there 21 Usque hodie profecto inter illos, cum cetera omnia com- munia sint cum nostris, solus prohibetur accessus lucorum et is clear evidence of the veneration of springs throughout fontium, quos autumant pollui Christianorum accessu (Hel- the Middle Ages, from the pagan Early Middle Ages to moldi I.1). the High and Late Middle Ages. In the sources listed, the 22 Nam lucorum et fontium ceterarumque supersticionum manifestations of this veneration only appear as slight multiplex error apud eos habetur (Helmoldi I.47). hints: as ritual purifications ( regenerati purificantur), 23 Et inhibiti sunt Slavi de cetero iurare in arboribus, fon- healing power ( sanam... fuisse fontem), sites of oaths tibus et lapidibus, sed offerebant criminibus pulsatos sacerdoti ( iurare), taboo for unbelievers ( pollui Christianorum ac- ferro vel vomeribus examinandos (Helmoldi I.83). cessu), isolated places outside settlements ( in arboribus) 24 ... usque ad locum de Cavoreto, ejusdem Dyocesis, ubi and places with supra-local significance ( innumerabiles... inter montes Sciavi innumerabiles arborem quandam et fon- tem, qui erat ad radices arboris, venerabant pro Deo; ... venerabant (Bian- ). The descriptions are scant, because they chi 1845, No. 699). were written exclusively from an antagonistic point of 66 8 SPATIAL ORGANISATION OF THE CEMETERY AREA Fig. 42: Bled Island. Left: meaningful places (1 − water spring; 2 − green stone; 3 − fireplace; 4 − grave oriented with respect to the line; 5 − line; 6 − normal). Right: graves oriented in relation to the line spring-fireplace. view, and they were understood by their authors in the church is 33.3 m), it was dedicated to Mary (of the As- context of the Old Testament (Vlasto 1970, 145−146). sumption), and it was a pilgrimage church (see Chap. However, the sources span almost a mil ennium and 11.1). most of the European continent, which is a clear indi- In addition to springs, stones or rocks ( petris, lapi- cation that it was indeed a widespread and persistent dibus) are mentioned as objects of veneration in three custom. of the above sources (Charlemagne, Bishop Theodore of Written sources contain enough information to Canterbury, and Helmold’s writing on Pribislav). This is learn that worship activities in or near springs did not evidence that in the pre-Christian Middle Ages springs leave direct archaeological traces, such as deposited and stones were objects of the same system of venera- objects or constructions (Kajkowski, Kuczkowski 2017, tion, which is further confirmed by recent research of 22). However, there is indirect material evidence of the Early Medieval pre-Christian religions (e.g., Hook 2010, venerated springs in the form of churches that were 31−46 with references to sources; Szczepański 2015, pas- later built near them. These churches, often dedicated sim). Written sources thus support the notion that the to Mary, can thus be used as proxy data (on the term cut piece of green tuff stone and the spring on the Bled and its use in archaeology, see Dincauze 2000, 30) to Island were objects of the same system of veneration. identify the specific springs that were venerated in the pre-Christian Early Middle Ages. So, the Early Medieval cemetery without the church There is sufficient proxy data to identify the spring on Bled Island had three objects of veneration that were on Bled Island as a venerated spring: The nearby church clearly recognised by mourners and other visitors: the is located in the immediate vicinity of the spring (the spring, the stone, and the fireplace. The latter two are distance between the spring and the altar of the earliest stratigraphical y contemporaneous to the cemetery and 67 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND the fragments of charcoal are evidence, that the fire- confidently measured ( Fig. 36). It has a cartographic place was in use during the interment (see Chap. 4.2). azimuth of 111° ± 1°, which is a strong deviation from Proxy data in the form of the church is evidence that the neighbouring graves ( Fig. 38). The reason for this the spring belonged to the same system of veneration. deviation is that it was oriented perpendicular to the All three were well visible and thus by definition also axis between the fireplace and the spring. The orien- served as spatial markers. More importantly, the use of tation of this grave is further independent evidence the fireplace firmly connects this system of veneration that the area of the cemetery was planned according with the burial practices taking place in the Bled Island to spatial markers. However, since this is the only cemetery. In other words, the spring, the stone and the sufficiently preserved grave with this orientation, any fireplace were spatial markers, part of a system of venera- further analysis into the meaning of this grave’s orienta- tion that was intertwined with the burial practices. To tion is not possible. our knowledge, this is the first time that such a direct The second axis connects the spring and the stone. connection can be made. The angle between these two axes is 23° ( Fig. 42). Once a direct connection between these spatial All other 9th/10th century cemeteries in the Bled markers and the cemetery has been established inde- area were organised according to a similar system (Žale pendently from the spatial analysis, the organisation of near Zasip: Pleterski 2014, 250−255; Dlesc near Bodešče: the cemetery area can be examined. The result shows Pleterski 2014, 259−261; Pristava: Pleterski 2014, that Bled Island cemetery was organised along two axes 264−274). This is further evidence of a single mortuary connecting the three spatial markers. landscape existing in the Bled area in the 9th and 10th The first axis connects the spring and the fireplace. centuries. This is also true for at least two other compara- It served as a guide for determining the positions and ble cemeteries in north Slovenia (Mali Grad in Kamnik: orientations of three graves (1, 27, 38). In the exten- Sagadin 2001; Štular 2007; Pleterski 2014, 290−293 and sion of this axis lies another Early Medieval grave Puščava near Stari Trg: Pleterski 2014, 354−356), which with female jewellery (83). Among those, only Grave is an indication that such an organisation of the Early 38 was sufficiently preserved for its orientation to be Medieval cemeteries was a broader phenomenon. 68 9 SPATIAL ANALYSIS OF GRAVE ATTRIBUTES 9.1 INTRODUCTION presence of charcoal, mortar or pottery in the backfill of the graves; graves carved into the bedrock). This chapter presents the results of the spatial analysis of grave attributes. Unfortunately, due to the poor preservation the selection of the observed attrib- 9.2 PITS utes was based on the accessibility of the data rather than on methodological considerations or specific scientific The excavators recorded relatively systematical y questions. Thus, the attributes discussed are those that the ground into which the individual grave pits were dug. could be systematical y recorded for all graves. The data They classified the graves as dug into the soil, partial y sources are the plan drawings (reburials and the arm dug into the bedrock, carved into the bedrock, or tombs. posture) and the descriptions in the field diaries (the Fig. 43: Bled Island, graves with charcoal in the filling (black); other graves in grey. 69 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND g089 g092 g096 g804 g805 g082a g084 g087 g077a g079a g095a g076a g078 g081a g086 g085 g098 g077b g079b g095b g076b g076c g077c g081b g082b g082c g082d g082e g093a g077e g101 g106 g094 g077d g102 g093c g093b g107 g103 g105 g080 g104 g083 Fig. 44: Bled Island, Phases 2b and 3. Top Harris matrix of cemetery group G4. Below Plan of the church buildings in Phase 2b (brown for the nave, yellow for the annex) and 3 (purple) and location of the contemporary grave chambers (coloured rectangles); graves with mortar in the filling are marked in yellow. Mapping of this attribute shows that most of the 9.3 BACKFILLS Early Medieval graves ( Fig. 18: G1 and G2) were dug into the top soil. Three graves on the south-west edge of the As already mentioned, the excavations on the Early Medieval cemetery were carved into the bedrock. Bled Island did not adhere to the stratigraphic method. This probably indicates that the edge of the area suitable Hence, the data on the backfilling of grave pits were not for burials had been reached. Similar happened at the systematical y recorded. There is also no description of northern edge of this group, where individual graves are the criteria according to which a particular artefact was either partial y cut into the bedrock or the remains lie recorded as part of a backfill rather then a part of a grave. on the bedrock. So, the interment of the Early Medieval So, the spatial distribution of artefacts and ecofacts is graves was limited to the area where the soil was deep the only information for interpretation. enough to dig a grave pit. Charcoal fragments are not uncommon in the South of the church graves carved into the bedrock backfill of Early Medieval graves (e.g., Sagadin 2001, predominate ( Fig. 18: G3−G5). This confirms the valid- 362−367). In the cemetery of the Bled Island, they were ity of the reconstruction of the geomorphology of the found in the backfill of four Early Medieval graves (18, island at the time when the cemetery was used ( Fig. 13). 22, 46, 63; Fig. 43). These graves were distributed over a 70 9 SPATIAL ANALYSIS OF GRAVE ATTRIBUTES Fig. 45: Bled Island, graves with pottery in the filling: green – prehistoric pottery; red – Early Medieval pottery; yellow – Late Medieval pottery; grey – no pottery (by B. Štular). small area no more than 2.95 m from the stratigraphi- other wal s were too far away for the mortar to fall into cal y contemporaneous fireplace (see Chap. 4.2). One of the graves: The west wall of the first church was 6 m the graves (46) was partial y dug into the burnt layer of away and the wall of the second church was 3 m away. the fireplace. This could have been the possible source This finding has chronological implications, since of charcoal in the grave. However, the charcoal was these graves belong to the stratigraphical y latest graves not preserved in the fireplace itself, implying that it of the High Medieval cemetery ( Fig. 44). The mortar was cleared away each time after it had been used. The thus proves the immediate proximity of the third church other three graves (18, 22, and 63) were also positioned at the time of the interment of the stratigraphical y later in close proximity to the fireplace. The proximity of graves, but not of the stratigraphical y earlier graves. the fireplace and the graves with charcoal fragments The latter must have been contemporary to the second in the backfil , on the one hand, and the fact that the church (see Chap. 10.3). fireplace was avoided by the graves, on the other, do Sand was also recorded in the backfil s of two graves not leave many possibilities for interpretation. The (67 and 71). Although a more precise interpretation is source of charcoal in the graves is precisely this fire- not possible, it is probably not a coincidence that these place. Moreover, the fireplace was used in the period two graves are the only graves of Phase 2. between the digging and the backfilling of the grave The pottery fragments were recorded in 13 graves pits, i.e. during the interment. ( Fig. 45). Eight graves (1, 24, 39, 45, 46, 49, 55, 120) Mortar fragments were found in seven High Me-contained prehistoric pottery. Their distribution cor- dieval graves (77a, 82a, 84, 87, 88, 93a, 101). The only responds to the distribution of the prehistoric finds, possible source of mortar was a mortar-plastered build- which means that these are residual finds, i.e., pottery ing. The only possible source was the adjacent west wall in secondary position (see Bitenc, Knific 2020a, 22−40). of the nave of the third church ( cf. Šribar 1966, 89). The Four graves (61, 71, 84, 115) contained fragments of 71 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 46: Bled Island, map of reburials. wheel-thrown pottery dated to the 13th century and able to identify the grave pits in exceptional cases, there later (Bitenc, Knific 2020b, Pl. 6: 5, 6, 8−13). These is only indirect evidence for this, namely the position fragments are interpreted as infiltrated finds. One of the skeletons. So, I define reburials as those where a grave (45) contained two fragments of possibly Early grave contains a well-preserved skeleton with a skul Medieval pottery (Bitenc, Knific 2020b, Pl. 6: 3, 4). and/or other bones placed (more or less) carefully As individual fragments of contemporary pottery are along the edges (for example, Graves 81 and 93; field known from other Early Medieval cemeteries (e.g., documentation N 11). Pleterski et al. 2019, 68, 69, 72 etc.), it is reasonable In the majority of the cemetery, reburials appear to assume that the pottery is related to the grave. A only sporadical y and only as a single or double reburial. more precise interpretation − for example, whether Outside of the High Medieval cemetery, a triple reburial the fragments were intentional grave goods, remains of is only found in two graves (39 and 109). In the first the interment ceremony, remains of other activities in case (39), it was a deliberate reburial. Judging from the the vicinity of the cemetery, or something else − must position of the individual bones, the individual bones be sought in other comparable cemeteries excavated of several deceased persons were placed in an organic, with modern methods. rectangular receptacle, probably a wooden box. In the second case (109), there are bones in the backfill that can probably be interpreted as the reburial of bones 9.4 REBURIALS discovered by the builders of the Late Medieval Gothic church in Phase 4. In the cemetery on Bled Island, 27% of the graves In terms of reburial, the High Medieval cemetery show traces of subsequent burials or reburials in the differs considerably from the rest of the cemetery. The same grave pit ( Fig. 46). Since archaeologists were only 72 9 SPATIAL ANALYSIS OF GRAVE ATTRIBUTES Fig. 47: Bled Island, map of the position of the deceased‘s hands: 0 − no data; 1 − in the lap (hypogastric); 2 − one in the lap, the other ful y extended; 3 − on the chest as in prayer; 4 − ful y extended. vast majority of graves contain reburials, including the and High Medieval cemeteries, this arm posture is clearly cases of quadruple and quintuple reburials. predominant (92%) only in the High Medieval cemetery. Reburials in the same grave pit were thus found The deceased with one arm in the lap and the other only sporadical y in most of the cemetery and are inter- outstretched or with both arms extended along the body preted as a consequence of the lack of space. However, are evenly distributed throughout the site, except in the reburials in High Medieval cemetery and in the Early High Medieval cemetery. The only deceased with hands Medieval Grave 39 were intentional and, as such, part clasped on the chest (Grave 67) was buried in front of of the interment customs. the second church’s threshold. I interpret these findings as follows: In Early Medi- eval cemetery ( Fig. 18: G1 and G2) the position of the 9.5 ARM POSTURE arms was not an important component of burial rites, as all postures are equal y represented. The only skeleton The arm posture of the deceased is another topic in from Phase 2, which is contemporary with the second the Early Medieval archaeology of death and burial with church, is also the only deceased in the entire cemetery a long history of research but few quantitative analyses. to have its hands clasped on its chest in the prayer posi- In this case, the arm postures were taken from the plan tion; this is probably less a chronological feature than drawings and divided into four relatively loosely defined a peculiarity of this grave. The posture of the arms was groups ( Fig. 47) in which the arms are positioned: in the especial y important in High Medieval cemetery ( Fig. lap; one arm in the lap, the other outstretched; clasped 18: G4), where the arms of all but one of the deceased on the chest; extended along the body. are in the lap. The deceased with one arm in the lap and Deceased with their hands in their laps are found the other outstretched are prevalent in the church graves throughout the cemetery. Although found in both Early of Phases 3 and 4 ( Fig. 18: G5). 73 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND 9.6 CONCLUSIONS The arm posture was not an important part of burial rites in this period. Given the sparse data, the spatial analysis of the High Medieval (and later) graves were habitual y grave attributes yielded a surprising amount of valuable cut into the bedrock. In the High Medieval cemetery information. west of the church the grave pits cut deep into the Early Medieval cemetery was positioned exclusively bedrock can be described as tombs. Mortar frag- where there was relatively deep soil. The charcoal in ments are evidence that these tombs were positioned the graves showed that a fireplace was burning during directly against the church wal . Repetitive reburial of interment. Reburials were rare and mainly due to a lack the bodies with arms in the lap was the norm in High of space. However, there was one intentional reburial. Middle Ages. 74 10 ABSOLUTE CHRONOLOGY OF THE BLED ISLAND CEMETERY 10.1 INTRODUCTION According to the current state of research, C14 dat- ing and typo-chronological dating for Early Medieval It might seem that archaeology is a straightfor- graves are similarly (in)accurate and allow dating with ward process of discovery fol owed by descrip- the accuracy of approximately one century. Namely, (in) tion, yet it is accompanied by interpretation at accuracy of C14 and other radiometric dating methods every step. Interpretation draws on theory − (e.g., Scott, Cook, Naysmith 2007; Michczyński 2007; cf. our rationalizations of our experiences in the Pleterski 2010, 86) is too often overlooked in archaeol- world − in order to make sense of how and why ogy, leading to overly high expectations. people of the past treated their dead, disposed of Only jewellery was suitable for a typo-chrono- their remains, and provided ways for the dead logical analysis of the grave inventories to explain the to coexist with the living chronological development of the site as a whole, as it (Parker Pearson 1999, 29). was the only type of artefact from this site that allowed dating with the precision of at least a century. The I could not have hoped to describe the arduous chronological analysis was based on the typo-chrono- path of archaeological inference as succinctly as Parker logical scheme developed by Pleterski (2013a) for grave Pearson did two decades ago (for a general overview of inventories dated by the C14 method and updated by the topic see Lozić, Štular 2007; Fahlander, Oestigaard Rihter (2020) on the basis of a stratigraphic analysis of 2008). And this path begins with a chronological analy- the Župna cerkev cemetery in Kranj. sis, that is, dating first the individual graves and then Twenty-seven temple rings and three finger rings the cemetery. could be dated ( Figs. 48−51). Twenty-seven datable In the Bled Island case study, three data sources artefacts came from Early Medieval cemetery, one from were available for absolute dating: architectural ele- High Medieval cemetery and three were found outside ments, artefacts, and radiocarbon analysis of organic the graves. All dates, including the C14 date (σ1 934 ± samples (hereafter C14). However, only three architec- 38 CE) of the skeleton from Grave 72, are between tural elements and 30 artefacts could be dated to the 800 and 1030 CE. The σ1 mean value of the C14 date Early Middle Ages. Moreover, the fact that the exca- was the year 934 CE, while the mean value of all typo- vations occurred more than six decades ago severely chronological dates was the year 935 CE, which is a limited the availability of suitable C14 samples. remarkable correlation. From a methodological point of view, the dating As already described, the graves in the Bled Island of architectural elements is the least reliable method. cemetery are neither spatial y nor chronological y evenly In this method, the remains of church architecture are distributed. So, the graves were divided into five groups usual y dated as either pre-Romanesque or Romanesque. (G1−G5) on the basis of their stratigraphic and spatial However, the only data available for Bled Island were the proximity ( Fig. 18). I was able to date four cemetery ground plans and, to some extent, the building materials. groups ( Fig 18: G1, G2, G3, G4) and the earliest three Due to the paucity of preserved remains and because church buildings. No data was available for dating the there were very few suitable securely dated analogies, remaining cemetery groups ( Fig 18: G5, G6, G7) or the dating of architectural elements was dispensed with. later churches, which is why they were not discussed The dating tools used were thus the two available in this chapter. C14 dates, and the typo-chronological analysis of 30 artefacts from the graves. 75 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND NO0100_0506 NO0100_0607 NO0100_0608 NO0100_0708 NO0100_0808 NO0100_0909 NO0500_0610 NO0504_0610 NO0700_0810 NO1700_0610 PR0103_0000 PR0103_0200 PR0105_0000 PR0201_0000 PR0501_0000 Fig. 48a: Bled Island, types of grave jewellery mentioned in Fig. 48b. 10.2 EARLY MEDIEVAL CEMETERY The only datable find from the east group of graves in the Early Medieval cemetery ( Fig. 18: G2) is the ring Typo-chronological y, the most informative was from Grave 111 ( Fig. 51: 13), which is dated between the Early Medieval cemetery ( Fig. 18: G1 and G2). In 840 and 1010 CE. The find comes from the grave backfil Early Medieval archaeology, the most commonly used and it is not certain whether it belongs to this grave at al . date is the intersection of the dates of the objects in the Regardless, it confirms that this group of graves belongs grave ( Fig. 52). I used it with the exception of Grave 4, to the Early Medieval cemetery, but this single artefact which did not have an intersection date. The finger ring does not allow any further interpretation. it contained appears to be a so-called “old object”, i.e. an Several conclusions could be drawn about the Early artefact made long before it was deposited in a grave. This Medieval cemetery based on this chronology ( Fig. 52). grave was thus dated on the basis of the latest artefact. First, the typo-chronological dating matches with Thus, 14 graves could be absolutely dated on the the stratigraphic superposition, that is, the absolute and basis of the artefacts. Six of these could only be dated relative dating are in concord. Since the two analyses imprecisely on the basis of the temple rings with a single were carried out independently of each other, such an thickening ( Fig. 48). overall correlation confirms the accuracy of the results. 76 10 ABSOLUTE CHRONOLOGY OF THE BLED ISLAND CEMETERY Dating Type Grave No. No. on Figs. 49−51 dating source earliest latest NO0500_0610 840 960 1a 1 JR KŽC NO0504_0610 840 960 1a 2 BŠ NO0100_0808 870 1000 3a 3 JR KŽC NO0100_0808 870 1000 3a 4 JR KŽC NO0700_0810 990 1030 4a 6 AP13 PR0501_0000 x x 4a 7 PR0201_0000 800 950 4a 8 JR KŽC PR0501_0000 x x 4a 10 NO0100_0808 870 1000 12 14 JR KŽC NO01mm_0808 870 1000 12 15 JR KŽC NO0100_08mm 870 1000 24 17 JR KŽC NO01mm_08mm 870 1000 24 18 JR KŽC NO0100_0608 830 1000 27 19 JR KŽC NO0100_0608 830 1000 27 20 JR KŽC NO0100_0808 870 1000 34a 22 JR KŽC NO0100_0808 870 1000 36a 1 JR KŽC NO0100_0909 920 1000 36a 2 JR KŽC NO0100_0808 870 1000 37 3 JR KŽC NO0100_0808 870 1000 37 4 JR KŽC PR0105_0000 840 1010 37 5 JR KŽC NO0100_0808 870 1000 38 6 JR KŽC PR0103_0000 870 1010 41 8 JR KŽC PR0103_0200 x x 45 10 NO1700_0610 870 970 45 11 JR KŽC NO1700_0610 870 970 45 12 JR KŽC NO0100_0506 830 1000 49 13 JR KŽC NO0100_0808 870 1000 49 14 JR KŽC PR0103_0200 x x 58 20 NO0100_mmmm x x 59 21 NO0100_0607 830 950 83 9 JR KŽC NO0100_0708 900 1010 83 10 JR KŽC NO0100_0808 870 1000 96 11 JR KŽC NO0100_0909 920 1000 15 JR KŽC NO0100_0808 870 1000 16 JR KŽC PR0201_0000 800 950 17 JR KŽC special type 18 Fig. 48b: Bled Island, typo-chronology of grave jewellery. 77 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 49: Bled Island, grave goods in Graves 1, 3–6, 12, 16, 24, 27, 30, 34 and 35. 1–4,6–10,14–20, 22 bronze; 5,11–13,23 iron; 7,9 glass, 21 lead. Scale = 1:2. Drawing: 1,2,7,16,21 I. Murgelj, 3,4,6,7,10–15, 17–20,22,23 D. Knific Lunder; 5, 8 V. Stare; 2,9,10 (Štular 2020, Pl. 1). 78 10 ABSOLUTE CHRONOLOGY OF THE BLED ISLAND CEMETERY Fig. 50: Bled Island, grave goods in Graves 36–38, 41, 45, 49, 54–56, 58, 59, and 61. 1–6,8,10–14, 20,21 bronze; 7,9,16–19, 22 iron; 7 bone or horn, 11,12 enamel; 15 lead. Scale = 1:2. Drawing: D. Knific Lunder (Štular 2020, Pl. 2). 79 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Fig. 51: Bled Island, grave goods in Graves 68, 69, 75, 77, 78, 83, 96, 100 and 111 (1–13) as well as scattered finds. 1–5,8,19,20 iron; 6,7,9–13,15–18 bronze; 14 silver. Scale = 1:2. Drawing: 1,2,4,7,12,20 I. Murgelj; 3,5,6,8–11,13–19 D. Knific Lunder (Štular 2020, Pl. 3). 80 10 ABSOLUTE CHRONOLOGY OF THE BLED ISLAND CEMETERY Second, the reason why most of the finds are from evidence that individual bones, rather than limbs or the most recent graves is because the earlier graves have body parts, were moved during reburial. In particular, been disturbed and not because more people would be the bones moved were mostly skul s and long leg and buried in the later period. arm bones. This is evidence that the decomposition of Third, it is possible to define the time span of the body was already in the final stages of skeletonization the Early Medieval cemetery based on three pieces of (i.e., the final stage of decomposition, during which the information. To begin with, the longest possible time soft tissues have decayed to the point that the skeleton span (range of dates) is between 830 and 1030 CE, and is exposed), as the ligaments of the legs are the last soft the shortest (intersection of dates) is between 960 and tissue to decompose (Pinheiro 2006, 110−112). To give 990 CE. The range of mean dates is between 900 and an example, skeletonization of exposed bodies of large 1010 CE; the latter is consistent with my intuitive dat- mammals takes up to five years (Hil , Behrensmeyer ing. Then, the predominant type of jewellery is temple 1984; Rebmann, David, Sorg 2000, 126; cf. Clark, Worel , ring with a single thickening (Type NO0100_0808; all Pless 1997, 159−160). However, in the case of buried type codes after Rihter 2020), which occur in both the human bodies, the process is longer and depends on stratigraphical y earliest and latest graves. The span of several factors (Daniell 1997, 109−121; Pinheiro 2006). use of these rings between the years 870 and 1000 CE is The only known factor in the case study on Bled Island thus another indication of the duration of the cemetery. (and in most other archaeological examples) is the soil And final y, also noteworthy is the absence of temple conditions after burial. The clayey soils on Bled Island rings with a hook, which were ubiquitous in this region slow down the skeletonization process ( cf. Esteves da in the ninth century, but occurred only exceptional y Silva et al. 2009). In view of this, one can roughly es- after the first decade of the 10th century (Rihter 2020; timate that the skeletonization process on Bled Island cf. Pleterski 2013a). So, their absence indicates that this took at least a decade. However, after only a decade, hair cemetery was not used until approximately the first and nails are still preserved and due to the adipocere the decade of the 10th century.25 I inferred on the basis of al unpleasant smel s are still very much present ( cf. Parker three pieces of information, that the interment occurred Pearson 1999, 14). throughout the 10th century and in the first decade of Second, the taboo of reburial. Since Medieval peo- the 11th century. ple believed that the soul does not leave the body until Fourth, in eleven decades four generations were it is completely disintegrated (Daniel 1997, 62−64; cf. interred in the Early Medieval cemetery on the Bled Williams 2006, 83−84), the taboo of reburial must have Island. The stratigraphy evidences at least four genera- persisted at least as long as the unpleasant smel s. Indi- tions; the duration of the use of the cemetery evidences rect evidence of such a taboo can be found in a spell from no more than four generations. Taken together, then, the end of the 15th or beginning of the 16th century: Die exactly four generations were interred there. The evi- chunnen mir hewt als wenig geschaden als der man, der dence and inference for this are presented in more detail vor xxx Jaren ist pegrabenn (Javor-Briški 1998, 9).26 This in the discourse below. spel is to be understood in the sense that the remains in a grave lost the social status of the deceased after three In cemeteries with marked graves, with sufficient decades; at that point the remaining bones acquired the space, and with clear rules for choosing the burial plot, social status of an inanimate objects. In oder words, after centuries could pass without a single reburial. In the Bled three decades, the body of the deceased ceases to be a Island cemetery none of these conditions was met and taboo and becomes a mere disposable object. In the case grave disturbance or even reburial was commonplace. of the Bled Island reburials, the attitude of the mourn- In the Early Medieval cemetery stratigraphic analysis re- ers towards the exhumed bones was disrespectful from vealed three triple reburials, i.e., four deceased (original the point of view of a modern observer, e.g., individual burial plus three reburials) were interred successively bones were carelessly disposed of. This indicates that in the same plot. This means at least four generations. the time of the respectful taboo − 30 years in the above The question remains, however, whether there were example − was already over at the time of the reburial. only four generations or more? As noted above, the Third, the generation time, that is, the average time cemetery was in use for 11 decades. The only missing between two successive generations in the lineage of a piece of evidence is the duration of a generation. There population. It can be calculated with various formulae are three clues from different sources that can answer and is typically between 22 and 33 years in human this question: grave taphonomy, reburial taboo, and populations (e.g., Bienvenu, Legender 2015). However, biological generation. the data for the Early Medieval population in the Bled First, grave taphonomy, or more precisely, observa- area is insufficient. For other archaeological popula- tion of the disarticulation of the corpse ( cf. Knüsel, Robb tions, there are rough estimates ranging from 25 (e.g., 2016, 667−668). The plan drawings of the graves provide 26 Today they can do me as little harm as a man buried 25 I would like to thank A. Pleterski for this information. 30 years ago. 81 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND first decade of the 11th century four generations were Grave No. Intersected dat. Stratigraphy buried in the Early Medieval cemetery on the Bled from to mean Island ( Fig. 54). 1a 840 960 900 x Based on the stratigraphic analysis, it can be said 27 830 1000 915 1/2 that at least seven deceased were buried in the first generation ( Fig. 53: graves 1b, 7b, 27, 38, 45, 61, 72). 45 870 970 920 1/2 Considering the chronology of the artefacts and the 83 900 950 925 x duration of each generation, the first two generations 72 895 972 934 x were interred in the first half of the 10th century. The earliest graves are 1 and 27. Together with most of the 38 870 1000 935 1/2 first generation graves (38, 45, 72), they belong to a group 37 870 1000 935 3 that is homogeneous both in terms of their location in 41 870 1000 935 3 the cemetery and their unorthodox orientation (see Chap. 7; Fig. 54a: B). 49 870 1000 935 3 The second generation probably included 14 de- 3a 870 1000 935 4 ceased ( Fig. 53: graves 1a, 7a, 13, 14, 15, 28, 31, 35, 37, 12 870 1000 935 4 42, 46, 54, 57, 64b), but due to the insufficient number of stratigraphic relationships, the precise distinction 34a 870 1000 935 4 between the first two generations cannot be established. 36a 920 1000 960 4 With this generation, a new system of organisation of 4a 990 1030 1010 4 interment was introduced: burial in rows or Reihen- gräberfeld ( Fig. 54a: C). This means that approximately Fig. 52: Bled Island, chronological distribution of graves based in the second quarter of the 10th century a significant on the earliest artefact in the grave in chronological order change in burial practice was introduced in the Bled (according to the median date); the stratigraphic position is Island cemetery. simplified (for the actual stratigraphic position see Fig. 54). At least 18 deceased were buried in the third gen- eration between the sixth and eighth decades of the 10th century ( Fig. 53: graves 4b, 5, 6, 8, 9, 12, 16, 21, 23, 34b, 36b, 40, 41, 43, 56, 59, 62, 63). At the beginning of this generation, the entire area of the cemetery suitable for Hazelwood, Steele 2004, 677) to 35 (e.g., Arnold 2002, burials was already occupied and the burials thus had 136) or exceptional y 40 years (Pleterski 2013a). If the to be “squeezed in” ( Fig. 54b: D). latter example is excluded, the average generation time The fourth generation includes at least 14 graves in archaeological populations is estimated to be 30 years. ( Fig. 53: graves 3, 4, 18, 19, 22, 25, 30, 34a, 36a, 39a, It can be concluded that there is direct evidence that 39b, 39c, 49, 60) dating to the last two decades of the the graves in the Bled Island cemetery were not exhumed 10th century and the first decade of the 11th century or disturbed for at least a decade, which is as long as the ( Fig. 54b: E). skeletonization takes. However, based on the Medieval The most interesting among them is Grave 4, in taboos and the generation time, we estimate that this which by far the latest artefact in the cemetery was period lasted approximately three decades. found. The artefact in question is a temple ring (Type This means that the burial of four generations NO0700_0810; Fig. 49: 6) dated between 990 and required at least nine decades between the end of the 1030 CE. The mean date is 1010 CE, which is consistent first generation and the end of the fourth generation. with the estimated end of use of this cemetery. This grave However, all members of a generation are never buried also differs from all other Early Medieval graves in that in the same year. The range of mean dates of the first it is the only grave that is oriented parallel to the church generation graves dated with artefacts ( Fig. 52: graves building. This is a reliable indication that the church was 1, 27, 45) tells us that the burial of this generation already built at the time of this burial. took place over two decades. Adding these two to the Grave 4 is thus a biritual burial of a woman. The nine decades gives an expected time span of eleven grave pit was dug parallel to the church building, as was decades for the burial of four generations on the Bled customary for Christian church burials. However, the Island. This corresponds exactly to the time span of the body was placed in the grave together with headdresses cemetery calculated in the typo-chronological analysis in accordance with pre-Christian burials (i.e., burials in presented above. cemeteries without a church). As this is the only such grave in the cemetery, it is evidence that this biritualism Returning from the discourse, I can thus say was a unique event. One could even assume that this with some conviction that in 10th century and in the woman was buried after the start of the construction 82 10 ABSOLUTE CHRONOLOGY OF THE BLED ISLAND CEMETERY 24 1a 34a 34b 1b B02 115a 116 36b 53 31 23 52 30 B01 115b 114 32 113 37* * 117 17 29 51 33 27 50 45 22 B03 111 112 12 21 123 35 121c 124 10 36a 121a * * 26 11 9 49 20 38 122 3 39b 39a 63 121b 16 8 15 48 39c 39d 47 7b 7a 44 41 19 42 62 6 14 28 64 43 40 58 61 4 18 60 5 59 13 55 25 54 56 57 2 70 125 92, 93b,c 83 120 69 84 85 86 93a 105 78 87 81b 10388,101 68c 72 96 81a 77b,c 66 82b 68b 82a82c,d,e 80 68a 89 102 77a 109d 77d,e 71 78 65a 94 107 79a 65b 106 109c 95b 95a 67 B05 76c 76a B04 109a 98b 76b 98c 109b 100a 98a 100b 90b 99 97 91 90c 90a 118 119 0 5 10 m Fig. 53: Bled Island, cemetery plan. Wal s of the present church are marked light grey. Scale 1:200. work but before the dedication of the church. Namely, A child’s grave (83) with two temple rings lies before the church was built the grave could not have somewhat distant from the rest but is contemporary been oriented parallel to it; but after the dedication of with the Early Medieval cemetery. The first temple the church only burials without grave goods would have ring (Type NO0100_0607; Fig. 51: 9) is dated to 830 to been allowed. This is thus the latest grave in the Early 950 CE, and the second (Type NO0100_0708; Fig. 51: Medieval cemetery and, according to the grave goods, 10) to 900 to 1010 CE. The intersection dates the grave the burial took place most likely in the first decade of to 900 to 950 CE. The grave is thus contemporary the 11th century. with the first two generations of the Early Medieval There are eighteen graves for which there is no cemetery. chronological data ( Fig. 53: graves 10, 11, 17, 20, 24, 26, 29, 32, 33, 44, 47, 48, 50, 51, 52, 53, 55, 58). 83 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND A B C Fig. 54a: Bled Island, A: Harris matrix of cemetery group G1; B: First generation; C: Second generation. 84 10 ABSOLUTE CHRONOLOGY OF THE BLED ISLAND CEMETERY A D E Fig. 54b: Bled Island, A: Harris matrix of cemetery group G1; D: Third generation; E: fourth generation. 85 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND 10.3 HIGH MEDIEVAL CEMETERY legal y exist. In Medieval practice, this meant that from that time onwards only individual burials could occur There were only three artefacts found in the graves either in the church or immediately next to it (the latter that belong to the High Medieval cemetery ( Fig. 18: G4), seems to have been legal y equal to the former). The end all of them in the backfill of the graves. of the 12th century is thus the terminus ante quem for The backfill of Grave 96 contained a temple ring the High Medieval cemetery. with a double thickening (Type NO0100_0808; Fig. Based on the stratigraphic analysis, the shortest 51: 11). Such temple rings are the most common type possible period of use can be determined using the of grave goods in the Early Medieval graves only a few same method as for the Early Medieval cemetery. In metres away. The artefact can be interpreted as a residual one of the central tombs ( Fig. 53: graves 81a, 82a, 82b, find, i.e., an artefact belonged to a nearby disturbed Early 102, 103, 104; Fig. 44) there is reliable evidence for five Medieval grave and was inadvertently incorporated in reburials, which means that at least six generations (the the backfill of this grave. The same applies to a small original burial plus five reburials) were interred. Under fragment − probably a fragment of the loop of a temple the same assumption that the duration of one generation ring with a forged loop − from the backfill of Grave 100. was 30 years and the time of burial of the first generation In Grave 77a a bronze ring was found, also as a 20 years, this cemetery was used for at least 170 years. residual find in the backfil . The ring has a rectangular This period corresponds directly to the time between cross-section and is decorated on the outside with a the post and ante dates mentioned above. Burial in the series of small dimples ( Fig. 51: 7). Such finger rings High Medieval cemetery thus began approximately in can only be roughly dated to the High Middle Ages. the second third of the 11th century and lasted approxi- For example, they do not occur in the Župna cerkev mately until the end of the 12th century. cemetery in Kranj, where there are no artefacts in graves Thus, the first generation was interred approxi- after the first decades of the 11th century. This indicates mately between 1030 and 1050, the second between that the first decade of the 11th century is a terminus post 1050 and 1080, the third between 1080 and 1110, the quem. On the other hand, the Late Medieval cemetery fourth between 1110 and 1140, the fifth between 1140 near the church of St Bartholomew in Šentjernej, for and 1170, and the sixth generation between 1170 and instance, does not contain any such rings either (Pre- 1200. However, it should be stressed that the accuracy dovnik, Dacar, Lavrinc 2008). Late Medieval period is of these dates is much lower than for the Early Medieval thus a terminus ante quem and the ring from the backfil cemetery and should consequently only be understood of Grave 77a is the only High Medieval artefact in the as a chronological framework. cemetery. Unlike the temple rings mentioned above, it cannot be interpreted as a residual find from a nearby High Medieval grave, as there are none there. On the 10.4 CHURCHES other hand, Grave 77a belongs to a series of five reburi- als in the same tomb. Consequently, this ring almost Dating the church buildings was one of the main certainly belonged to one of the earlier burials in the goals of the 1960s excavations on the Bled Island. Unfor- tomb, which was used at least since sometime after the tunetely, the correlation of church buildings with graves first decades of the 11th century. On this basis, these is the only available information for absolute dating. graves, which are very homogeneous in terms of burial A terminus post quem for the construction of the rites, can be dated very loosely to the High Middle Ages. first Pre-Romanesque church is provided by Grave 72, In the absence of other artefacts, the only option which is C14 dated to the time after 779 CE or with left is to date these graves on the basis of the wider con- 94.2% probability between 869 and 1013 CE (see Chap. text. These graves were located next to the church. In 4.2). More precise chronological information is provided the High Middle Ages (almost) all graves were located by Grave 4, which contained the so-called astragal tem- near a church. However, not all churches were allowed ple ring (Type NO0700_0810; Fig. 49: 6). As already to foster burial grounds. Höffler (2016a, 64) has convinc- mentioned, such temple rings were used between 990 ingly demonstrated that the church on Bled Island was a and 1030 CE, but the grave was dated in the first decade proprietary church, i.e., owned by a landlord. Between of the 11th century. If the above interpretations that this the end of the 10th and the end of the 12th century, burial in the Early Medieval cemetery was contempora- some owners were more successful than others in con- neous with the construction of the church are correct, verting such churches into benefices. The owners of the then the first pre-Romanesque church was built in the Bled Island church were not successful and the church first decade of the 11th century. was downgraded to the level of a filial church. One of The dating of the second Pre-Romanesque church the consequences of this downgrading was the loss of roughly to the end of the Early Middle Ages is supported burial rights. Thus, at the latest after the end of the 12th not only by the architecture but also by two fragments century, the cemetery on Bled Island could no longer of Early Medieval pottery discovered in the layer sepa- 86 10 ABSOLUTE CHRONOLOGY OF THE BLED ISLAND CEMETERY Fig. 55: Bled Island. Top: Harris matrix of cemetery group G4; bottom: all possible stratigraphic positions within G4 of the grave with mortar No. 93a (yellow). rating the first and the second church (see Chap. 4.3). still adhered to when the third church was built, i.e., the Further chronological clues result from the interpreta- church wall was not allowed to be built on top of these tion of the stratigraphy and dating of the High Medieval tombs. This interpretation is further confirmed and cemetery: The southernmost grave is in line with the clarified by mortar fragments in the backfill of several southern wall of the second pre-Romanesque church graves (77a, 82a, 84, 87, 88, 93a, 101). These fragments ( Fig. 44). The accuracy of the alignment leaves little could only have originated from the west wall of the possibility for coincidence and leads to the conclusion third church (see Chap. 9.3). Stratigraphical y, the earli- that the High Medieval cemetery was delimited while est grave with mortar fragments is Grave 93a. Since it the second church was still standing, that is, the earli- is in the middle of a stratigraphic chain of five graves, it est graves were contemporary with the second church. belongs either to the third or to the fourth generation Thus, the terminus post quem are the earliest burials in of the High Medieval cemetery ( Fig. 55). The terminus the High medieval cemetery, which occurred after ap-ante quem for this church is thus approximately the year proximately 1030. 1140, or more broadly, the church was already standing The third church, built in the Romanesque style, sometime between 1080 and 1140 CE. can also be correlated to the tombs of the High Medieval The archaeological data agrees well with the inter- cemetery. In particular, the eastern ends of the tombs, pretations of historians, who point to the deed of dona- i.e., the legs of the skeletons, almost touch the west wall tion in 1004 for the first church and the dedication of of the church building. This proves that these tombs were a church in 1142 for the third church (see Chap. 11.1). 87 11 INTERPRETATION OF THE BLED ISLAND IN THE MIDDLE AGES 11.1 WRITTEN SOURCES 18 farms, a fee, a mill in Mlino,27 and four plots in Zaka. These posessions came from the Bishopric of Brixen and Until now, it was assumed that the Church of the from various private donors from the nearby vil ages of Assumption of the Blessed Virgin Mary on Bled Island Mlino, Koritno, Bodešče, Bitnje, Zaka and Moste near was built between the second half of the 9th and the Žirovnica. All listed donations were made after 1140, third quarter of the 10th century. It was supposedly starting with the one from Bishop Hartmann of Brixen founded as a proprietary church of the local landlord, (Štih 2004, 22; Bizjak 2006, 59−60; Pleterski 2013b, župan, but came into the hands of the East Frankish 173, passim). Before the 1140s, the provostry − a legal lords before 1004 (Pleterski 2013b, 169; cf. Höfler 2016b, body comprised of the church and the estate − probably 212). According to Höfler, the church and the estate included only Bled Island with the church and a small passed into the possession of the Bishopric of Brixen non-arable plot of land called Vadiše, but its income soon after 1004. They allegedly rebuilt the church and probably also included donations from pilgrims. Indeed, appointed a provost. analysis of the ownership described in the sources sug- This retrograde interpretation is based on the fact gests that the church on the Bled Island was probably that the Bishopric of Brixen held the patronage and already a place of pilgrimage in 1004 (Gornik 1967, other rights to the island church in 1185 (Höfler 2015, 146; Gornik 1990, 173−176; Pleterski 2013b, 97−98; cf. 50; 2016b, 212). In the king‘s charter that bestowed the Štih 2004, 21−22). Judging by the example of the nearby Brixenes the Bled estate in 1004 − the earliest written chapel of the Assumption of Mary in Lesce (Bizjak 2012 document that mentions Bled − the church on the island with source references), pilgrims eagerly flocked to is not listed. However, the 1004 charter contains a sum- the church with offerings on the feast day of Mary on mary of the infrastructure elements of the Bled estate: 15 August from near and far. churches ( aecclesi s), buildings, mil s, unfree inhabitants, The sudden popularity of donations to the prov- various arable and nonarable plots, rights and sources ostry (and thus indirectly to the Bishopric of Brixen) of income. after 1140 can be understood as a consequence of the The same formula is found in a very similar charter Bishopric‘s active policy of trying to exploit people‘s from 1011. Such formulas ( pertinentio) were more or concern for their souls (Pleterski 2013b, 173) to establish less formalised in similar charters of the time (Štih 2004, the estate of the provostry and, in a broader context, to 22−23; Štih 2011, 14−19 with source references). Inter- further consolidate its domain of the Bled estate. estingly, according to Štih‘s analysis of the document of One way of achieving this goal may have been a 1004, such formulae take into account, at least to some church building programme, as can be identified in the extent, the natural, economic and social realities of the case of the above-mentioned Chapel of the Assump- property given away (Štih 2004, 22), whereas when he tion of St Mary in Lesce, only a few kilometres away. writes about the document of 1011, which was drawn A wooden chapel was built there in the middle of the up according to the same formula, it is to be understood 11th century at the latest. After a long time had passed as a purely formal element (Štih 2011, 23). ( post multorum vero cursum temporum), Dietmar, The church on Bled Island was thus first mentioned Archdeacon of Aquileia (Bishop of Trieste 1121−1145), in written sources only in 1185 as the Chapel of St Mary and Herwig, parish priest of Rodine, demanded that with a provost ( prepositi sanctę Marię in lacu Veldes) the owner tear down the wooden church and build a (Bizjak 2006, 59−60; Höfler 2015, 49−50; Gornik 1990, stone one instead, which could then be consecrated. 150−152; all with source references). With this docu- The owner, Brixen‘s ministerialis (local lord) Nantwin, ment, Bishop of Brixen Henry II (Bishop 1178−1196) did so. The church was consecrated with the assistance confirmed the estate of the provostry, which included 27 Mlino is a derivative of a Slovenian word for mil . 89 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND of the aforementioned persons ( presidio eorum), which disputes between Aquileia and Brixen about jurisdic- very probably meant a formal permission of Patriarch tion over the church. Ulrich I of Aquileia (Patriarch 1085−1121). The church, Höfler‘s interpretation of the situation in the 13th rebuilt in stone, was consecrated by Bishop Eberhard, century is that the patriarchs of Aquileia acquired the the Patriarch‘s vicar, as can be seen from a document church on the island with full rights between 1185 and issued between 1115 and 1121 (Bizjak 2012). 1247. Nevertheless, two centuries later, in 1459, the In this case, several elements can be understood to papal court in Rome confirmed all rights to the island indicate that this action was part of a broader scheme church to the Bishopric of Brixen. This means that at of replacing earlier wooden churches with stone ones: the end of the Middle Ages the Bishopric of Brixen the appeal was initiated by the regional (ecclesiastical) administered the church as their own proprietary administration, the construction was carried out by church with full jurisdiction, but the dispute ended the owner (who at the same time represented the local only with an agreement between the dioceses of Brixen administration) and the execution was supervised by the and Ljubljana (the latter as successor to the Aquileian regional (ecclesiastical) administration and confirmed patriarchs) in 1688 (Gornik 1990, 195−204; Höfler by the act of consecration. 2015, 50 with sources). The next document pertinent to the island of Bled A brief comment should be made on the patron concerns the consecration of a church in the area of saint of the Bled Island church. While the present Bled by Patriarch Pellegrinus I of Aquileia (1131−1161) church is dedicated to the Assumption of St Mary, it is in 1142 (Gornik 1990, 153; Štih 2004, 22 with source mentioned in the medieval sources only as the church references). Since the church was consecrated by the of St Mary ( sancte Marie in lacu Veldes). Based on a patriarch himself − unlike the ordinary church of Lesce, marginal note from the 16th or 17th century in the Ma- which was consecrated by a bishop, the patriarch‘s tricula of Radovljica, Höfler (1988, 224−225) assumed vicar − it must have been an important church. that the church was original y dedicated to the Nativity There are only two possibilities, either the parish of St Mary. This would be typical for the time around church of St Martin in Bled or the church on the Bled Is- 1004 (Höfler 2016b, 212), while the earliest Marian land. The building programme of the patriarch assumed churches were dedicated to the Assumption. However, above and the simultaneous efforts of the Bishopric of the orientation of the church convincingly shows that its Brixen to expand the possession of the provostry on original dedication was to the Assumption of St Mary, Bled Island do not completely rule out the parish church. while the note in the margin, dated three quarters of a However, these two arguments strongly suggest that millennium after the church was built and a century the stone church, consecrated on 11 December 1142 in before the birth of modern historiography, is likely to place of a wooden predecessor, is the church of St Mary remain an interesting curiosity. on Bled Island. The above interpretation of the written sources In this context, it should be noted that the construc- must be understood in the light of the fact that the tion of such a church could have taken between a few provost, and with him indirectly the church on the years and a few decades ( cf. Štular 2013, 138; Trueman island of Bled, was first mentioned only in 1185, which 2019). It is possible that the construction of the church was followed by two modest notes in the 13th century. on the island and the church in Lesce both started in It is important to recognise that before 1185 all infer- the second decade of the 12th century; however, the ences based on written sources were no more than completion of the former with a much more ambitious that: inferences. Moreover, all the interpretations of the architecture took two decades. The consecration by the written sources mentioned above, with the exception of patriarch could be seen primarily as a political event that of Gornik (1967), were influenced by Šribar‘s inter- linked both to the beginning of the ambitious career of pretations of the same archaeological data that are the Bishop Hartmann of Brixen (bishop 1141−1163) and to subject of my analysis. In other words, the existence of the completion of the church building. the church on the island before 1185 cannot be proven The next mention of the church on Bled Island is by written sources but only by archaeological analysis. in a tithe list of the Patriarchate of Aquileia from 1247, Otherwise, circular reasoning occurs: Interpretations where the church is listed together with monasteries of written sources that include earlier interpretations of and other provostries ( ecclesia Valdensis). It is also archaeological data influence the archaeological inter- mentioned as a provostry in a tithe list from 1296 pretation of the same archaeological data. ( prepositura Insule, prepositura de Veldis). These two Written sources, on the other hand, give a good references are better understood in the context of an overview of the legal status of the Church of the Assump- instal ation document from 1309, in which the Patri- tion on Bled Island from the late 12th century onwards. arch of Aquileia confirms the appointment of a chapel The church never rose above the level of a provost‘s rector, who was installed by the Bishop of Brixen. It is chapel. In the course of the regulation of proprietary clear from the document that there had previously been churches, which took place between the end of the 10th 90 11 INTERPRETATION OF THE BLED ISLAND IN THE MIDDLE AGES Fig. 56: Bled Island, view of the ongoing excavation north of the present church building (north-western corner, top left); the unused potential cemetery area mentioned in the text, top right (source: NMS archive OA film No. 8541). and the 12th century, it belonged to the proprietary 11.2 EARLY MEDIEVAL CEMETERY churches whose owners could not convert them into benefices. Thus, the Bled Island church on the island The Early Medieval cemetery without a church was of Bled dropped to the level of a filial church and it lost used in the 10th century and the first decade of the 11th the right of burial at the end of the 12th century at the century. It was located mostly in a small flat area in the latest (Höffler 2016a, 64). centre of the island. The cemetery consists of graves in A comment on the congruency of dating between clusters and graves in rows ( Fig. 53). archaeological and historiographical sources is in order. The westernmost row contains six graves (54, Archaeology dates the first church to the first decade 55, 59−62). Two stratigraphical y later graves of non- of the 11th century and the terminus ante quem for the adults partial y protrude from the orderly row. North third church is approximately 1140. This agrees very of this row is an unused area large enough for two or well with the interpretations of historians, who point three graves. As can be said from the documentation to the deed of gift from 1004 for the first church and ( Fig. 56), the unused space was suitable for burial, but to the consecration of a church in 1142 for the third was never used. church. Since these interpretations are based on com- The second row is the largest, containing the re- pletely different sources, the concept of consilience of mains of 19 graves (3−17, 25, 34, 56, 57) with at least 24 induction can be applied. This scientific principle is also deceased. The row is somewhat irregular, but the desire known as convergence of evidence and it states that to form a row while adapting to the terrain configuration the same conclusion is much stronger when it comes is clear. Typical of this row are reburials, which are more from independent and unrelated sources. Confidence numerous than in the adjacent rows. The position of the is strongest when evidence from different fields is con- graves in relation to each other clearly shows that the sidered because the methods and/or data are different diggers of the later graves were not aware of the exact (e.g., Štular et al. 2022, 7). The dating of the first church location of the earlier graves. This is direct archaeological to the first decade of the 11th century and the terminus evidence that the graves were not permanently marked. ante quem for the third church in the 5th decade of the Although only three graves (49−51) can be assigned 12th century are therefore very conclusive. with certainty to the third row, it probably contained 91 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND the remains of two others (44, 53). The row is bounded to the north by a rocky ridge, while to the south it was not ful y excavated due to later pavement. Taking into account the density of burials in the row and the loca- tion of Grave 44, which indicates how far the graves in this row extend, one can assume that at least one more grave remained unexcavated under the said pavement. The existence of the fourth and fifth rows identified by Šribar (1972, 390−391) is questionable. The graves around the fireplace were certainly not arranged in rows. The possible fourth row might have contained four graves (35−37, 45) with an unused space in the middle. Six graves (1, 20, 22, 29, 30, 38) may have been placed in the potential fifth row. However, in this part of the cemetery, the gaps alternate with areas of high burial density in a kind of chequerboard pattern, which does not occur in the first three rows. The main argument against the existence of the fourth and fifth rows is the overlap: the legs of the graves in the fourth row overlap with the graves in the fifth row. If there were actual y two rows, the entire fourth row would be later than the fifth, which is not likely. Instead of rows four and five, the graves in the east part of the cemetery seem to have been arranged in clusters, perhaps reflecting a kind of cemetery plots. At least 13 deceased were packed around the contem- porary fireplace in the southern cluster (Graves 18, 19, 28, 39, 40−43, 46, 63). The second cluster from the south consists of two graves (20, 38) and a small remnant of another earlier burial (20). There is a large difference (17°) between the orientations of the stratigraphical y earlier and later graves within the cluster. The third cluster contained at least five deceased (35−37, 45). In one case, the difference in orientation between an ear- Fig. 57: Bled Island, location of cemetery groups G1 (blue) lier and a later grave is again very striking (19°). In the and G2 (red) and reconstructed geomorphology of the island northernmost group, there are eight graves (21−23, 27, before Medieval building processes (by B. Štular). 29−31, 33) containing a triple reburial. North of the above clusters are two single graves (1, 24), each with a double reburial. The final part of the Early Medieval cemetery con- sisted of a cluster of 11 graves (111−117, 121−124) on immediately next to the green stone that at the time of a small terrace on the eastern slope, about 3 m below interment served as a spatial marker for (see Chap. 4.7). and 8.1 m east of the rest of the cemetery ( Fig. 57). They Similar isolated graves at the edge of the cemetery are were dug into a narrow crevice in the rock ( cf. Šribar known from two nearby cemeteries from the 9th and 1972, 390), which determined their orientation. The 10th centuries (Žale near Zasip − male Grave 55; Pleter- only artefact from these graves is a finger ring from the ski 2013b, 39, Fig. 2.3.4 − and Dlesc near Bodešče – a backfill of Grave 111, which can only be broadly dated cenotaph Grave 14; Pleterski 2013b, 51, Fig. 2.3.16) to the Early Middle Ages ( Fig. 51: 13). However, these and more prominently in the cemetery of Mali Grad in graves are stratigraphical y earlier than the first church Kamnik from the end of the 10th and beginning of the and are thus at least roughly contemporary to the burials 11th century (Grave 23; Štular 2007, 28, Fig. 4). described above. So, this appears to be just another clus- Similarly isolated is a child’s grave with a female ter in the Early Medieval cemetery, albeit isolated and headdress (Grave 83), which lies on one of the orienta- somewhat distant due to the restrictive geomorphology. tion axes of the cemetery. Although it is located within Even more isolated grave (72) further south was the High Medieval cemetery, it is clearly stratigraphical y probably the most important on the site. It was located below the High Medieval graves. Only the cranial base in the place with special significance for the mourners, was preserved and the head rings lay (presumably in 92 11 INTERPRETATION OF THE BLED ISLAND IN THE MIDDLE AGES Fig. 58: Bled Island, hotspot analysis (red − hot, yellow − neutral, blue − cold) of burial density of cemetery group G1 (percentage refers to confidence level; method according to Achino et al. 2019) confirms that the burial around the fireplace (green) was the most intense. situ) to its left and right. Comparable cemeteries often portant graves (72 and 83) are isolated from the rest. contained a separate area for children, for example, Four generations were buried in unmarked graves, in Mali Grad (Sagadin 2001; Štular 2007), Žale near Zasip total at least 71 deceased. The homogeneous gender (Pleterski 2014, 250−256 and Figure 3.3.6.17, as wel and age structure of the deceased indicates that a single as the literature cited there) and Dlesc near Bodešče community buried all (or most) of its deceased in this (Pleterski 2014, 259−261 and Figure 3.3.6.31, as well as cemetery. This community never consisted of more than the literature cited there). approximately 20 adults. However, Grave 83 is isolated and does not appear During the time of the burials of the second, third, to be the sole remnant of the children’s section of the and fourth generations, the fireplace in the central part cemetery. As such, it is very similar to the Grave 41 of of the cemetery became one of the focal points of the a child with female jewellery from the Dlesc cemetery burial rites. As it is evidenced by charcoal fragments near Bodešče, who was buried in the first generation in four graves (18, 22, 46, 63), the fireplace was lit at and was avoided by later graves (Pleterski 2014, 259). least four times during the interment. The thickness of Grave 83 from the Bled Island, too, is 3.7 m away from the burnt soil, ranging from 6 to 15 cm (Bitenc, Knific the next nearest contemporary grave and, judging by 2020a, 27), reflects the high temperatures caused by a the grave goods, could belong to the first generation. large fire. Another indication that this large fire was lit Consequently, the most convincing explanation is that during the interment is the fact that the fireplace was Grave 83 is an isolated grave of the Early Medieval avoided by the graves. cemetery, perhaps one of the graves of the first genera- Unfortunately, due to the smal size of the cemetery tion. This is supported by its location in the extension and the non-stratigraphic excavation method, there of the axis between the spring and the fireplace, which is no data on whether a fire was lit during each inter- dictated the orientation and location of most of the first ment. What can be demonstrated is that proximity to and second generation graves ( Fig. 50). the fireplace was considered prestigious, as evidenced The Early Medieval cemetery thus contains graves by a spatial analysis indicating the highest density of in rows and graves in clusters mostly located on the burials around the fireplace ( Fig. 58; for the method, largest, relatively flat area of the Bled Island. Two im- see Achino et al. 2019). 93 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Another argument underlining the importance of 11.3 HIGH MEDIEVAL GUARDIANS the proximity to the fireplace is the secondary burial of cranial and long bones (Grave 39). The position of the The stratigraphic phase that followed the Early bones proves that the disarticulated bones were placed in Medieval cemetery was marked by the construction of a rectangular recipient made of organic material, prob- the first wooden church in the pre-Romanesque style, ably a wooden box. The secondary burial is reminiscent dedicated to St Mary of the Assumption. The nave, of High Medieval ossuaries, which is highly unusual in with an area of 35.9 m2, was just large enough to ac- Early Medieval cemeteries. commodate the members of a small congregation the It is also similar to reliquaries, for example, the one size of that which used the earlier cemetery. This could made of lime wood in the 5th century and found in a 6th be a coincidence, however, because it is not known century church at Hemmaberg in present-day Austria which community used the cemetery or who built the (Binder, Ladstätter 2019). However, the Late Antiquity church for whom. However, we do know that many reliquaries in the Alpine region are usual y only half the details were taken into account when this church was size of the recepticle from Bled Island, which was about built. The church was oriented precisely in the direction 40 by 80 centimetres. of sunrise on the calendar day of the church’s patron The latest burial in this cemetery was that of a saint, that is, on 15 August, when the Assumption of woman buried in the first decade of the 11th century St Mary is venerated. The orientation possibly served (Grave 4). The orientation of the grave is parallel to the a purpose: during the morning service on that day, the church building, which must have been already in place rising sun il uminated the altar, which certainly made a at the time of this burial, or at least its ground plan was strong impression on churchgoers in a time of modest already marked. This was a biritual burial, combining artificial lighting. a pre-Christian element (grave goods) and a Christian The building was planned according to the of Pre- element (orientation paralel to the church building). Romanesque architecture guidelines, which stipulated that the interior should reflect the ratio 1 : √2. This Thus, the Early Medieval cemetery on Bled Island intent was abandoned because of the difficult terrain. was used in the 10th century and the first decade of the The exact position of the church was chosen so, that 11th century. During this time the burial rite changed the holiest part of the church, the apsidal area with three times. the altar, was built over the most important grave of First, the initial generation of the deceased was the preceding cemetery (Grave 72) and an even earlier buried in the first two decades of the 10th century spatial marker. From today’s perspective, this may seem perpendicular to an orientation axis connecting the unusual or even destructive, but in the Middle Ages it fireplace and the spring. Such burial rite is known in was a common practise. In Ireland, for example, several several cemeteries from the period between the 9th and Medieval churches were built over the central graves of the end of the 10th century. earlier cemeteries because the bones of an important In the second generation, approximately between person were believed to offer protection to the church 920 and 950, efforts were made to align the graves in (Fry 1999, 67−68). Thus, the seemingly modest church rows or arrange them in clusters. Some graves were care- of St Mary on the Bled Island had a powerful symbolic ful y oriented towards cardinal east, which demonstrates function. It combined the demonstration of the power a high degree of astronomical knowledge. The meaning of the new religion (the appearance of being able to of this orientation could not be determined, but the control sunlight) with the reliance on pre-Christian fact that it was the result of a conscious and deliber- power (protection by the bones of an important person). ate decision was proven beyond doubt. The cemetery Another element of continuity in the construction of the area expanded to cover all available space with deep first church is the above mentioned grave (4), which was enough soil. dug parallel to the church. Third, the last deceased was buried in this cem- Soon, perhaps after only a few decades, the small etery in the first decade of the 11th century. It was church was rebuilt. A somewhat larger building with an aligned parallel to the church building according to the area of 52.3 m2 was erected. The interior was painted Christian canon and contained jewellery according to with frescoes. The building was symmetrical, with the pre-Christian rites. sides approaching a ratio of √3 : 2, typical of the pre- So, the cemetery is a mixed ritual cemetery where Romanesque style. On the west side, the nave was later three different burial rites were used. The changes follow extended by a narrow room in the form of an irregular each other chronological y and do not seem to overlap trapezoid. in time. To our knowledge, this is the first time such Two graves were excavated right next to the two diversity has been documented in what is essential y a pre-Romanesque churches: Grave 71 next to the first small and short-lived Early Medieval cemetery. church and Grave 67 next to the second. Analysis of ground plans and field diaries shows that Grave 71 is 94 11 INTERPRETATION OF THE BLED ISLAND IN THE MIDDLE AGES the earlier of the two. At some point, it was exhumed Burials of adults in prominent places outside and filled with soil. Exhumation can perhaps be un- church entrances are mentioned by two Irish medieval derstood in the context of the phenomenon known written sources, the first on the life of St Senán of Scattery as the removal of a saint’s body for relics ( cf. Klevnas and the second on the burial of the king’s son Magnus et al. 2021; Kjel ström 2022). The only evidence that it in 1244 (Fry 1999, 170−171 with source references). original y contained a body are tiny fragments of human The latter contains three elements of interest. First, the bones. After the Grave 71 was emptied and backfilled, location under the threshold of the church was explicitly the pit for Grave 67 was dug about half a metre further mentioned, which confirms its significance. Second, south, disturbing the northern part of Grave 71. These the church had been desecrated at the time of burial, two graves are the only ones in whose backfill sand which could be the reason why the grave was not placed was found. The deceased in Grave 67 was the only one inside the church. Third, the deceased died during an in the entire cemetery who was buried with his hands attack on the church and thus had a direct connection folded on his chest in prayer. The location of the two to it. Similar burials are also known from Merovingian graves is particularly revealing: the earlier one (71) was Europe (Effros 2003, 212) and Early Medieval England immediately in front of the earlier church and the later (Kjølbye-Biddle 1975, 101).28 one immediately in front of the later church, while both Based on the analysis of the sources, such buri- were most likely under the threshold of the respective als can be understood as church burials of special church entrances ( Fig. 22; also Šribar 1972, 391). Both significance, which were believed to ward off enemies were covered by a mortar pavement that contained nu- or provide some protection for the church (Fry 1999, merous pottery fragments and animal bones in its lower 170−171). part. This means that the pavement was laid directly on On the basis of these analogies, the graves under the remains of an activity related to the manipulation of the thresholds of the Bled Island first three churches can foodstuffs. These activities took place after the backfil - be understood as burials with special meaning. The two ing of Grave 67 and before the laying of the pavement. earlier graves had a position as prestigious as the burials Given that the pavement was intended to cover the grave, inside the church would have had. Thus, they might have it can be assumed that the backfilling of the grave, the been substitutes for the more common burials inside the manipulation of the foodstuffs, and the laying of the church. Based on the Irish analogy one could assume pavement all took place in a relatively short period of that the church on the Bled Island was not (properly?) time, i.e. they were all part of the interment. This then is consecrated, just as the wooden church of St Mary of the direct evidence of a Christian burial rite involving ma- Assumption in Lesce was not (properly?) consecrated nipulation of foodstuffs. There is evidence of such burial until the middle of the 12th century. rites in Medieval written sources (Fry 1999, 93−94), and the term funeral feast is sometimes used to describe it The graves were thus clearly marked, but were (Makarovič 1995, 152). nevertheless disturbed several times. Each time the The same pattern was repeated in Grave 118 from entrance to the church was moved, a new grave with the the next stratigraphic phase (Phase 3), which was same characteristics was created, the third more than contemporaneous with the third church built in the a century after the first. This testifies to a continuing Romanesque architectural style. Grave 118 was carved awareness of the special meaning of these graves; an into the bedrock under the threshold of the Roman- enduring tradition rather than a one-off action following esque church, had the traces of several reburials, and an extraordinary event such as the death of the king’s was overlaid with a mortar pavement ( cf. Šribar 1966, son Magnus in the Irish example. Here, the location 173−174, 228). Šribar (1966, 231) concluded that this was more important than the deceased. The interpreta- grave had to be connected with the functionality of the tion that churchgoers believed that these burials would south-western entrance of the church, i.e., the grave was provide additional protection to the church seems very under the church’s threshold. likely. So, it can be said that the grave under the threshold These three graves (67, 71, 118) thus share five was indeed believed to be the guardian of the church. characteristics. First, they were not part of the contem- porary cemetery, but were directly connected to the contemporary church. Second, they were positioned 11.4 HIGH MEDIEVAL under the church’s threshold. Third, they had elements CHURCH CEMETERY of ritual acts that occurred during interment (pottery and animal bones). Fourth, their location was marked The third church was built from stone between with mortar pavement (which was the only paved sur- 1080 and 1140 in place of the wooden church. It was built face outside the church at that time). Fifth, there were in the Romanesque architectural style in stratigraphic multiple manipulations of the remains of the body 28 (exhumation of the body). The author would like to thank J. Rihter for the infor- mation. 95 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Phase 3. Stratigraphical y contemporaneous with the to the cowl shall be brought over this and at- thirdchurch were the graves located directly adjacent to tached in three places with thread. Gaiters of the west wall of the church. I understand this compact the same material reaching to the knees shall group of 24 graves ( Fig. 18: G4) with at least 42 deceased be put on the legs and night shoes on feet. The to be a High Medieval church cemetery. The position to hands shall be sewn together from arm to arm, the west of the church was unusual, as most Medieval and round the legs likewise. So dressed, the cemeteries extended around the church (e.g., Predovnik, corpse shall be set in the hearse and covered Dacar, Lavrinc 2008). The most prestigious locations with a pall.29 were to the east of the church (near the altar) or near the entrance, which in this case is to the south. The unusual However, Bled Island was not a monastery and location of the Bled Island cemetery cannot be explained there are also women and children buried there. This by a challenging geomorphology, because the graves is thus a cemetery of lay people, and it is unlikely that were carved into the bedrock, which would be possible their bodies were prepared for funeral in the same way anywhere. The cemetery was most likely restricted to a as those of the monks. Thus, the deceased from the small area west of the church due to other factors. One Bled Island were most likely wrapped in a shroud. The possible explanation is that the graves were located in uniformity of the bodies over almost two centuries dem- the antechamber of the portico. Thus, in the eyes of onstrates that this was a controlled action in which strict the church regulations, this would not be considered rules were observed, including the manipulation of arm a church cemetery, but burials in the church reserved posture. All this testifies that these were the tombs of a for the affluent. smal , orthodox, and conservative community. The structure of the cemetery provides us with fur- Forty-eight individuals in at least six generations ther information. The graves (79−89, 92−96, 101−107, means that on average eight deceased were buried per B04, B05) are characterised by deep pits, partly carved generation. This corresponds to the space available in into the bedrock. There were six contemporary grave four single and two double tombs. The number corre- plots, two of which were for double graves. sponds to a small social group in which men, women and Each grave was used for several reburials so that children were equal y represented, perhaps an extended there are at least 48 individuals buried in the cemetery. family. Compared to the Early Medieval cemetery, this is With each new interment, most of the bones were re- half the number of people per generation, a clear indica- moved or destroyed (the data are too scanty to distin- tion that this was a different community. guish between the two cases). Only in two tombs (82, 93) Due to its continuous use for almost two centuries, were the skul s of earlier burials careful y placed at the the equal number of deceased in each generation, the edge. The burials often disturbed neighbouring graves balanced gender and age structure, conservative and (e.g., 102 and 103 or 90, 107 and 94). orthodox burial rites, and the exceptional location near Taken together, the deeply cut pits, the reburials, or even inside the island church, the entire cemetery can and the secondary placement of the skul s are charac- be understood as the burial place of a lay but pious social teristic of tombs. group of the highest social status in the local community. In addition to the six tombs, there is an additional The local community at that time was, of course, the Bled grave on the southern edge of the cemetery ( Fig. 44). estate, which belonged to the Bishopric of Brixen. The The complete absence of finds in the graves (with high social status of the cemetery community on Bled the exception of a ring) on the one hand, and the uni- Island thus presupposes a very close connection with form position of the hands in the lap on the other, are the Bishopric of Brixen, which is confirmed by the piety indications that the bodies were probably naked and expressed in the burials. The most obvious choice is the wrapped in a shroud. There are no pins, which means family of a ministerialis, the estate manager. that the shrouds were either sewn or tied together ( cf. The above archaeological evidence testifies that an Daniell 1997, 38−39, 45, 109, and 155). interment in this cemetery can be imagined very much There is another way of preparing the body of like an English depiction from the late 1400s ( Fig. 59; the deceased for a funeral that would leave the same cf. Hamilton 1980): the tomb, which has not been used archaeological traces: dressing the body of a monk, as for at least 30 years, has just been emptied; bones of described by Lanfranc (+ 1089) in his monastic guide- earlier deceased lie scattered around, but do not disturb lines (Fry 1999, 125-126): the small pious congregation led by a clergyman; lay- men lover a body in a tomb; the body is wrapped in a The chamberlain should be present with grave- funeral shroud, and a priest reads prayers over it from clothes of the right kind, thread, and a needle for a juxtaposed missal. sewing... When it is (the corpse) washed, shall be clad in a new shirt, or one newly washed, and a cowl; a head-cloth of linsey-wolsey belonging 29 Translation Fry 1999, 125. 96 11 INTERPRETATION OF THE BLED ISLAND IN THE MIDDLE AGES Fig. 59: Leicester (UK), Wygston house, 18 Highcross Street: detail of painted window panel glass, 1490−1500: Burial scene, long bones, probably human, scattered below beside a typical medieval wooden spade. The inclusion of the latter in the depiction is a clear indication of the indifference to human bones in a medieval cemetery (source: https://www.storyofleicester.info/a-place-to-live/ wygston-s-house/; the window panel is kept in the Newarke Houses Museum, Leicester). In addition to this clearly delimited cemetery there Nul us mortuus infra ecclesiam sepeliatur, nisi were seven other stratigraphical y contemporaneous episcopi, aut abbates, aut digni presbiteri, vel graves whose position was directly related to the church. fideles laici.30 Some of the graves are located in the southeast part of the church, while others are outside, just south of the apse. This rule excluded practical y no one but the poor The grave with the most direct connection to the and it realistically reflected practices in the Middle church was Grave 120. It was located east of the altar in Ages, where the rich and powerful competed for the the central axis of the apse. To understand the excep- best burial places. tionality of this burial, one must be familiar with the Regarding where exactly, burials in churches were rules about who could be buried in a church and where. restricted to the area in front of the portico (a vestibule, Mediaeval Christian interment rituals and ceremo- lat. porticus; cf. Sorries 2003, 35) and in vicinity of, but nies had evolved from various pre-existing traditions. by no means under the altar (Fry 1999, 169−170; Hart- Thus, Christian interment was not uniform, as different mann 2003, 137−138; Schlokmann 2003, 204; Krznar authorities interpreted God’s will differently. The 12th 2012, 26). This tradition is exemplified by the order of century Camaldolese monk Gratian was one of the Theodulf of Orleans (+ 818): first to summarise laws on burial in his Decretum (e.g., Landingham 1993). He turned to a statute of the Council of Mainz of 813 to answer the question of who can be 30 buried in a church: Decretum 13 q. 2 c. 18 (Translation Landingham 1993, note 25): No one shall be buried inside the church, except for bishops, abbots, worthy priests, or faithful laymen. 97 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Prohibendum etiam secundum maiorum in- relationship between the two is and will likely remain stituta, ut in ecclesia nul atenus sepeliantur, unknown (unless new, currently unknown methods sed in atrio aut in porticu aut exedra ecclesiae. of dating emerge). Regardless, these two graves were Infra ecclesiam vero autprope altare, ubi corpus a strong element of continuity between the pre-Ro- domini et sanguis conficitur, nul atenus habeat manesque first and second churches (Phase 2) and the licentiam sepeliendi.31 Romanesque third church (Phase 3). “The guardian” (Grave 118) reflects the old tradition, and “the saint” Theoduf’s decree was still observed at the begin- (Grave 120) the beginning of a new tradition. ning of the 11th century, when it was included in the work of Burchard of Worms (Hartmann 2003, 134) and There were six more graves (65, 66, 68, B05, B04) in it was not overlooked by Gratian in the 12th century. the church. Most were badly damaged by the subsequent In the Late Middle Ages and even more so in the post- construction works. The exceptions were Graves 65 and Medieval period, the rules became more relaxed, but 66 positioned near the altar, whose pits were carved the immediate vicinity of the altar was stil the most into the bedrock. Grave 109 was positioned outside the esteemed position (Rodwell 1989, 153−160; Scholk- church, just south of the apse. The body was laid at the mann 2002, 212−217). bottom of the grave pit cut in the bedrock and the back- To be specific, the most prestigious position among fill contained numerous bones from disturbed graves. the already prestigious burials inside of the church was At least one other unexcavated grave lay to the south to the east of the altar (Hartmann 2003, 134; Pedersen of it and was badly damaged during the subsequent 2003, 172; Scholkmann 2003, 206−210; Wemhoff 2003, construction works. 102−103; Kenzler 2015, 150). The exceptionality of the There are no archaeological data to unequivocal y position is highlighted by Bede the Venerable in the date these six graves. The only element that can help in parable about an Irish saint Fursey. dating is that they are stratigraphical y contemporane- For the first 27 days after his death, Fursey‘s body ous with the third chur. The only thing that can be said lay in the portico. When the people went to bury him about these graves is that they are High Medieval burials near the altar, they discovered that the body had not yet in prestigious locations within or immediately adjacent begun to decay, and after four years they moved him, his to the church. Judging by the small number of graves body still in pristine condition, to a newly built church and their positions, they are not a separate cemetery or to bury him to the east of the altar (Miller 1999, 99). a separate part of a cemetery, but individual graves. As The parable shows that Fursey, a venerable missionary, mentioned above, the graves inside the church are likely was to be buried near the altar. However, when his body those of important ecclesiastical figures, while the graves did not decompose, indicating that he was a saint, a new outside could belong to high-ranking lay people who church was built. St Fursey had to be buried immedi- earned or paid for a position of honour ( cf. Fry 1999, ately east of the altar. In other words, any burial near 170). For example, some of the late 12th century donors the altar is exceptional, but a burial east of the altar and to the provostry of Bled Island could be among them. sometimes referred to as the founder grave (e.g., Halsal All graves of this phase are directly related to the 2009, 211−212 and 372) is unique. church building and thus to the Church. They can be Consequently, one should expect a cleric of excep- described as church graves. tional y high status in Grave 120. This is further empha- sized by the fact, that the above described tombs in the High Medieval cemetery were in the antithetical position 11.5 LATE MEDIEVAL AND to Grave 120. This was no doubt a deliberate connection POST-MEDIEVAL CHURCH GRAVES symbolising the shepard (a priest or a bishop) and his flock (lay population). But it was a child who was buried The Late Medieval graves are located in the south- there (Bitenc, Knific 2020b, 324−325). This makes the east part of the church or outside, directly south of the grave unique, not only on Bled Island but also in the apse. These graves are stratigraphical y contemporane- wider context of the European Middle Ages. ous with the Late Medieval church, which was built in There are thus two graves that have a strong and the Gothic style (Phase 4). direct connection to the third church: „the guardian“ in Five graves (90, 91, 98−100) with the remains of at Grave 118 and “the saint” in Grave 120. The chronological least 11 individuals were located next to the south wall of the church. Furthermore, a single grave (119) was 31 Capit. I.11. (Translation B.Š.): It is forbidden according located on the outside of the south wall of the portico. to the institutions of the ancients, that bodies could be buried Its unusual north-south orientation is perpendicular in the church, only in the churchyard, or portico, or in the to the church wall and suggests that it was marked with apse of the church. No one may be buried inside the church or close to the altar, where the body and blood of the Lord a gravestone set into the church wal . All six of these are prepared. graves with 12 individuals can also be described as 98 11 INTERPRETATION OF THE BLED ISLAND IN THE MIDDLE AGES church graves, possibly belonging to lay people of high which was worshipped on the island, and with the social status. central place of the world of the dead in the Early Me- Two graves (69 and 70) lie under the western dieval landscape of the area around Bled, the Višelnica gallery in the northwest part of the nave. They are of bonfire site. Archaeological evidence of this symbolism Post-Medieval and possibly Modern era date, as they is embodied in what is probably the most important are stratigraphical y later than the last reconstruction of Early Medieval grave (72) on the Bled Island. The de- the stone floor in the nave. They are also church graves. ceased was buried with his legs right next to the green stone, and the grave was oriented with its side facing the Višelnica bonfire site. Other archaeological evidence 11.6 CONCLUSION of this symbolism is the careful y selected locations of three first-generation burials and the fireplace where Based on the archaeological evidence it can be the fire burned during (some of) the burials. The concluded that the object of a pre-Christian worship second generation reorganised the cemetery. They on the Bled Island of supralocal importance was the arranged the graves in rows. Some faced the cardinal spring. Judging from the analogies in the written east, suggesting considerable astronomical knowledge. sources, it is likely that the spring was positioned in a The last interment in this cemetery occurred in the first grove or at least in the shade of a tree. The vegetation decade of the 11th century. The deceased woman was on the island and the appearance of the island seen placed in the grave with jewellery, but the orientation from the mainland were thus probably similar to today, of the grave already corresponded to the orientation only without the church and stairs ( Fig. 10). Ritual y of the first Christian church. Thus, this was a biritual connected to the spring was a nearby place of special burial, in which both the pre-Christian (jewellery) and meaning, which in the 10th century was marked with a the Christian (orientation) rites were followed. It was green stone. There is no direct archaeological evidence the only biritual burial on the Bled Island and at the on when this place gained its special meaning. The only same time the last deceased from this community to direct chronological clue is a pit under the stone, carved be buried on the islet. into the bedrock and surrounded by prehistoric pottery In the first decade of the 11th century, probably and an axe with a polished surface. Consequently, it is shortly after the Bishopric of Brixen formal y received fairly safe to assume that this part of Bled Island held a the Bled estate in 1004, a small wooden church was built special meaning long before the 10th century, perhaps on the island. It was dedicated to the Assumption of St since prehistory. Mary, and it was careful y placed so that its holiest part, Indirect evidence of pre-Christian worship on the apse, stood directly on the earlier place of worship Bled Island predating the cemetery may be the recently marked with the green stone. This was by no means discovered of a watercraft, specifically a logboat or accidental. Rather, it was entirely in keeping with the monoxylon. It was discovered sunk in the lake on the customary procedure of the time to Christianise pre- shore of the island and C14 dated approximately to Christian places of venerated springs. A burial under the second half of the 8th century and the beginning the threshold, the guardian, offered symbolic protection of the 9th century CE. The current working hypothesis to the church. of the researchers working on its recovery is that the Despite its apparent modesty, the construction of logboat was submerged on the shore of the isle for safe- the church required some cutting-edge architectural keeping, but was never recovered (Gaspari, Humerca, and astronomical knowledge at the time. The alignment Žvan 2021; ZVKD 2022). This finding is therefore a with the sunrise on the day of the Assumption is mate- direct evidence of activity on the Bled Island, since rial evidence that vividly demonstrates the difference someone was obviously visiting the isle. However, it between the pre-Christian and Christian models of does not provide direct evidence of the nature of the placing religious objects in space. The pre-Christian activity. In the context of the archaeological evidence model is multicentric and strives to accommodate as presented in this volume − which testifies only to ritual many direct connections to the divine as possible. In activity, whereas the settlement activity is highly un- this case, this was expressed through the connection likely given the restrictive geomorphology of the isle of the cemetery to the venerated spring, the venerated as reconstructed − the most likely inference is that green rock, and the wider mortuary landscape. In the pre-Christian worship on the Bled Island was already Christian model, the connection with God is monopo- taking place in the second half of the 8th century and lised and centralised in the church. The church is not the beginning of the 9th century. connected to a place in the landscape, but symbolical y In the early 10th century, a small community directly to the Divine. began to bury their dead in the flat area between the On the Bled Island, the symbolic connection to the spring and the green stone. The original design of the Divine was established by the astronomical orientation cemetery symbolical y connected it with the spring, of the church according to the sunrise on 15 August. 99 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND However, this direct connection existed only for those who were attended the service that was held in the church building each year on that day. And access was controlled by the Church. In particular, it seems that only pilgrims with offerings were allowed entry. In this way, Christianity legitimised the concentration of power ( Fig. 60; cf. Fabech 1999, 469−470). After only a few decades, the original wooden church was demolished and replaced by a slightly larger and more decorated building made (partly) of wood. This church was also protected by the burial of the guardian. To the west of the church, in the shadow of sunrise on the day of the Assumption, a small com- munity of mixed gender and age buried their dead. The community was orthodox and conservative, which was reflected in the strict observance of the same rules for Christian burials for almost two centuries. This and the exceptional location of the islet mean that this community belonged to the circle of those who were involved in the administrative and defensive apparatus of the Bishopric of Brixen. After almost a hundred years, perhaps in the second decade of the 12th century, the wooden church was demolished and the construction of a new stone church began. It was a relatively ambitious Romanesque architecture, whose area was more than three times that of its predecessor and which changed the image of Fig. 60: Illustration of the differences in centrality between Bled Island forever. The islet with a small church near the pre-Christian (top) and the Christian (bottom) landscape. a spring in a grove was transformed into the church on In the pre-Christian cosmos, the farm was a nodal point that an islet. Immediately east of the altar, the holiest place provided direct access to the gods, and no place was ranked above the others. In the Christian cosmos, the communication in the Christian world (usual y reserved for saints or at with God was only possible in church. Christianity legitimised least bishops), a non-adult person was buried in the new the concentration of power mostly in the churches and thus church. The continuity with the old church is reflected the Church representatives controlled access to God (by B. in the burial of the third guardian and in the continued Štular; source: Fabech 1999, Fig. 12). burials to the west of the church. 100 12 SOURCES AND BIBLIOGRAPHY Sources Herbord = Herbordi Dialogus de vita Ottonis episcopi Babenbergensis Ann. Car. Sax. = Annales aevi Caroli et Saxonici Köpke, R. (ur.) 1868, Herbordi Dialogus de vita Pertz, G. H. (ur.) 1841, Annales, chronica et historiae Ottonis episcopi Babenbergensis, MGH Scriptores aevi Carolini et Saxonici, Monumenta Germaniae rerum Germanicarum in usum scholarum separa- Historica 4, Hannover. tim editi 33, Hannover. Bianchi, G. 1845, Documenti per la storia del Friuli dal Orig. = Isidori Hispalensis Episcopi: Etymologiarum sive 1326 al 1332, Udine. Originum Lindsay, W. M. 1911, Isidori Hispalensis Episcopi Capit. = Capitula episcoporum Etymologiarum sive Originum libri XX, Oxford. Brommer, P. (ur.) 1984, Monumenta Germaniae Historica, Capitula episcoporum 1. Hannover. Paen. Theo. = Peon. pseudo-Theodore Liber poenitentialis Theodori archiepiscopi Cantuariensis ecclesiae Correctione = Martini Baracinensis De correctione Thorpe, B. (ur.) 1840, Ancient Laws and Institutes rusticorum o f England. Monumenia ecclesiastica. London. Barlow, C.W. (ur.) 1950, Martin of Braga. De cor- rectione rusticorum. In: Martini Baracinensis opera Thietmar = Thietmari Merseburgensis episcopi Chronicon omnia. New Haven, 183−203. Wattenbach, W., F. Kurze (ur.) 1889, Thietmari Merseburgensis episcopi Chronicon, MGH Scrip- Decretum = Decretum of Gratian or Concordia discor- tores rerum Germanicarum in usum scholarum dantium canonum separatim editi 54. Hanover. Decretum magistri Gratiani. Editio Lipsiensis secunda, post Aemilii Ludovici Richteri curas, ad librorum manu scriptorum et editionis Romanae Bibliography fidem recognovit et adnotatione critica instruxit Aemilius Friedberg. Leipzig, B. Tauchnitz, 1879. ACHINO, K. F., ŠTULAR, B., RIHTER, J., RIHTER, J. (Corpus iuris canonici 1). 2019, Assessing the intentionality of spatial organi- zation. Cemetery of Župna Cerkev (Kranj, Slove- Durand, Rationale = Durando Episcopo Mimatensi, nia) case study. − Arheološki vestnik 70, 297−313. Rationale Divinorum Officiorum ALI, J. R., CUNICH, P. 2001, The Orientation of Beletho, J. (ur.) 1859, Rationale Divinorum Offi- Churches: Some New Evidence. − The Antiquaries ciorum a rev domino Gulielmo Durando Episcopo Journal 81, 155−193. Mimatensi J.U.D. Clarissimo, Neapelj. ARNOLD, B. 2002, A Landscape of Ancestors: The Space and Place of Death in Iron Age West-Central Helmoldi = Helmoldi Presbyteri Chronica Slavorum Europe. − In: H. Silverman, D. B. Small (eds.), The Stoob, H. 1963, Helmold: Chronica Slavorum. Space and Place of Death, 130−143. Neu übertragen und erläutert von Heinz Stoob. In: AVGUŠTIN, C. 1971, Zeleni kamen v gorenjski arhitek- Ausgewählte Quellen zur deutschen Geschichte des turi. − Slovenski etnograf 23/24, 39−52. Mittelalters. Wiss, Darmstadt. AZINOVIĆ BEBEK, A., JANEŠ, A. 2016, Groblje oko crkve sv. Nikole biskupa u Žumberku. – In: Krznar et al., 123−140. 101 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND BARKER, P. 1977, The Techniques of Archaeological CONOLLY, J., LAKE, M. 2006, Geographical Information Excavation. − London. systems in Archaeology. − Cambridge. BERCE, R., KASTELIC, J., ŠRIBAR V., ŠTULAR, B., ČAVAL, S. 2010, Astronomska usmeritev romanskih 2012, Terenski dnevniki izkopavanj leta 1953. – cerkva v Sloveniji. − Unpublished PhD thesis, Uni- In: B. Štular, M. Belak (eds.) 2012, Grobišče Župna verza v Ljubljani, Filozofska fakulteta, Oddelek za cerkev v Kranju. Dokumentacija o izkopavanjih v arheologijo, Ljubljana. letu 1953. − E-Monographiae Instituti Archaeo- ČAVAL, S. 2014, Church Orientations in Slovenia. − logici Sloveniae 1, 17−186. Handbook of Archaeoastronomy and Ethnoas- BERK, S., BOLDIN, D. 2017, Slovenski referenčni koor- tronomy, 1719−1726. dinatni Sistemi v okolju GiS. − Geodetski vestnik ČIHÁKOVÁ, J. 2018, The Lesser Town of Prague in 61 (1), 91−101. the Tenth and Eleveth Centuries. In: – P. Kouřil, BIENVENU, F., LEGENDRE, S. 2015, A New Approach R. Procházka et al. (eds.), Moravian and Silesian to the Generation Time in Matrix Population Mod- Strongholds of the Tenth and Eleventh Centuries els. − The American Naturalist 185 (6), 834–843. in the Context of Central Europé, Brno, 283−303. BINDER, M., LADSTÄTTER, S. (eds.) 2019, Die Heilige ČIMIN, R. 2016, Osobitosti kasnosrednjovjekovnoga i vom Hemmaberg. Cold Case einer Reliquie. − novovjekovnoga groblja uz župnu crkvu sv. Martina Vienna. u Virju. – In: Krznar et al., 107−122. BITENC, P., KNIFIC, T. 2020a, Srednjeveško grobišče DANIELL, C. 1997, Death and burial in medieval Eng- na Blejskem otok / Medieval burial ground on Bled land, 1066−1550. − London, New York. Island. − In: Štular 2020, 11−44. DEMACOPOULOS, G. 2008, Gregory the Great and the BITENC, P., KNIFIC, T. 2020b, Katalog grobov in najdb / Pagan Shrines of Kent. − Journal of Late Antiquity Catalogue of graves and finds − In: Štular 2020, 1 (2), 353–369. 287−342. DINCAUZE, D. 2000, Environmental Archaeology. BIZJAK, M. 2006, Urbarji briksenske škofije 1253−1464 Principles and Practice. − Cambridge. (Die urbare des hochstifts brixen 1253−1464). − DOWDEN, K. 2000, European Paganism: The Realities Srednjeveški urbarji za Slovenijo 5. of Cult from Antiquity to the Middle Ages. − Oxon, BIZJAK, M. 2012, Listina o izgradnji in posvetitvi New York. lastniške cerkve v Lescah. – In: R. N. Klemenčič, S. EFFROS, B. 2003, Merovingian Mortuary Archaeology Štefanac (eds.), Historia artis magistra: amicorum and the Making of the Early Middle Ages. − Oak- discipulorumque munuscula Johanni Höfler septua- land (CA). genario dicata, Ljubljana, 33−42. EICHERT, S. 2012, Frühmittelalterliche Strukturen im BOŽIČ, D. 1995, O zakladu poznorimskih novcev v Ost alpenraum. Studien zu Geschichte und Archäolo- Slatni pod Dobrčo in o zgodnjekrščanskih mo- gie Karantaniens. – Aus Forschung und Kunst 39. zaikih v Lescah (A Note on a Hoard Find of Late ESTEVES da SILVA, J. C. G., DURÃES, N., CORTEZ, Roman Coins at Slatna under Dobrča and on Early D., BOBOS, I., ALGARRA, M., RODRÍGUEZ- Christian Mosaics at Lesce). − In: J. Sinobad et al. BORGES, J. E. 2009, Factores ambientais para a (eds.), Radovljiški zbornik 1995, 38−57. formação de adipocera em cadáveres . – Geologos BRONK RAMSEY, C., LEE S.2013, Recent and Planned 8, 25−31. Developments of the Program OxCal. – Radiocar-EVANS, H. M. A. 1989, The Early Mediaeval Archaeol- bon 55 (2–3), 720–730. ogy of Croatia A.D. 600−900. − BAR international BROWN, M. A. 1983, Grave Orientation: A Further series 539. View. − Archaeological Journal 140 (1), 322−328. EVIS, L. H., HANSON, I., CHEETHAM, P. N. 2016, BRUNDKE, N. 2013, Das Mittelalterliche Gräberfeld An experimental study of two grave excavation Mockersdorf. Archäologie im Schatten des Rauhen methods: Arbitrary Level Excavation and Strati- Kulm.− Pressath. graphic Excavation. − STAR: Science & Technology CHAPLOT, V., DARBOUX, F., BOURENNANE, H., of Archaeological Research 2 (2), 177−191. LEGUÉDOIS, S. A. N., PHACHOMPHON, K. FABECH, C. 1999, Centrality in sites and Landscapes. 2006, Accuracy of interpolation techniques for the − In: C. Fabech, J. Ringtved (eds.), Settlement and derivation of digital elevation models in relation to landscape. Proceedings of a conference in Århus, landform types and data density. − Geomorphology Denmark, May 4−7 1998, Aarhus, 455−474. 77, 126−141. FILIPEC, K. 2016, Pokopi u zgrčenom i poluzgrčenom CLARK, M. A., WORRELL, M. B., PLESS, J. E. 1997, položaju na groblju uz crkvu majke božje gorske u Postmortem Changes in Soft Tissues. − In: W. D. Loboru. – In: Krznar et al. , 49−56. Haglund, M. H. Sorg (eds.), Forensic taphonomy: FLETCHER, R. J. 2004, Materiality, Space, Time, and the postmortem fate of human remains. − Boca Outcome. – In: J. Bintliff (ed.), A Companion to Raton, London, New York, Washington, 151−164. Archaeology, Malden, Oxford, Carlton. 102 12 BIBLIOGRAPHY FRY, S. L. 1999, Burial in Medieval Ireland, 900−1500. A HAZELWOOD, L., STEELE, J. 2004, Spatial dynamics Review of the Written Sources. − Dublin. of human dispersals. − Journal of Archaeological GASPARI, A. , HUMERCA, R. , ŽVAN, J. 2021, Science 31 (6), 669–679. Zgodnjesrednjeveška deblaka v Blejskem jezeru. − HERMANN, B. 1989, Anthropologische Untersuchun- In: A. Gaspari, P. Novaković (eds.), Arheologija v gen am Skelettmaterial der Burgwal bevölkerung. − letu 2020 dediščina za javnost. Zbornik povzet- In: A. von Muller, K. von Muller-Muči (eds.), kov. Strokovno srečanje Slovenskega arheološkega Ausgrabungen, Funde und naturwissenschaftliche društva na spletni videokonferenci 10. decembra Untersuchungen auf dem Burgwall in Berlin- 2021, Ljubljana. Spandau, Berlin, 175−197. GEARY, P. J. 2001, Peasant Religion in Medieval Europe. HERZOG, I. 2004, Stratify, 1.5. − Bonn. (internet source: − Cahiers d'Extrême-Asie 12 (1), 185−209. http://www.stratify.org/Download/Stratify_Manu- GLEIRSCHER, P. 2006, Frühmittelalterlicher Kirchen- al.pdf; last access: 14.9.2022). bau zwischen Salzburg und Aquileia. Ein Diskus- HILL, A., BEHRENSMEYER, A. K. 1984, Disarticula- sionsbeitrag. − Beiträge zur Mittelalterarchäologie tion patterns of some modern East African mam- in Österreich 22, 61−80. mals. − Paleobiology 10 (3), 366–376. GONZÁLEZ-GARCÍA, A. C., BELMONTE, J. A. 2015, HINTON, I. 2006, Church Alignment and Patronal The Orientation of Pre-Romanesque Churches in Saint’s Days. − The Antiquaries Journal 86, 206−226. the Iberian Peninsula. − Nexus Network Journal HINTON, I. 2012. The Alignment and Location of Me- 17 (2), 353−377. dieval Rural Churches. − BAR British Series 560. GORDON, B. L. 1971, Sacred Directions, Orientation, HOARE, P. G., SWEET, C. S. 2000, The orientation of and the Top of the Map. − History of Religions 10 early medieval churches in England. − Journal of (3), 211−227. Historical Geography 26 (2), 162−173. GORDON, C. C., BUIKSTRA, J. E. 1981, Soil pH, Bone HÖFLER, J. 1988, Gradivo za historično topografijo Preservation, and Sampling Bias at Mortuary predjožefinskih župnij na Slovenskem. Pražupniji Sites. − American Antiquity 46 (3), 566−571. Radovljica in Kranj . − Acta Ecclesiastica Sloveniae GORNIK, F. 1967, Bled v fevdalni dobi. − Bled. 10, 201−251. GORNIK, F. 1990, Zgodovina blejske župnije. − Celje. HÖFLER, J. 2015, Gradivo za historično topografijo GRAFENAUER, B. 1960, Struktura in tehnika zgodovin- predjožefinskih župnij na Slovenskem: Kranjska. − ske vede. – Ljubljana. Ljubljana. HAGLUND, W. D., SORG, M. H. (eds.) 1997, Forensic ta- HÖFLER, J. 2016a, Lastniške cerkve zgodnjega sred- phonomy: the postmortem fate of human remains. − njega veka na Gorenjskem in njihovo zgodovinsko Boca Raton, London, New York, Washington. ozadje. − Varstvo spomenikov (49), 64−90. HALSALL, G. 2009, Cemeteries and Society in Merov- HÖFLER, J. 2016b, O prvih cerkvah in župnijah na Slo- ingian Gaul. − Selected Studies in History and venskem. K razvoju cerkvene teritorialne organizaci- Archaeology, 1992−2009. je slovenskih dežel v srednjem veku. − Ljubljana. HAMILTON, A. 1980, Orthodoxy in late fifteenth- HOOKE, D. 2010, Trees in Anglo-Saxon England: Lit- century glass at Leicester. − Transactions of the erature, Lore and Landscape. Leicestershire Architectural and Archaeological Internet source 1, Raster pyramids. – http://desktop. Society 55, 22−37. arcgis.com/en/arcmap/10.5/manage-data/raster- HANULIAK, M. 1984, Problematik der Gräberorien- and-images/raster-pyramids.htm (last access: tierung vom Gesichtspunkt der Aussagefahig-16.9.2022). keit. – In: B. Chropovský (ed.), Interaktionen der JAVOR-BRIŠKI, M. 1998, Geistesgeschichtliche und mitteleuropäischen Slawen und anderen Ethnika im literar historische Aspekte eines spätmittelalter- 6.−10. Jahrhundert: Symposium, Nové Vozokany lichen Privatgebetbuches der National- und 3.−7. Oktober 1983, Nitra, 109−119. Universitätsbi- bliothek von Ljubljana. − Acta HARRIS, E. C. 1979, Principles of archaeological stratig- Neophilologica 31, 3–33. raphy. − London, San Diego, New York, Boston, KAISER, K. 2004, Obredni kot in zgodnjesrednjeveško Sydney, Tokyo, Toronto. grobišče na Brezju nad Zrečami. − Studia Mytho-HARRIS, E., OTTAWAY, P. 1976. A recording experi- logica Slavica 7, 79−88. ment on a rescue site. − Rescue News 12: 6−7. KAJKOWSKI, K., KUCZKOWSKI, A. 2017, Water HARTMANN, W. 2003, Bestattungen und Bestattungs- in pre-Christian beliefs in Pomerania (northern rituale nach dem kirchlichen und weltlichen Recht Poland) of the early medieval period. − Studia des frühen Mittelalters. – In: J. Jarnut, M. Wemhoff Mythologica Slavica 20, 15−32. (eds.), Erinnerungskultur im Bestattungsritual, KENZLER, H. 2015, Religion, Status and Taboo. München, 127−144. Changing Funeral Rites in Catholic and Protestant 103 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND Germany. – In: S. Tarlow (ed.), The Archaeology of MICHCZYŃSKI, A. 2007, Is it Possible to Find a Good Death in Post-medieval Europe, 148−169. Point Estimate of a Calibrated Radiocarbon Date? − KLEVNÄS, A., ASPÖCK, E., NOTERMAN, A. A., VAN Radiocarbon 49 (2), 393–401. HAPEREN, M. C., ZINTL, S. 2021, Reopening MILLER, T. 1999, Th e Old English version of Bede’s Eccle- graves in the early Middle Ages: from local practice siastical history of the English people . − Cambridge. to European phenomenon, − Antiquity 95 (382), MINISTR, Z. 2007, Géniové dávnověku. 7000 let orien- 1005−1026. tace ke slunci. − Praha. KJELLSTRÖM, A. 2022, From Saint to Anthropological MÜLLER, E. 2013, Slawische Bestattungssitten im Specimen: The Transformation of the Alleged Skel- Saalegebiet − die Gräberfelder von Niederutünsch etal Remains of Saint Erik. − In: E. Weiss-Krejci, S. und Oechlitz (Saalekreis, Sachsen- Anhalt). – Becker, P. Schwyzer (eds.), Interdisciplinary Explo- In: F. Biermann, T. Kersting, A. Klammt (eds.), rations of Postmortem Interaction, Springerbooks, Soziale Gruppen und Gesel schafts strukturen im Leiden, 167−188. west slawischen Raum. Beiträge der Sektion zur KJØLBYE‐BIDDLE, B. 1975, A cathedral cemetery: slawischen Frühgeschichte der 20. Jahrestagung des Problems in excavation and interpretation. − World Mittel- und Ostdeutschen Verbandes für Altertums- Archaeology 7 (1), 87−108. forschung in Brandenburg (Havel), 16. bis 18. April KLEJN, L. S. 1987, Arheološki viri. − Ljubljana. 2012, Langenweissbach. KNIFIC, T. 2004, Arheološki sledovi blejskih prebivalcev MYTUM, H. 2006, Popular attitudes to memory, the iz pozne antike in zgodnjega srednjega veka. –In: J. body, and social identity: the rise of external Dežman (ed.), Bled 1000 let, Blejski zbornik 2004, commemoration in Britain, Ireland and New Eng- Bled, 93–117. land. − Post-Medieval Archaeology 40 (1), 96−110. KNÜSEL, C. J. 2014, Crouching in fear: Terms of en- NIEDERLE, L. 1931, Rukověť slovanské archeologie gagement for funerary remains. − Journal of Social (Manuel de ľ archéologie slave) . − Praha. Archaeology 14 (1), 26−58. NOVAKOVIĆ, P. 2001, Prostorska in pokrajinska arhe- KNÜSEL, C. J., ROBB, J. 2016, Funerary taphonomy: ologija: študija na primeru Krasa. – Unpublished An overview of goals and methods. − Journal of Phd, Univerza v Ljubljani, Filozofska fakulteta, Archaeological Science: Reports 10, 655−673. Oddelek za arheologijo, Ljubljana. KRZNAR, S. 2012, Arheološka slika kasnosrednjo- PAPAGEORGOPOULOU, C., RENTSCH, K., RAGHA- vjekovnih groblja na prostoru sjeverne Hrvatske. − VAN, M., HOFMANN, M. I., COLACICCO, G., Unbublished PhD thesis, Filozofski fakultet, GALLIEN, V., BIANUCCI, R., RÜHLI, F. 2010, Pres- Sveučilište u Zagrebu, Zagreb. ervation of cell structures in a medieval infant brain: KRZNAR, S., SEKELJ IVANČAN, T., TKALČEC, T., A paleohistological, paleogenetic, radiological and BELA (eds.) 2016, Groblja i pogrebni običaji u physico-chemical study. − NeuroImage 50, 893−901. srednjem i ranom novom vijeku na prostoru sjeverne PARKER PEARSON, M. 1999, The Archaeology of Death hrvatske. − Zbornik radova prvog međunarodnog and Burial. − Stroud. znanstvenog skupa srednjovjekovne arheologije PEDERSEN, A. 2003, Reiche Bestattungen einer Auf- Instituta za arheologiju Zagreb, 4. lipnja 2014 , bruchszeit − Tradition und Erneuerung in den Zbornik Instituta za arheologiju 4. Grabsitten des 10. Jahrhunderts in Altdänemark. – LANDINGHAM, M. van 1993, The Dying and the Dead In: J. Jarnut, M. Wemhoff (eds.), Erinnerungskultur in Gratian's Decretum. Comitatus. − A Journal of im Bestattungsritual, München, 145−174. Medieval and Renaissance Studies 24 (1), 61−78. PINHEIRO, J. 2006, Decay Process of a Cadaver. − In: lupa = UBI ERAT LUPA – F. und O. Harl, www.ubi- A. Schmitt, E. Cunha, J. Pinheiro (eds.), Forensic erat-lupa.org (Bilddatenbank zu antiken Stein-Anthropology and Medicine: Complementary Sci- denkmälern). ences From Recovery to Cause of Death, Totowa, LOZIĆ, E., ŠTULAR, B. 2007, Od arheologije grobišč NJ, 85−116. proti arheologiji o smrti. Analiza arheoloških PINTAR, A., GLAVAŠ, V. 2017, The archaeology dog raziskav grobišč v Arheološkem vestniku. − Arheo project. – In: C. Schuler, P. Kaul (eds.), Faszinosum (24), 69−79. Spi rhunde: Gefahren sichtbar machen − Gefahren MAKAROVIČ, G. 1995, Slovenci in čas: odnos do časa abwenden Tagungsergebnisse des 3. Symposiums fur kot okvir in sestavina vsakdanjega življenja. − Ljub- Odorologie im Diensthundewesen an der Hochschule ljana. Bonn-Rhein-Sieg, Hamburg, 30−38. MĚŘÍNSKÝ, Z. 1985, Velkomoravské kostrové pohřebiště PLETERSKI, A. 1996, Strukture tridelne ideologije v ve Velkých Bílovicích. K problematice venkovských prostoru pri Slovanih. − Zgodovinski časopis 50 pohřebišt, 9.-10. stol. na Moravě. − Brno. (2), 163−185. PLETERSKI, A. 2003, Struktur des Gräberfeldes Alten- erding. – In: H. Losert, A. Pleterski, Altenerding 104 12 BIBLIOGRAPHY in Oberbayern. Struktur des frühmittelalterlichen PREDOVNIK, K., DACAR, M., LAVRINC, M. 2008, Gräberfeldes und “Ethnogenese” der Bajuwaren, Cerkev sv. Jerneja v Šentjerneju: arheološka izkopa- Berlin, Bamberg, Ljubljana, 505−684. vanja v letih 1985 in 1986. (St Bartholomew Church PLETERSKI, A. 2008, Zgodnjesrednjeveška naselbina in Šentjernej: Archaeological Excavations in 1985 na Blejski Pristavi. Najdbe / Frühmittelalterliche and 1986). − Ljubljana. Siedlung Pristava in Bled. Funde . − Opera Instituti RAJCHL, R. 2001, Astronomické prvky v orientaci Archaeologici Sloveniae 14. pohřebiště Břeclav-Pohansko . − Archaeologia me- PLETERSKI, A. 2010, Zgodnjesrednjeveška naselbina diaevalis Moravica et Silesiana I/2000, 127−132. na blejski Pristavi. Tafonomija, predmeti in čas. / RAPAN PAPEŠA, A. 2016, Prethodna razmišljanja o FrühMittelalterliche Siedlung Pristava in Bled. prilozima keramičkih lonaca i njihovom sadržaju Taphonomie, Fundgegenstände und zeitliche Einord- u avarodobnim grobovima na lokalitetu Nuštar − nung. − Opera Instituti Archaeologici Sloveniae 19. dvorac. – In: Krznar et al., 37−48. PLETERSKI, A. 2013a, Korak v kronologijo zgodnje- REBMANN, A., DAVID, E., SORG, M. H. 2000, Cadaver srednjeveškega naglavnega nakita vzhodnih Alp / dog handbook : forensic training and tactics for the A step towards the chronology of early medieval recovery of human remains. − Boca Raton, London, head ornaments in the Eastern Alps. − Arheološki New York, Washington. vestnik 64, 299−334. REIMER PJ, BARD E, BAYLISS A, BECK JW, BLACK- PLETERSKI, A. 2013b, The invisible Slavs. Župa Bled WELL PG, BRONK RAMSEY C, BUCK CE, in the “Prehistoric” Early Middle Ages. − Opera CHENG H, EDWARDS RL, FRIEDRICH M, Instituti Archaeologici Sloveniae 28. GROOTES PM, GUILDERSON TP, HAFLIDA- PLETERSKI, A. 2014, Kulturni genom. Prostor in njegovi SON H, HAJDAS I, HATTÉ C, HEATON TJ, ideogrami mitične zgodbe. − Studia Mythologica HOFFMANN DL, HOGG AG, HUGHEN KA, Slavica − Supplementa, Supplementum 10. KAISER KF, KROMER B, MANNING SW, NIU PLETERSKI, A. 2017, Die strukturelle Verwirklichung M, REIMER RW, RICHARDS DA, SCOTT EM, des slawischen Altglaubens. − In: F. Biermann, T. SOUTHON JR, STAFF RA, TURNEY CSM, VAN Kersting, A. Klammt (eds.), Religion und Gesell- DER PLICHT J. 2013. IntCal13 and Marine13 schaft im nördlichen westslawischen Raum. Beiträge radiocarbon age calibration curves 0–50,000 years der Sektion zur slawischen Frühgeschichte der 22. cal BP. − Radiocarbon 55(4):1869–1887. Jahrestagung des Mittel- und Ostdeutschen Ver- REMPEL, H. 1966, Reihengräberfriedhöfe des 8. bis 11. bandes für Altertumsforschung in Chemnitz, 29.−31. Jahrhunderts aus Sachsen-Anhalt, Sachsen und März 2016, Langenweissbach, 31−47. Thüringen. − Berlin. PLETERSKI, A. 2020, A mythical landscape. Tests of RIHTER, J. 2018, Staro pokopališče ob župnijski cerkvi its existence with predictive models for the Bled sv. Žalostne Matere Božje v Leskovcu pri Krškem in case. − In: Štular 2020, 235-278. arheološka odkritja. − In: J. Spahalić (ed.), Leskovec PLETERSKI, A., ŠTULAR, B., BELAK, M. (eds.) 2016, pri Krškem skozi čas, Leskovec pri Krškem, 7−40. Izkopavanja srednjeveškega in zgodnjenovoveškega RIHTER, J. 2020, Stratigrafija in tipokronologija grobišča pri Župni cerkvi v Kranju v letih 1964−1970 zgodnjesrednjeveškega nakita: Študijski primer (Excavations of Medieval and Post-Medieval cem- grobišča Župna cerkev v Kranju. − Unpublishd PhD etery at the Župna cerkev in Kranj from 1964 to thesis, Univerza v Ljubljani, Filozofska fakulteta, 1970) . − Opera Instituti Archaeologici Sloveniae Oddelek za arheologijo. 35, Grobišče Župna cerkev v Kranju 1. RODWELL, W. 1989, The archaeology of religious places: PLETERSKI, A., ŠTULAR, B., BELAK, M. (eds.) 2017, churches and cemeteries in Britain. − Philadelphia. Arheološka raziskovanja grobov najdišča Župna SAGADIN, M. 1985, Nekateri elementi socialne struk- cerkev v Kranju med letoma 1972 in 2010 (Archaeo- ture prebivalstva z zgodnjesrednjeveškega grobišča logical investigation of graves from Župna cerkev site ob tovarni Iskra v Kranju. – In: Kranjski zbornik, in Kranj between 1972 and 2010) . − Opera Instituti Kranj, 95−97. Archaeologici Sloveniae 37, Grobišče Župna cerkev SAGADIN, M. 2001, Staroslovansko grobišče na Malem v Kranju 2. gradu v Kamniku (The Early Slavic Cemetery at PLETERSKI, A., ŠTULAR, B., BELAK, M., BEŠTER, H. Mali grad in Kamnik). − Arheološki vestnik 52, (eds.) 2019, Začetek in konec raziskovanj grobišča 359−375. pri Župni cerkvi v Kranju (1953 in 2011−2013). SAGADIN, M. 2017, Prva cerkev v luči arheoloških (The beginnning and end of exploration at the Župna raziskav. – In: R. Peskar (ed.), Župnijska cerkev cerkev cemetery in Kranj (1953 and 2011−2013)). − sv Kancijana v Kranju − njeno obličje in pomen. − Opera Instituti Archaeologici Sloveniae 38, Grobišče Ljubljana, 35−58. Župna cerkev v Kranju 3. 105 GRAVE ORIENTATION IN THE MIDDLE AGES. A CASE STUDY FROM BLED ISLAND SASSIN ALLEN, A. 2015, Church Orientation in the ŠPRAJC, I. 1991, Arheoastronomija. − Ljubljana. Landscape: a Perspective from Medieval Wales. − ŠRIBAR, V. 1966, Poročilo izkopavanj na Blejskem otoku. Archaeological Journal 173 (1), 154−187. − Unpublished report, National museum Slovenia, SCHOLKKMANN, B. 2003, Die Kirche als Bestattung- arch. No. AO NMS 489. splatz. Zur Interpretation von Bestattungen im ŠRIBAR, V. 1967, Tisoč let slovenske arhitekture na Kirchenraum. – In: J. Jarnut, M. Wemhoff (eds.), Blejskem otoku. − Sinteza 5−6, 60−64. Erinnerungskultur im Bestattungsritual, München, ŠRIBAR, V. 1969, O raziskovalnih in koservatorskih 189−218. problemih pri odkrivanju Freisinškega trga Guten- SCOTT, E. M., COOK, G. T., NAYSMITH, P. 2007, Error werth (Otok pri Dobravi na Dolenjskem). − Varstvo and Uncertainty in Radiocarbon Measurements. − spomenikov 13−14, 29−38. Radiocarbon 49 (2), 427–440. ŠRIBAR, V. 1971, Blejski otok – oris zgodovine. – Bled. SÖRRIES, R. 2003, Der mittelalterliche Friedhof, Das = Die Insel im See von Bled (Historischer Abriß). Monopol der Kirche im Bestattungswesen und der = The island in lake Bled (a summary of its past). so genannte Kirchhof. − In: Arbeitsgemeinschaft ŠRIBAR, V. 1972, K dataciji zgodnjesrednjeveške Friedhof und Denkmal e. V and Zentralinstitut und cerkvene arhitekture na Slovenskem (Zur Datier-Museum für Sepulkralkultur (eds.), Raum für Tote. ung frühmittelalterlichen Kirchenarchitektur in Die Geschichte der Friedhöfe von den Gräberstraßen Slowenien). – Arheološki vestnik 23, 384–396. der Römerzeit bis zur anonymen Bestattung, Braun- ŠRIBAR, V. 1974, Ob dokumentaciji arheološkega schweig, 27–53. odkrivanja freisinškega trga Otok pri Dobravi − SPINAZZÈ, E. (2014) La consuetudine medioevale Gutenwerth. − Varstvo spomenikov 17−19/1, 7−18. nel ’orientazione degli edifici sacri secondo il trat- ŠRIBAR, V. 1976, K problemu arheološke dokumentacije tato di Guido Bonatti. – Atti dell' Accademia “San na terenu. − Materijali 12, 371−376. Marco” di Pordenone 16, 521−570. ŠRIBAR, V. 1977, Novi stativ za ortografsko fotogra- SPINAZZÈ, E., 2016a, The Alignment of Medieval firanje. − Varstvo spomenikov 21/1977, 141−143. Churches in Northern-Central Italy and in the ŠRIBAR, V. 1981, Izpopolnitev opreme za izkopavanje. − Alps and the Path of Light Inside the Church on Varstvo spomenikov 23/1981, 117−120. the Patron Saint’s Day. – Mediterranean Archaeol- ŠTIH, P. 2004, Prva omemba Bleda v pisanih virih − ogy and Archaeometry 16(4), 455−463. Listina kralja Henrika II. za briksenškega škoda SPINAZZÈ, E. 2016b, La luce nel ’ architettura sacra: Albuina z dne 10. aprila 1004 (D. H. U. 67. – In: J. spazio e orientazione nelle chiese del X-XII secolo Dežman (ed.), Bled tisoč let. Blejski zbornik 2004, tra Romandie e Toscana. – Beihefte zur mediaevis- Radovljica, 7−34. tik. Monographien, Editionen, Sammelbände 20, ŠTIH, P. 2011, Tisoč let od prve omembe Blejskega gradu Frankfurt am Main. (1011−2011) in začetki blejskega gospostva škofov STRAHLER, A. 2011, Introducing Physical Geography iz Briksna. – In: M. Vidic (ed.), Blejski grad : 1000 (Fifth edition). − Hoboken. let prve omembe, Bled, 8−27. STOPAR, I. 2017, Arhitektura predromanike in romanike ŠTULAR, B. 2007, Posamezniki, skupnost in obred v v Sloveniji. – Ljubljana. zgodnjem srednjem veku. Primer grobiščnih po- STRMČNIK-GULIČ, M. 1994, Legen, Sveti Jurij, za- datkov z Malega gradu v Sloveniji (The Individu- kladnica podatkov. − Slovenj Gradec. als, the Community, and the Ritual in the Early SZCZEPAŃSKI, S. 2015, Old Prussian “Baba” Stones: An Middle Ages. An Analysis of Cemetery Data from Overview of the History of Research and Recep- Mali grad, Slovenia), Studia mythologica Slavica tion. Pomesanian-Sasinian Case / Baby pruskie – 10, 23−50. z dziejów badań i adaptacji kulturowej. Casus ŠTULAR, B. 2008, Analiza dokumentacije nestratigraf- pomezańsko-sasiński. − Analecta Archaeologica skih izkopavanj. − Arheo (25), 45−52. Ressoviensia 10, 313−364. ŠTULAR, B. 2009, Mali grad. Visokosrednjeveški grad v ŠAŠEL KOS, M. 1999, Promet v prazgodovini. − In: Kamniku / Mali grad. High Medieval Castle in Kam- Božič, D. et al. (eds.), Zakladi tisočletij. Zgodovina nik. − Opera Instituti Archaeologici Sloveniae 15. Slovenije od neandertalcev do Slovanov, Ljubljana, ŠTULAR, B. 2010, Jožefinski vojaški zemljevid kot vir za 254−256. preučevanje preteklih pokrajin (Josephine military ŠKERLJ, B. 1952, Orientacija grobov v nekaterih map (1763−1787) as a source for a study of past srednjeveških nekropolah Slovenije (The Orienta- landscapes). − Geografski vestnik / Geographical tion of Graves in Some Mediaeval Necropoles of Bulletin 82 (1), 85−96. Slovenia). − Arheološki vestnik 3 (1), 99−136. ŠTULAR, B. 2013, Grad Smlednik. − In: B. Štular (ed.), ŠOLLE, M. 1959, Knížecí pohřebiště na Staré Kouřim Grad Smlednik. Raziskave 2011−2012, Ljubljana, (Fürstliche Bestattungsstätte in Stara Kouřim). − 135−148. Památky archeologické 50 (2), 353−506. 106 12 BIBLIOGRAPHY ŠTULAR, B. (ed.) 2020, Sred njeveški Blejski otok v arheo- VLASTO, A. P. 1970, The Entry of the Slavs into Chris- loških virih / Medieval archaeology of Bled Island. − tendom. An Introduction to the Medieval History of Opera instituti archaeologici sloveniae 42. the Slavs. − Cambridge. ŠTULAR, B., BELAK, M. (eds.) 2012, Grobišče Župna VRANOVÁ, V., MARFO, T. D., REJŠEK, K. 2015, Soil cerkev v Kranju. Dokumentacija o izkopavanjih v scientific research methods used in archaeology - letu 1953. − E-Monographiae Instituti Archaeo- promising soil biochemistry: a mini-review. − Acta logici Sloveniae 2. universitatis agriculturae et silviculturae mendeli- ŠTULAR, B., BELAK, M. (eds.) 2012, G robišče Župna anae brunensis 63 (4), 1417−1426. cerkev v Kranju. Dokumentacija o izkopavanjih v VYROUBAL, V., PLEŠE, T., NOVAK, M. 2016, Rezul- letih 1964, 1965 in 1966. − E-Monographiae Insti- tati antropološke analize osteološkog materijala tuti Archaeologici Sloveniae 3. pronađenog u crkvi pavlinskog samostana svih ŠTULAR, B., HROVATIN, I. M. 2002, Slovene pagan svetih u strezi − prijedlog povijesno-arhivske sacred landscape. Study case: the Bistrica Plain. − kontekstualizacije i interpretacije. – In: Krznar et Studia mythologica Slavica 5, 43−68. al., 87−106. ŠTULAR, B., LOZIĆ, E. 2016, Primernost podatkov pro- VOGEL, C., NÉDONCELLE, M., BOTTE, B., De jekta Lasersko skeniranje Slovenije za arheološko BRUYNE, L., GRABAR, A., MARICHAL, R., interpretacijo: metoda in študijski primer. – In: R. MOHRMANN, C. 1962, Sol æquinoctialis. Pro- Ciglič, M. Geršič, D. Perko, M. Zorn (eds.), Digi- blèmes et technique de l’orientation dans le culte talni podatki, Ljubljana, 157−166. chrétien . − Revue des Sciences Religieuses 36 (3−4), ŠTULAR, B., LOZIĆ, E., BELAK, M., RIHTER, J., 175−211. KOCH, I., MODRIJAN, Z., MAGDIČ, A., KARL, WEMHOFF, M. 2003, Bestattungen des Früh- und Hoch- S., LEHNER, M., GUTJAHR, C. 2022, Migration of mittelalters an und in westfälischen Kirchen. – In: Alpine Slavs and machine learning: Space-time pat- J. Jarnut, M. Wemhoff (eds.), Erinnerungs kultur im tern mining of an archaeological data set. – PLOS Bestattungsritual, München, 97−106. ONE 17 (9), https://journals.plos.org/plosone/ WILLIAMS, H. 2006, Death and Memory in Early Me- article?id=10.1371/journal.pone.0274687. dieval Britain. − Cambridge. TRUEMAN, C. N. 2019, Building a Medieval Cathedral, ZVKD 2022, Bled, Blejsko jezero Dvig deblaka danes, (internet source: www.historylearningsite.co.uk/ 14.10.2022, Zavod za varstvo kulturne dediščine, medieval-england/building-a-medieval-cathedral/; Novice, 14. September (https://www.zvkds.si/ last access: 15.9.2022). sl/novice/bled-blejsko-jezero-dvig-deblaka- VALCÁRCEL, J. P., PALMERO, V. P. 2018, La orien- danes-14102022, accessed 21.10.2022). tación de las iglesias mozárabes. − En la España Medieval 41, 171−197. 107 Document Outline 1_zacetek_kk 2_e_Blejski_otok_besedilo_kk 3_e_Blejski_otok_lit_kk