UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ancient Peruvian sprang fabrics Frame, Mary Patricia 1982

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1982_A8 F73.pdf [ 30.87MB ]
Metadata
JSON: 831-1.0095501.json
JSON-LD: 831-1.0095501-ld.json
RDF/XML (Pretty): 831-1.0095501-rdf.xml
RDF/JSON: 831-1.0095501-rdf.json
Turtle: 831-1.0095501-turtle.txt
N-Triples: 831-1.0095501-rdf-ntriples.txt
Original Record: 831-1.0095501-source.json
Full Text
831-1.0095501-fulltext.txt
Citation
831-1.0095501.ris

Full Text

ANCIENT PERUVIAN SPRANG FABRICS hy MARY PATRICIA FRAME B.A., The University of B r i t i s h Columbia, 1978 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF • THE REQUIREMENTS FOR THE DEGREE OF MASTER OF-. ...*.'. ARTS i n - THE FACULTY OF GRADUATE STUDIES (Fine Arts Department) We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA Octoher 1982 (c} Mary P a t r i c i a Frame, 1982 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the head of my, department or by his or her representatives. I t i s understood that copying or publication of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of FINE ARTS  The University of B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 D a t e OflTCRER 18, 1982 DE-6 (3/81) ABSTRACT This paper attempts to develop a technical feature of ancient Peruvian f a b r i c s as a dating tool by describing i t s occurrence as f u l l y as possible. The fab r i c s i n t h i s study share a common technique or method of manufacture which has been c a l l e d "sprang". Sprang i s a weftless tech-nique of interworking a set of p a r a l l e l elements fi x e d at both ends and i t i s characterized by the duplication of fab r i c i n mirror-image symmetry at both ends of the warp. Structural p e c u l i a r i t i e s which remain i n the finish e d f a b r i c allow the i d e n t i f i c a t i o n of the method of manufacture. A large sample of sprang f a b r i c s was i d e n t i f i e d , analyzed, r e p l i -cated and diagrammed. They were grouped on the basis of structural s i m i l a r i t y . The fabrics were dated through grave association, where pos-s i b l e , "but primarily through comparative s t y l i s t i c and iconographic analysis. The andent Peruvian sprang f a b r i c s are presented chronologically by horizon and period. Within each time segment, the sample i s grouped by structure. What re s u l t s i s a p r o f i l e of the sprang technique - the variations i n structures, the modifications to the basic technique and the d i f f e r i n g functions of the finished f a b r i c s - as i t i s used through time. By cross-checking technical groupings with grave association, style and iconography, i t emerges that certain technical features are diagnostic of the originating culture. The application of t h i s informa-t i o n includes the i d e n t i f i c a t i o n of undocumented fabrics i n museum c o l -l e c t i o n s and the expansion of the iconographic repertory of some s t y l e s . The evolution of the techniques of sprang i n ancient Peru i s broadly sketched and some in t e r n a l sequencing of the larger groups of technically associated sprang f a b r i c s i s proposed. The iconographic study of the sprang f a b r i c s leads to some insights into the importance of f a b r i c making within ancient Peruvian culture. In p a r t i c u l a r , the repeated association of the serpent with images of fab r i c structures i n the sprang sample of the Early Horizon illuminates the significance of fabric and f a b r i c making. i v TABLE OF CONTENTS ABSTRACT i i LIST OF TABLES v LIST OF FIGURES v i ABBREVIATIONS x i i i ACKNOWLEDGEMENT . . . • x i v CHAPTER I : INTRODUCTION 1 General Considerations Terminology I d e n t i f i c a t i o n of the Sprang Technique Chronology and Style I d e n t i f i c a t i o n CHAPTER I I : PRECERAMIC AND INITIAL PERIODS ...... Ik CHAPTER I I I : THE PARACAS-NASCA TRADITION OF THE EARLY HORIZON AND EARLY INTERMEDIATE PERIOD . . 2 6 A. I n t e r l i n k i n g 1. C e r r i l l o s hand 2. Openworks a) Sample b) Iconography c) Dating 3 • Striped Sample B. Interlacing 1. Small bags 2. Bags from the Early Intermediate Period 3• Openwork 4. Garment t i e s C. Intertwining CHAPTER IV: THE MIDDLE HORIZON 128 A. Tassel sample 1. Category 1 and IB 2. Category 2 3. Category 3 4. Category 4 B. Iconography C. Chronological Implications CHAPTER V: SUMMARY AND CONCLUSIONS BIBLIOGRAPHY 218 2 2 5 LIST OF TABLES TABLES I Chronological Table I I Occurrence of the sprang technique i n the thesis sample LIST OF FIGURES PAGE F i g . 1 Working a fabric i n the sprang technique 5 F i g . 2 Chaining the terminal area of a sprang fa b r i c 6 F i g . 3 The basic structures of the sprang technique 8 F i g . 4 Structure of a preceramic bag from Asia 1 5 F i g . 5 Structural d e t a i l indicating a non-sprang technique 1 7 F i g . 6 Asia f a b r i c described as "loose sprang" . 18 F i g . 7 Three bar loom with a false c i r c u l a r warp 20 F i g . 8 Diagrams of weft-twined fabrics from Huaca P r i e t a with h e l i c a l warp paths _ 22 F i g . 9 Diagrams of weft-twined f a b r i c s from Huaca P r i e t a with l i n k i n g warps 2 3 F i g . 10 Technical d e t a i l s of the e a r l i e s t , confirmed sprang fabric 2 7 F i g . 11 Structure of south coast headdresses 31 F i g . 1 2 General features of Early Horizon i n t e r l i n k e d sprang 3 2 F i g . 1 3 Interlinked sprang turban from Paracas Gavernas 34 F i g . 14 Pattern repeat on an i n t e r l i n k e d sprang turban from Paracas Gavernas 35 F i g . 1 5 Arrangement of headdresses, Paracas Gavernas 3 6 F i g . 16 D e t a i l of a Paracas turban showing gathering loops 37 F i g . 17 Paracas Gavernas turban with double-headed serpent repeat 39 F i g . 18 Paracas turban with serpentine g r i d design 40 F i g . 1 9 Interlinked sprang bag (?) with zigzag design 41 F i g . 20 Interlinked sprang bag (?) with branching serpent design 42 F i g . 21 Interlinked sprang turban with branching serpent design, Gabeza Larga 43 \ v i i FIGURES PAGE F i g . 22 Interlinked sprang turban with serpentine g r i d 45 F i g . 23 Interlinked sprang hood with double-headed serpents, Ocucaje 4-7 F i g . 24- Interlinked sprang hood with "Oculate Being", Ocucaje 4-8 F i g . 25 Branching serpentine design on a fragmentary i n t e r l i n k e d sprang f a b r i c from Ocucaje 4-9 F i g . 26 Interlinked sprang hood with branching serpentine design, Ocucaje 5 f t F i g . 27 Interlinked sprang hood with serpent head design 52 F i g . 28 Five variations of the branching serpent design on i n t e r l i n k e d sprang fabrics 54-F i g . 29 Branching serpent designs and three strand braids 56 F i g . 30 Braided serpent belts on a Ghavinoid painted t e x t i l e 57 F i g . 31 Possible derivation of a serpent design from a link e d structure 58 F i g . 32 Images of linked structures on other fabrics 59 F i g . 33 Serpentine grids on i n t e r l i n k e d sprang openworks 60 F i g . 34- Twisted (or twined?) strand image 62 F i g . 35 Twisted (or twined?) strand image 63 F i g . 36 Double-headed serpents within a complex g r i d 65 F i g . 37 Dawson's t r a i t chart f o r features of painted mummy masks from Ocucaje 68 F i g . 38 Three versions of the Oculate Being on an i n t e r -l i n k e d sprang hood from Ocucaje 70 F i g . 39 Imagery on a gauze mantle from Paracas Gavernas 72 F i g . 4-0 Gauze and p l a i n weave structure contemporaneous with i n t e r l i n k e d sprang openworks 73 F i g . 4-1 D e t a i l of a striped, i n t e r l i n k e d sprang bag from •Paracas Necropolis 76 v i i i FIGURES PAGE F i g . 42 Two interlaced sprang bags from Paracas 78 F i g . 43 Interlaced structure of F i g . 4-2 bags 79 F i g . 44- Interlaced sprang bag from the Early Intermediate Period 81 F i g . 45 Interlaced sprang bag from Poroma, Tunga Valley 83 F i g . 46 Interlaced structures of sprang bags from the Early Intermediate Period 84 F i g . 47 Interlaced sprang bag with three pendant pockets 85 F i g . 48 Discontinuous warp set-up f o r the F i g . 47 bag 86 F i g . 49 Interlaced bag with three pockets shaped by making s l i t s i n the sprang f a b r i c 88 F i g . 50 Interlaced sprang openwork 92 F i g . 51 Paracas Necropolis s k i r t with interlaced sprang -ties 93 F i g . 52 Details of s k i r t t i e s 95 Fig« 53 Structure of interlaced sprang t i e s 96 F i g . 54 Method of working two t i e s on a folded warp 97 F i g * 55 D e t a i l of an intertwined sprang openwork 101 F i g . 56 Structure of intertwined sprang openworks 102 F i g . 57 Fragmentary intertwined sprang openwork, Gahuachi 103 F i g . 58 Unfinished intertwined sprang warp 106 F i g . 59 Large, intertwined sprang fa b r i c that may have been a mummy wrapping 107 F i g . 60 Geometric designs on intertwined sprang fabrics 109 F i g . 6 l Figurative designs on intertwined sprang fabrics 110 F i g . 62 Serpent braid with geometric markings 111 F i g . 63 Intertwined fragment with a variant structure 113 i x FIGURES PAGE F i g . 64 Intertwined openworks with a variant structure 114 F i g . 6 5 Intertwined openwork with a variant structure 1 1 5 F i g . 6 6 Hold l i n e d e t a i l s of an intertwined openwork made i n two sections 117 F i g . 67 D e t a i l showing where two separately worked sections abutt 118 F i g . 68 Schematic drawing of the reconstructed f a b r i c i n F i g . 59 120 F i g . 6 9 An openwork f a b r i c , probably intertwined sprang, from Paracas Gavernas 122 F i g . 70 Intertwined sprang t a s s e l , Middle Horizon IB 1 2 9 F i g . ?1 Intertwined sprang tassel from Gahuachi 1 3 1 F i g . 72- P a i r of intertwined sprang tassels 1 3 2 F i g . 73: Category 1 ta s s e l 1 3 6 F i g . 74 Unfinished t a s s e l i l l u s t r a t i n g the technique of sprang was used 1 3 7 F i g . 7 5 Structure of Category 1 tassels shown i n a single layer of fab r i c 1 3 9 F i g . 7 6 Category 1 tassels: the structure i n two interpene-t r a t i n g layers with v e r t i c a l joins 140 F i g . 7? Category 1 t a s s e l d e t a i l showing top layer only 141 F i g . 7 8 Replicating intertwined sprang on two interpenetrating layers 142 F i g . 7 9 D e t a i l of cross-knit looped warp cover 144 F i g . 80 Maze-like style of patterning i n Category 1 tassels 146 F i g . 81 Category 1 ta s s e l with diagonal l i n e s b u i l t up of r e c t i l i n e a r modules 148 Fig.. 82 Tassel warp cover with more c l e a r l y d i f f e r e n t i a t e d figure 149 X FIGURES . PAGE F i g . 8 3 A favoured colour combination i n Category 1 tassels 1 5 0 F i g . 84 P r o f i l e figures are repeated figures i n Category 1 tassels . 1 5 2 F i g . 8 5 A repeating figure with a p r o f i l e head and a dorsal view of the body 1 5 3 F i g . 8 6 A single, central image shown from a dorsal viewpoint 1 5 4 F i g . 87 A single image shown from a f r o n t a l viewpoint 1 5 5 F i g . 8 8 P a i r of Category IB tassels with four colours 1 5 8 F i g . 89 P a i r of Category IB tassels with four colours -quadruple layer intertwined sprang 1 6 0 F i g . 9 0 Category 2 ta s s e l 1 6 2 F i g . 9 1 Intertwined structure of a Category 2 ta s s e l shown as a single layer of fa b r i c 1 6 4 F i g . 9 2 Intertwined structure of a Category 2 ta s s e l showing the two interpenetrating layers of fa b r i c s 1 6 5 F i g . 9 3 Category 2 ta s s e l 1 6 6 F i g . 9 4 P a i r of Category 2 tassels 1 6 ? F i g . 95 An aberrant t a s s e l with embroidered designs, grouped with Category 2 tassels 168 F i g . 9 6 Category 3 ta s s e l 1 7 0 F i g . 97 Diagram of the obliquely twined and interlaced structure of Category 3 tassels 1 7 2 F i g . 9 8 Inside a Category 3 tas s e l : The warps of both layers have heading cords 1 7 3 F i g . 9 9 Category 3 ta s s e l 1 7 5 F i g . 1 0 0 Category 3 ta s s e l ' 1 7 6 F i g . 1 0 1 Category 3 ta s s e l • 1 7 7 F i g . 1 0 2 Category 4 ta s s e l 1 7 9 FIGURES PAGE Fig. 1 0 3 Category 4 tassel 181 Fig. 104 Category 4 tassel 182 Fig. 1 0 5 Tassel figures that may relate to the "Ayacucho serpent" 184 Fig. 1 0 6 Tassel figures that may relate to the "humped animal" 1 8 5 Fig. 1 0 7 Ventrally extended tassel figures 186 Fig. 108 Tassel figures relating to the head of the "stinger animal" 187 Fig. 1 0 9 Tassel figures with barbed t a i l s 189 Fig. 1 1 0 Various representations of the "bird head" element which i s probably a trophy head motif 1 9 0 Fig. I l l Tassel figures that may be "full-bodied" trophy heads 1 9 2 Fig. 1 1 2 Zoomorphs with outstretched foreleg, possibly r e l a -ting to the humped animal and implying trophy head 1 9 4 Fig. 1 1 3 Three zoomorphs. with outstetched forelegs 1 9 5 Fig. 1 1 4 Varied ray motifs 1 9 7 Fig. 1 1 5 Ray motifs terminating extensions from mouths, heads and headdresses 1 9 8 Fig. 1 1 6 Distorted variant of a ray motif 1 9 9 Fig. 1 1 7 Ray motifs as outlines 2 0 0 Fig. 118 Geometric motifs i n the tassel sample 2 0 2 Fig. 1 1 9 Tassel figures that appear to be ungulates, possibly llamas 204 Fig. 1 2 0 Warp-patterned figures on Category 3 tassels 2 0 5 Fig. 1 2 1 Tassel with a staff holding human and a scrambled figure 2 0 6 x i i FIGURES PAGE Fi g . 1 2 2 Nasca 7 pot with a warrior carrying a feathered s t a f f • •- 2 0 7 Fi g . 1 2 3 P r o f i l e human carrying a s t a f f 208 Fi g . 124 Two horizontal figures that may.be humanoid 2 0 9 Fi g . 1 2 5 Warp cover figure possibly related to Tiahuanaco staff-bearers 2 1 0 Fi g . 1 2 6 Zoomorphs with humanoid heads 2 1 2 Fi g . 1 2 ? Tassel with anatomically scrambled figures 2 1 3 x i i i ABBREVIATIONS AMNH American Museum of Natural History BME Basel Museum of Ethnography BMS Burke Museum, Seattle CAI Chicago Art I n s t i t u t e , DO Dumbarton Oaks FI Collection of F r i t z I k l e FM F i e l d Museum JCTM J u l i o C. Tello Museo KO Kanebo, Osaka LM Lowie Museum MAI Museum of the American Indian MDH Musee de 1'Homme MM Museum of Mankind MNAA Museo Nacional de Antropologia y Arqueologia MRI Museo Regional de lea . OAG Ohara Art Gallery, Kobe, Japan PMH Peabody Museum, Harvard PT Collection of Paul Truel TM The Textile Museum ACKNOWLEDGMENTS The experience of writing a thesis has shown me that research i s not an independent undertaking. I would l i k e to acknowledge my g r a t i -tude to the people who assisted me. As my major professor, Alan R. Sawyer has encouraged me since I stumbled through the door of my f i r s t Peruvian art course. His b e l i e f that I could contribute to the f i e l d of Peruvian studies eventually transferred to me and gave me the impetus to start graduate work. He offered tangible help of many sorts - h i s l i b r a r y and s l i d e archive, introductions to colleagues, grant recommendations, a study t r i p to Peru - i n a phrase, complete access to h i s resources as a Peruvianist of many years standing. Alan Sawyer brings together the rare combination of devotion to the study of Peruvian art and devotion to h i s students. I have met with extraordinary generosity i n other researchers. Nobuko K a j i t a n i s i g n i f i c a n t l y increased the sample of sprang f a b r i c s i n t h i s study by f r e e l y sharing her own unpublished research on the subject. Peter Gollingwood, through h i s book, classes and correspondence, shared h i s research on sprang, offered advice and provided a l u c i d model for technical matters. Noemi Speiser has been a stimulating correspon-dent who led me to consider technical and terminological problems more clos e l y . The l i s t of those who have kindly pointed out further examples of sprang fabrics or offered information or comments reads l i k e a Who's Who of Peruvian studies: Elizabeth Benson, Junius Bird, Penny Bateman, Barbara Conklin, Larry Dawson, Pat Lyon, J i l l Mefford, Anne Paul, Ann Rowe, Daniel R i f k i n , Dwight Wallace and Noemi Speiser. I am grate-f u l and encouraged by the generous help I received from these people and from curators i n museums where I studied and photographed t e x t i l e s . Funding f o r my graduate program came from the Social Sciences and Humanities Research Council of Canada and from a University of B r i t i s h Columbia Graduate Fellowship. F i n a l l y , I want to express gratitude to my husband, Gary, who pati e n t l y and serenely coped with a not-always-mature student i n the house over a number of years. His encouragement and support were sus-taining and I would l i k e to dedicate the thesis to him. 1 CHAPTER I INTRODUCTION The f a b r i c s of ancient Peru have received the acclaim of c u l t u r a l h i s t o r i a n s , anthropologists and antiquarians ever since European archae-ologists began excavating and publishing t h e i r finds over a hundred years ago. The discovery and excavation of the Paracas tombs i n 1925-27 brought to l i g h t a huge cache of f a b r i c s , u n r i v a l l e d among ancient c u l -tures for t h e i r r i c h colour, elaborate imagery and technical v i r t u o s i t y . The Paracas f a b r i c s , more than any single f i n d , aroused widespread i n t e r -est and established Peruvian t e x t i l e s as a s i g n i f i c a n t a r t i s t i c t r a d i t i o n . Although the beauty and the c u l t u r a l importance of fa b r i c s from ancient Peru are re a d i l y acknowledged, they have been under-utilized i n reconstructing the pre-history of Peru. Ceramics, which are more numerous and which do not require the same s t r i c t l y a r i d conditions f o r preservation, have provided the c r i t e r i a used i n constructing chrono-log i e s . Despite the degradable q u a l i t y of fa b r i c s and th e i r intermittent presence at archaeological s i t e s , f a brics have p a r t i c u l a r virtues as repositories of information. Many fabrics are large or f i n e l y worked and can carry iconography that i s more f u l l y expressed than the often abbreviated symbols that occur on ceramics. For instance, the Carhua t e x t i l e s , a south coast Chavinoid cache, display a range of iconography f a r more extensive than the range on contemporary ceramics. The tech-n i c a l variety and complexity of f a b r i c s from ancient Peru afford another source of information. The number of processes involved i n the pro-duction of f a b r i c (spinning, plying, interworking of the elements, - 2 dyeing, f i n i s h i n g ) and innumerable quantitative variants that are pos-s i b l e at each stage of production, provide a large body of concrete factors that can be described and compared. Junius B i r d has demonstra-ted the importance of technical analysis of fabrics i n h i s work on the preceramic s i t e of Huaca P r i e t a ( 1 9 5 1 - 2 , 1 9 5 2 ) where he was even able to retrieve i n v i s i b l e iconography through laborious, thread by thread analysis. Inspired and encouraged by h i s work, others have made focussed studies of t e x t i l e features. William Conklin, i n h i s weave typology f o r Moche f a b r i c s , has found certain specialized structures to be diagnostic of Moche culture ( 1 9 7 9 , P« 1 & 5 ) • Dwight Wallace has collated spin and p l a i n weave data from early s i t e s which demonstrate a high degree of uniformity i n these basic processes within a geogra-phical area (1975, 1 9 7 9 ) . Ann Rowe has recently supplied the technical information on fabrics from a b u r i a l at Ghanchan and has taken the f i r s t steps toward distinguishing North and Central Coast fabrics i n the Late Intermediate Period (1980). A typology of fabric processes and structures by culture i s beginning to emerge through the c o l l a t i n g of technical information of excavated t e x t i l e s . Some factors i n the typology are proving to be c u l t u r a l l y diagnostic. The technical data from f a b r i c analysis provide a set of features that appear to be as useful as ceramic features, l i k e shape and s t y l e , i n i d e n t i f y i n g c u l -t u r a l context. This paper i s concerned with a group of fabrics sharing a technique c a l l e d sprang which has a s i g n i f i c a n t , though l i m i t e d , presence i n ancient Peru. Sprang i s a weftless technique of p l a i t i n g on stretched threads (Collingwood, 1 9 7 4 ) that i s characterized by the duplication of fabric i n r e f l e c t i v e symmetry at both ends of the f i x e d warp. The 3 sample of sprang fabrics include published pieces and ones i n the photo archive of Alan R. Sawyer. As w e l l , many examples have been added from personal study i n museums i n Peru, the United States and Great B r i t i a n and through extensive correspondence with researchers and curators. I have proceeded to study the sample by technical analysis and r e p l i c a t i o n and have recorded the structural variants i n diagrams. I then ordered the sample chronologically, using grave context and s t y l i s t i c and icono-graphic association. In analyzing and ordering the fab r i c s , my intent has been to provide a p r o f i l e of the technical variants of sprang f a b r i c s as they occur through time. The choice of the sprang technique as the focus of t h i s study was made on two bases. The need for investications of the occurrence of t h i s technique i n c u l t u r a l contexts has been pointed out by Annemarie Seiler-Baldinger (1977» p. 18). A pioneering survey of the world-wide d i s t r i b u t i o n of sprang has been done by Irmgard Weitlaner Johnson ( 1 9 5 0 ) • My own study of Peruvian sprang has shown that the use of t h i s technique i n ancient Peru was more extensive than her study i n d i -cates. In The Techniques of Sprang ( 1 9 7 4 ) , Peter Gollingwood devises and describes methods for reproducing many of the structures of sprang, including h i s t o r i c a l l y occurring types and innovative types. With these works as foundation, i t now seems important to trace the appearance of sprang within the c u l t u r a l framework of ancient Peru to t r y to plot the development of the technique. The second basis for choosing sprang as a subject arises from i t s s i n g u l a r i t y . Sprang i s an unusual technique i n that some evidence of the process or technique i s retained i n the structure of the finished f a b r i c , unless i t i s too fragmentary or cut into paired pieces and separated. This means the technique, i n many cases, can be accurately and c e r t a i n l y i d e n t i f i e d . The structures that can be produced i n the technique of sprang are l i m i t e d to variations of three basic types. From the sample assembled, i t appears that only s p e c i f i c , and often highly complex, variations of the basic structures were used. William Conklin has suggested that, i n general, the more complex the structure of the f a b r i c , the more diagnostic the t e x t i l e i s of the originating culture (l975a» P- 18). Sprang fa b r i c s seem i d e a l l y suited for a tech-n i c a l investigation which might produce s i g n i f i c a n t c u l t u r a l indicators. Furthermore, many sprang fabrics have iconography that i s i d e n t i f i a b l e to a s p e c i f i c culture. The iconography provides comparative material for checking the r e l i a b i l i t y of the technical data as indicators of cu l t u r a l a f f i l i a t i o n . Description and Terminology •To further expand the d e f i n i t i o n given above, the term sprang i n t h i s paper refers to a weftless technique of making fabric on a set of stretched elements which are f i x e d at both ends. The interworking of adjacent elements i s transmitted to both ends of the warp. This weft-less procedure produces fa b r i c simultaneously at both ends of the warp. The fa b r i c grows toward the centre from both directions ( F i g . l ) . In general, when the unworked warps i n the centre grow very short, some method of fastening, l i k e chaining across the warps (Fig . 2) or i n s e r t -ing a weft, i s necessary to prevent the fab r i c from unworking i t s e l f . The transverse halves of the fab r i c are i n t o t a l mirror-image symme-tr y - i n structure, design and even mistakes - because both halves are the product of the same set of hand motions. The basic fabric Fig. 1 Working a fabric in the sprang technique by finger-manipulating a set of fixed-end elements. Photo taken by Alan R. Sawyer. F i g . 2 Chaining the central, terminal area of a f a -b r i c made i n the sprang technique. Photo taken by Alan R. Sawyer. 7 structures which can be worked i n t h i s way are i n t e r l i n k i n g , i n t e r l a c i n g and intertwining ( F i g . 3a,b,c). E a r l i e r l i t e r a t u r e on Peruvian f a b r i c s uses a variety of terms to describe sprang fabrics: twining (O'Neale, 194-2, p. 161) , p l a i t i n g (O'Neale, 194-2, p. 162) , twine p l a i t i n g (O'Neale and Kroeber, 1930, p. 54 , PI. 2 1 ) , loom p l a i t i n g (Bird, 1963b, p. 56 , P i . 10 ) , twined "lace" (O'Neale, 1937, P L 5 4 ) , p l a i t i n g with yarns held firm at both ends (d'Harcourt, 1962, p. 80, P I . 54A, 56B). Some of these terms are used i n the same work to describe f a b r i c s made i n non-sprang methods. I t i s often not clear whether the words are meant to describe a struc-ture only or a structure and the technique. Part of the problem i s that a l l the structures that can be produced by sprang, can be produced by non-sprang methods as w e l l . When dealing with fragments, an i n t e r -l i n k e d sprang f a b r i c may be i d e n t i c a l with one made i n single element l i n k i n g , i f the central, terminal area of construction i s not present. Likewise, an interlaced sprang fragment can be i d e n t i c a l to a fab r i c braided on free ends and an intertwined sprang fragment can be i d e n t i c a l to intertwined bobbin lace. The presence of i n t e r l i n k e d , interlaced or intertwined structures i s not proof that sprang was the^method of work-ing (Collingwood, 1974, p. 3 4 ) . A previous lack of separate, d i s t i n c -t i v e terms for technique and structure as well as the fragmentary na-ture of the archaeological fabrics gave r i s e to the confusing descrip-tions by some authors, especially O'Neale. The f a b r i c structure terminology used i n t h i s paper draws on the monumental work of Irene Emery, The Primary Structures of Fabrics: An  I l l u s t r a t e d C l a s s i f i c a t i o n ( I 9 6 6 ) . Additional terminology for f a b r i c technique r e l i e s extensively on Peter Collingwood 1s excellent book, The Techniques of Sprang (1974). Further advice on fab r i c terminology F i g . 3 The basic structures of the A. I n t e r l i n k i n g B. Interlacing G. Intertwining sprang technique 9 was generously given i n personal communication by Noemi Speiser, Peter Collingwood and Ann Rowe. While acknowledging my indebtedness to these people, I take r e s p o n s i b i l i t y for the terminology used i n t h i s paper. I d e n t i f i c a t i o n of the Sprang Technique Sprang fa b r i c s can be i d e n t i f i e d by certain structural p e c u l i a r i -t i e s that earmark the technique, i f the fabric i s not too fragmentary and i f ' i t has not been cut apart. Many of the complete pieces of an-cient Peruvian sprang have loops at both ends of the warp. In a com-plete piece of i n t e r l i n k i n g , i n t e r l a c i n g or intertwining that has loops at both ends, proof of sprang can be a central hold l i n e that may be chained or wefted, and exact, r e f l e c t i v e symmetry across t h i s l i n e . In i n t e r l i n k i n g and intertwining, the direc t i o n of the l i n k i n g twist or twining twist (S or Z) w i l l be opposite on either side of the hold l i n e (Collingwood, 1 9 7 4 , p. 2 7 3 - 4 ) . Some Peruvian variants of t h i s type have more than one meeting l i n e because an unusually long warp i s worked i n sections. Mirror-imaging of structure and design across each hold l i n e i s then the needed evidence. Centrally unworked warps, with warp loops at both ends, which j o i n two symmetrical fabrics i n any of the possible sprang structures i s also proof that sprang was the tech-nique . Some Peruvian sprang fabrics with centrally unworked warps are cut apart into two pieces. I f the two symmetrical pieces are used to-gether i n one a r t i c l e or i f a symmetrical mistake or i r r e g u l a r i t y i s located, then i t " i s l i k e l y that sprang was the technique. I f the two pieces from one warp are made into separate a r t i c l e s , firm evidence for sprang may be l o s t e n t i r e l y (Collingwood, 1 9 7 4 , p. 1 7 4 ) . Some fragments which have no evidence of the technique, but are si m i l a r to more complete f a b r i c s with proof of sprang, w i l l be considered to have been made i n the same technique. While t h i s i s d i f f i c u l t to substan-t i a t e , the t r a d i t i o n a l and conservative nature of fab r i c processes i n Peru add some weight to t h i s interpretation. Chronology and Style I d e n t i f i c a t i o n A number of authors have constructed chronologies, or abstract space-time gr i d s , to order style changes i n ancient Peruvian a r t . Alan R. Sawyer's book, Ancient Peruvian Ceramics ( 1 9 6 6 ) , i s a major contribution to the d e f i n i t i o n and ordering of ceramic s t y l e s . The chronological table published by John Rowe and Dorothy Menzel (Table l ) i s also based on ceramic s t y l e s . Since the publication of the Rowe and Menzel chronology i n 1 9 6 7 , a number of a r t i c l e s and dissertations have been written that cross-tie t e x t i l e style to segments of t h i s cera-mic chronology and, i n some cases, r e f i n e the in t e r n a l evolution of styl e as i t applies to iconography on t e x t i l e s . Among such writings, the works of Dawson (1979), J . Dwyer ( 1 9 7 1 , 1 9 7 9 ) , N. Dwyer ( 1 9 7 9 ) and A. Rowe (1979») (araPparticularly c r u c i a l to dating the fabrics i n t h i s study. Rowe and Menzel's chronology i s a convenient, r e l a t i v e chronology and i t i s being used throughout t h i s paper, i n order to keep to a single set of chronlogical terms and dates. This chronology employs units of time, or contemporaneity which are c a l l e d horizons and periods ( I 9 6 2 ) . The units of time correspond to style change i n the master ceramic sequence of l e a . The three h o r i -zons, Early, Middle and Late, are times of c u l t u r a l or a r t i s t i c unity i n ancient Peru. The Chavin style dominates the Early Horizon (14-00-400 B.C.), the Huari style dominates the Middle Horizon ( 5 5 0 - 9 0 0 A.D.) and the Inca style dominates the Late Horizon ( 1 4 7 6 - 1 5 3 4 A.D.). The CHRONOLOGICAL TABLE OF THE PERUVIAN COAST Master Sequence North Coast Central Coast lea (South Coast) Relative Chronology accounts, bronze tools bronie jewelry silver, lead molded adobes l e n d fortified towns! pottery molds gold and cooper metallurgy fortified towns large cities (may be earlier) pottery I South Coast) m a i n ( C e n t r a l Coast) tor m i l temples adobes, llamas permanent buildings cotton cultivated N Moche <" IMochlca) Tembladera (Tecapa) Cuplsnlque (Chavin) Guanape lea 10 Chancay P a c h a c a a i o H u o r l I n f l . t N l i . t f l o Lima (Interlocking) ~ Mlramar (White-on-rtd) - J A n t o n (Chavin) Rio Sees (Chuquitanta) Ocucaje (Paracas) Chavin I n f l u en ci Colonial Period Late Intermediate Period Middle Horizon Early Intermediate Period Early Horizon Initial Period CHRONOLOGICAL TABLE OF THE PERUVtAN SIERRA Relative North North-Central S o u t h - C e n t r a l South Bolivia Chronology Ancash ' Ayaeucho Cuzco P u n o La Paz C e I o n l o I L . I . H o , I . Q Late Intermediate Period Middle Horizon Early I ntermediate Period Early Horizon Initial Period Local Styles R e a l t y t —\ EF 1-ca • • • I n a n e * l - c w l e l (Cuice) btca Tanta 'Urqo Huarpa - 1 -? w l c h q a m ? Earty Inca IK'llim H u o r l l n f l . . n < Darlved Chanapata total Stylas _ , ... Qaluyu Early Tlahuanaco ( Q e y a l .Table 1^ Chronological Table (Rowe and Menzel, 196?) Early Intermediate Period (400 B.C.-550 A.D.) and the Late Intermediate Period (900-14-76 A.D.) are times of regional d i v e r s i t y and f a l l between the horizons. The Preceramic Period (3000-2100 B.C.) and the I n i t i a l Period (2100-1400 B.C.) are e a r l i e r than the horizons or intermediate periods. The absolute dates given are s t i l l open to question and precise cross-tying from the lea master sequence to other ceramic sequences i s not complete. Archaeologists are not i n complete agreement on the spa-t i a l or temporal extent of some styles nor on the absolute dates that correspond to st y l e change. Although chronological matters are f a r from settled, there i s considerable accord among scholars on the r e l a -t i v e ordering of styles and the i n t e r n a l evolution within s t y l e s . Most of the fabrics i n t h i s study can be placed i n a general c u l -t u r a l context through a combination of comparative techniques. F i r s t l y , a few have been excavated s c i e n t i f i c a l l y and t h e i r ceramic or other associations have been recorded. These few pieces are i d e n t i f i e d on the basis of stratigraphy, carbon 14 dating of associated remains, or association with ceramics of a known s t y l e . Other f a b r i c s , of unknown o r i g i n , can be grouped with these f a b r i c s i f they share important fea-tures. The shared features may be s t y l i s t i c and iconographic or they may be technical and s t r u c t u r a l . Because the str u c t u r a l features of sprang fabrics are p a r t i c u l a r l y complex and s p e c i f i c , they have been used to add i n additional pieces which lack comparative iconography or s t y l i s t i c features. The v a l i d i t y of associating fabrics through speci-f i c technical features w i l l be i l l u s t r a t e d throughout the paper as f a -b r i c s that have been grouped on the basis of a consistent style and iconography also exhibit a common technique and structure. Some technically homogenous groups of fabrics i n t h i s study are without a single excavated example. Neither does the iconography relate closely to that found on ceramics. Even for placing them i n a general i. c u l t u r a l context, i t i s necessary to r e l a t e them to broad s t y l i s t i c fea-tures i n other t e x t i l e s which have grave associations with ceramics that have been i d e n t i f i e d . As Andean chronology i s based on ceramic sequences t h i s circuitous method i s sometimes the only one open for dating f a b r i c s . Even when the general placement of a group of sprang fabrics r e l i e s on several slender connections, a contribution to the chronology of fabrics can sometimes be made. A large group of sprang tassels which lack exca-vation data have been arranged i n an i n t e r n a l sequence, based on small changes i n structure and technique. The sequencewas then cross-checked with the d i r e c t i o n of s t y l i s t i c and iconographic change which i s known for ceramics or other f a b r i c s . The incomplete state of the knowledge of Andean art styles demands the use of the diverse techniques employed i n t h i s paper f o r i d e n t i f y i n g f a b r i c s . CHAPTER I I PRECERAMIC AND INITIAL PERIODS In chronicling the appearance of the sprang technique i n ancient Peru, the published data from excavations should be the prime reference for setting the e a r l i e s t dates. However, the incompleteness of the archaeological record i n the early periods, as well as errors i n d i s t i n -guishing the products of the sprang technique, weaken the a b i l i t y of t h i s source to r e f l e c t the approximate date that sprang came into use. A more speculative method of comparing the l e v e l of technology required f o r sprang with the l e v e l of technology exhibited by extant f a b r i c s can only suggest where the sprang technique might be expected to appear, i f the archaeological picture were more complete. The e a r l i e s t report of sprang i s from the l a t e preceramic s e t t l e -ment of Asia on the central coast (Engel, I 9 6 3 ) • However, I have studied the report arid doubt the accuracy of the i d e n t i f i c a t i o n of sprang as the technique. Six photos or diagrams of one bag (Cat. No. 0 . 7 9 8 ) , said to be loom-plaited or sprang, suggest a single element constructional method rather than sprang. The close-up photo of one area of the bag does show a structure consistent with both the techniques of sprang and single ele-ment l i n k i n g . The structure combines S and Z l i n k i n g with oblique i n t e r -l a c i n g (Fig. 4-). The supporting photos and diagrams, however, show fea-tures of t h i s bag that are inconsistent with the sprang technique. The bottom of the bag (Engel, 1963i Pig- 58) shows a central point with radiating and increasing numbers of r i b s . The diagram of an area where the r i b s increase i n number show threads i n jumbled paths of varying F i g . 4 Structure of a preceramic bag from Asia A. Engel, I963 , Fig. 62, Cat. No. 0.798 B. Combination of S and Z linking and oblique interlacing lengths (F i g . 5 ) - In sprang, the set of elements i s of a uniform length. These i r r e g u l a r thread movements, and the varying thread lengths they require, could not be encompassed on a sprang warp. They are, on the other hand, quite consistent with single element l i n k i n g . Nowhere are a hold l i n e or symmetrical i r r e g u l a r i t i e s mentioned or shown. A l l the i n d i -cators suggest a single element technique where the bag i s begun at the bottom with a single thread and b u i l t up i n s p i r a l l i n g rows, with extra stitches added to increase the size and shape the bag. The reconstructed bag (Engel, 1963, F i g . 7 9 ) i s similar i n shape and shaping to other bags from Asia made i n the single element techniques of l i n k i n g and looping. I t must be concluded that the analyst working with these t e x t i l e s did not have a clear conception of sprang, perhaps confusing the presence of a lin k e d structure that can (but need not) be produced by the sprang technique. The second example, described as having "loose sprang" (p. 4-0) add-ed to the side selvedges of a looped rectangle, i s an equally dubious case. Though fragmentary, i t appears that parts of both end selvedges of each addition are present (Fig. 6 ) . The end selvedges appear to have looped ends. I f these additions were separate i n t e r l i n k e d sprang pieces, a hold l i n e and a change of twist d i r e c t i o n would be v i s i b l e i n the centre of each added piece. I f the two additions were part of one sprang warp, then each addition would have one cut warp end. Neither asf these fea-tures appears i n the photograph. A l l the v i s i b l e features are consistent with single element l i n k i n g and not with i n t e r l i n k e d sprang. The report of sprang from the preceramic s i t e of Asia appears to be based on an erroneous understanding of the sprang technique. Sprang has not been reported from other preceramic s i t e s though i t would not be Fig. 5 Structural detail of the Asia bag indicating a non-sprang technique (Engel, 1963, Fig. 64, Cat. No. O.798) (arrow indicates increased r i b area) FIG. 87. A belt or pouch; use unknown. The central part is looped, with long edges sewn together. Sprang fabric was added to both ends; yarns are of cotton. F i g . 6 A preceramic fabric from Asia that i s described, probably erroneously, as having "loose sprang" attached (Engel, I 9 6 3 , Fig. 8 7 , Gat. No. 0.1018) surprising i f i t had been. Junius Bird's careful analysis and descrip-t i o n of the predominantly weft-twined fabrics of Huaca P r i e t a , particu-l a r l y of the warp selvedges, gives the best insight into the methods of working used by preceramic people (1951-2, 1952, I 9 6 3 , I 9 6 8 ) . The l e v e l of t e x t i l e development and the technology applied to the production of the weft-twined and p l a i n weave fab r i c s i s s u f f i c i e n t for the production of fabrics by sprang. The making of firm, continuous elements by spin-ning f i b r e s and plying them i s well established at the preceramic site s of Huaca P r i e t a (Bird, 1951-2, p. 74), Asia (Engel, 1963, p. 24-5) and at the Ancon-Chillon area s i t e s (Moseley, 1975> P« 25-33)- The presence of warp loops at both selvedges of some fab r i c s (Engel, 1963» P» 27; B i r d , 1951-2, p. 76) shows that the preparation of a set of p a r a l l e l elements by continuous warping was a usual practice at these s i t e s . S t r i c t main-tenance of warp order i s amply demonstrated by the elaborate patterns produced by transposing warps i n the weft-twined fabrics (Bird, I963)• The use of loom bars to support and tension the warp of weft-twined f a -b r i c s at Huaca P r i e t a ( F i g . 7) has been convincingly hypothesized by B i r d (1951-2, p. 76) on the basis of the large size of the f a b r i c s and the looped warp ends. P l a i n weave f a b r i c s are also found at preceramic s i t e s , though infrequently compared to twined ones (Bird, 1951-2, p« 76, 1952, p. 4-5; Engel, I 9 6 3 . p. 43-4 and F i g . 97) ' Unlike twining, plain-weave separates the warp into two equal groups made up of alternate warps, with a shed or opening between them. Doing sprang requires the separa-t i o n of a warp by making a shed, for without i t , any interruption i n work would allow the f a b r i c to undo i f a shed retainer (cord, hand, st i c k ) did not lock the l a s t warp movements. In p l a i n weave, the f i n a l weft retains the shed. The l e v e l of technical development needed for 20 F i g . 7 Three bar loom with a false c i r c u l a r warp (constructed after B i r d , 1951-2, p. 76) sprang - a set of continuous elements of equal length and s u f f i c i e n t strength, a support and tension system and the separation of the warp by a shed - are present i n the twined and p l a i n weave fabrics from pre-ceramic s i t e s . Several Huaca P r i e t a fabrics exhibit some in t r i g u i n g s t r u c t u r a l s i m i l a r i t i e s to sprang f a b r i c s . The warps of three weft-twined pieces, diagrammed by Junius B i r d , t r a v e l i n h e l i c a l paths ( F i g . 8), unlike the more usual cross and re-cross paths. A h e l i x i s the warp path i n i n t e r -l i n k e d sprang. Other diagrams show the l i n k i n g of adjacent warps be-tween the rows of spaced weft-twining ( F i g . 9)• The l i n k i n g i s repeated but reversed i n direction between the next rows of weft-twining. This reciprocal movement i s necessary when a warp i s f i x e d at both ends i n order to move the threads back to t h e i r o r i g i n a l position and prevent a disordering of the warps at the terminal area. Even i f t h i s i s a mis-take as Junius B i r d believes i t i s ( p . c ) , the correction of i t i n the succeeding row exhibits an understanding of the duplicating action ( i n reverse) of a twist when warp ends are f i x e d . The mirror-imaged du p l i -cation of warp movements on a set of fixed-end elements i s the major charac t e r i s t i c of the sprang technique. The orderly r e p e t i t i o n of the i n i t i a l mistake, without the correction, and the omission of twined wefts would r e s u l t i n i n t e r l i n k e d sprang. Although speculative, i t does appear that the l e v e l of technical development exhibited i n other fabrics i s s u f f i c i e n t for the advent of sprang i n the Preceramic Period and that the invention of sprang only awaits a fortuitous accident ... but how long a wait? I t i s clear that sprang did not become the major fabric making technique at any point i n the history of Peru, even with i t s advantage F i g . 8 Diagrams of weft-twined f a b r i c s from Huaca P r i e t a having h e l i c a l warp paths (after B i r d , 1968, Figs. 32 , 22 and 91) F i g . 9 Diagrams of weft-twined f a b r i c s from Huaca P r i e t a with l i n k i n g warps (after B i r d , 1968, Figs. 41, 51) of simultaneously producing f a b r i c at both ends of a warp. During the Preceramic Period, weft-twining, knotting and looping, which are t o t a l l y hand-manipulated, were the major techniques. In the I n i t i a l Period, weaving progressively became the ascendent technique, perhaps evolving out of twining (Conklin, 1975°) • The entrenchment of weaving as the major technique throughout the rest of Peruvian pre-history may have re-sulted from the invention of shed controls to move warps mechanically. The archaeological record i s not complete enough to know when the shed-ding appliance of the heddle came i n to use. The best estimate i s that i t may have been i n use during the Chavinoid phase of the Early Horizon, when large, evenly battened f a b r i c s , l i k e the Garhua fa b r i c s , were being made (Conklin, 1978, p. 3)• The sprang technique requires a lesser l e v e l of technical development than the revolutionary method of weaving with heddles, yet archaeologically, i t i s not represented e a r l i e r (aside from the questionable Asia report) . I t i s possible sprang developed before weaving, but for reasons embedded i n the culture, i t remained a minor and, at present, unverified technique. During the 1350 years when i t was quite d e f i n i t e l y i n use, sprang was always a minor technique, used for s p e c i f i c types of f a b r i c s . Though sprang has not been correctly recorded from.Preceramic and I n i t i a l Period s i t e s , comparison to the technical l e v e l exhibited by other fabrics from these time periods suggests that the sprang technique could have developed from the Preceramic Period onward. We may never know when i t came into use because i t might have originated i n areas where there i s l i t t l e or no fab r i c preservation, l i k e the highlands. This lengthy speculation on the o r i g i n of the sprang technique has been pursued because, among the e a r l i e s t recorded examples, are very complex and competent fa b r i c s which do seem to indicate a substantial develop-ment period. CHAPTER I I I THE PARACAS-NASCA TRADITION OF THE EARLY HORIZON AND EARLY INTERMEDIATE PERIOD In t e r l i n k i n g ; C e r r i l l o s At present, the e a r l i e s t , correctly i d e n t i f i e d sprang f a b r i c from a controlled excavation i s a long narrow s t r i p , possibly a belt (Wallace, 1962, p. 311)' Dwight Wallace excavated the piece at C e r r i l l o s , l e a Valley and dates i t to the I s l a Phase or approximately Ocucaje 7 (upper valley ) i n the chronology of Menzel, Rowe and Dawson (1964), c. 800-700 B.C. The dimensions are 84 cm by 2.5 cm and there are warp loops at both ends. The threads are three ply cotton, Z spun and S p l i e d . The structure i s almost a l l i n t e r l i n k i n g , t i g h t l y beaten. A single row of in t e r l a c i n g (threads pass over two, under two) appears 11 cm on either side of the central, terminal area ( F i g . 10A). The terminal area ( F i g . 10B) i s interlooped (p.c. Dwight Wallace, 2l/l2/79). The i n t e r l i n k e d structure connects adjacent threads i n an elbow-l i k e l i n k throughout. The only change i n t h i s regular structure i s i n the direction of the h e l i c a l path of each thread at points 11 cm from the central area. At that point, the S twist i n t e r l i n k i n g switches to Z twist i n t e r l i n k i n g , ( i n the mirror-imaged h a l f of the f a b r i c , the switch i s from Z to S twist) . The change of direction i n i n t e r l i n k i n g takes place along the row of 2:2 i n t e r l a c i n g . Such a change of twist d i r e c t i o n w i l l make a fa b r i c l i e f l a t t e r by reversing the torque imparted to the over-all f a b r i c , but i t i s d i f f i c u l t to t e l l i f t h i s was intention-a l l y done. I t could have resulted from turning the loom end f o r end and repeating one row of the two row sequence used i n the rest of the f a b r i c , B. F i g . 10 Technical d e t a i l s of the e a r l i e s t confirmed sprang f a b r i c , G e r r i l l o s , Early Horizon 7 A. Structure: i n t e r l i n k i n g and i n t e r l a c i n g B. Chained closure at the centre C. Warp loops at both ends (After sketches supplied by Dwight Wallace, 2l/l2/79) perhaps after an interruption i n work. Several insights into the method of working are preserved i n the technical data supplied by Wallace. The warp loops at both ends show a continuously wound warp was prepared. The fabric has no heading cords at either end, nor are the warp loops enlarged, as sketched by Wallace (Fi g . 1 0 c ) . I t appears that t h i s piece was not mounted on bars, either d i r e c t l y through the loops, l i k e the twined fabrics of Huaca P r i e t a , or by lashing a heading cord area to bars, l i k e the majority of woven and sprang f a b r i c s . As the f a b r i c has only sixteen threads and i s l e s s than a metre long, i t i s quite possible that a temporary cord was passed through the warp loops at both ends and attached to fixe d points for tensioning. In 1978, I watched an Ollantaytambo woman weave on a warp prepared i n t h i s way and tensioned between her big toe and her waist. Because of the l i m i -ted size of the C e r r i l l o s piece, special tools or equipment would not be needed. There i s good evidence i n the single row of int e r l a c i n g that the fab r i c was worked i n horizontal rows against a transverse fa b r i c f e l l or working edge. This method requires establishing a shed i n the warp, even i f only with the hands or a cord (see Collingwood, 1974, pp. 58-64 for a f u l l description of a working method). This point i s raised because other narrow sprang fabrics i n t h i s study are worked along oblique fabric f e l l s . ' The difference i n the orientation of the f a b r i c f e l l may indicate separate developments f o r these two variant techniques. The technique of the C e r r i l l o s sprang f a b r i c with the transverse f e l l has more i n common with the techniques of p'receramic twining and weaving where the fabric i s b u i l t up i n horizontal rows on a set of longitudinal elements. The technique with oblique working edges, found i n the straps of Paracas-Nasca s k i r t s , may have evolved out of braiding on free ends, where the fabric f e l l i s almost always oblique. I n t e r l i n k i n g may be the e a r l i e s t f a b r i c structure made i n the sprang technique i n Peru. Non-sprang, linked and looped fa b r i c s are numerous i n preceramic s i t e s , l i k e Asia (Engel, 1963) which indicates a use or appre-c i a t i o n f o r the stretchy, expandable qu a l i t y shared by those single ele-ment fabrics and in t e r l i n k e d sprang f a b r i c s . The impetus for developing the technique for i n t e r l i n k e d sprang may have arisen from an attempt to speed up the slower method of working with a single element. Although i n t e r l i n k e d sprang fabrics resemble single element, lin k e d fabrics most closely, the technical requirements and the probable evolution of interna, linked sprang share more with twining and weaving i n the e a r l i e r periods. I n t e r l i n k i n g ! Openwork Sample The e a r l i e s t extensive group of fa b r i c s produced i n the sprang tech-nique are: (interlinked headdresses and bags from the South Coast area. Some of the pieces have provenience. The Paracas s i t e s of Paracas Cavernas and Cabeza Larga and the Hacienda Ocucaje i n the lea Valley have yielded i n t e r l i n k e d sprang. Most of these technically homogenous fabrics have d i s t i n c t i v e iconography which supports the inclusion of other fabrics that are without provenience. Grave association and s t y l i s t i c comparisons to objects outside the grouping indicate these fabrics are primarily from Early Horizon 9 with a few examples that are t r a n s i t i o n a l to 10 (c. 600-450 B.C.). The workmanship and the patterning complexity of t h i s group do indicate a substantial developmental period preceded them. However, t h i s has not yet been borne out archaeologically and, so f a r , the use of the sprang technique for t h i s type of patterned i n t e r l i n k i n g i s confined to the lat e stages of the Early Horizon with no examples from outside the South Coast area. This i s a period where a new and a l i e n imagery appears on the South Coast. I t s origins are unknown but are sus-pected of l y i n g i n an intermountain v a l l e y (Sawyer, 1966, p. 87) . I f t h i s i s the case, t e x t i l e preservation might be so poor that the d i r e c t prece-dents of the i n t e r l i n k e d sprang hoods w i l l never be known. Most of the fabrics i n t h i s group appear to be headdresses, either long, f l a t turbans or hoods. Several pieces of diff e r e n t sizes and con-structional features may be small bags or mantles. The structure of the fabr i c s i s 1:1 i n t e r l i n k i n g combined with "holes" or omitted l i n k s (Fig.11). The patterning depends solely on the contrast between the regularly worked, dense cloth and the openwork created when l i n k s are omitted. No other structure or variant i s introduced except the chained hold l i n e which secures the threads i n the centre of each complete piece and the heading areas at either end which are generally strengthened with three to seven wefts inserted i n p l a i n weave sheds (F i g . 12). The chained hold l i n e and the r e f l e c t i v e symmetry of design and structure on either side of the hold l i n e i s the proof that sprang was the technique. The designs are all-over patterns with approximately equal dense and open areas. Some of the patterns interlock and repeat exactly while others repeat variations on a simi l a r theme. A l l the fabrics are mono-chromatic: dark blue, orange, maroon, green, gold, off-white and l i g h t blue pieces are present. The majority use deeply dyed wool, probably alpaca, while a few cotton ones are either natural coloured or pale i n hue. The two best documented pieces comes from the Paracas Cavernas s i t e on Cerro Colorado, Terrace I I I , Tomb 2 and 3« These excavations were made under the direc t i o n of Dr. Luis Valcarcel of the Museo Nacional, Lima i n 1931 and the t e x t i l e finds were published the following year 31 F i g . 11 Structure of South Coast headdresses: I n t e r l i n k i n g with "holes", or l i n k s omitted F i g . 12 General features of Early Horizon interlinked sprang A. Chained hold line B. Heading area with wefts (AMNH 41.2/5346) (Yacovleff and Muelle, 1932). MNAA #8430 i s an orange, wool oblong of of i n t e r l i n k e d sprang from Tomb 2 which has dimensions of 1.35 M x 0.60 M (p. 36). A detailed photograph shows the structure and the chained hold l i n e and a drawing shows the pattern of a serpentine g r i d with birds (?) i n the i n t e r s t i c e s ( F i g . 13). MNAA #8509 from Tomb 3 i s also orange, wool i n t e r l i n k e d sprang but i t i s longer, 2.20 M x 0.40 M (p. 44) . The design i s a branching serpentine repeat (Fig. 14). Both #8430 and #8509 were found on heads that were separated from torsos. Unlike the t e x t i l e s that were wrapped around the torsos, the sprang fabrics were ca r e f u l l y arranged on the head as they might have been worn i n l i f e : "Gontrarimente a esto, l a s cabezas, hasta en l o s casos de hallarse separados del tronco, conservaban su tocado cuida-dosemente puesto y en forma que s i n duda corresponde a su man- • era de usar." (p. 50, Yacovleff and Muelle, 1932) The two headdresses are described as being i n the form of hammocks, f l a t but with gathered ends (p. 52) . Another openwork headdress i n a non-sprang technique i s used to i l l u s t r a t e the manner i n which hammock-shaped headdresses of Tomb 2 and 3 were arranged,on heads (Fig. 15)-In addition to the two well-documented turbans i n F i g . 13 and 14, Nobuko K a j i t a n i has supplied pictures of four other i n t e r l i n k e d sprang fa b r i c s i n the Museo Nacional de Antropologia y Arqueologia, Lima. One, at le a s t , i s from an excavation by J u l i o C. T e l l o as i t s t i l l re-tains a number (12/9030) from h i s numbering system, i n addition to i t s current number, Lima 1091. I t i s long and narrow with the ends gathered through loops of the heading cord ( F i g . 16) i n a manner similar to that described for the two previous headdresses. The dimensions are 1.8? M x 0.99 M (p.c. Nobuko K a j i t a n i ) . Although t h i s piece i s narrower, i t seems l i k e l y that i t , too, was fashioned into a hammock shape and wound onto a head. I t i s possible that t h i s turban came from Tello's 1925-6 34 A F i g . 13 Interlinked sprang turban, Paracas Gavernas, Terrace III, Tomb 2, MNAA #84-30 A. Detail of terminal area (d'Harcourt, 1962, PI- 56B) B. Extension of pattern repeat (after Yacovleff and Muelle, 1932, Fig. 10) 35 F i g . 4 4 Pattern repeat on i n t e r l i n k e d sprang turban, Paracas Cavernas, Terrace I I I , Tomb 3 , MNAA #8509 36 F i g . 15 Arrangement of headdresses on truncated heads i n Tombs 2 and 3, Terrace I I I , Paracas Gavernas (after F i g . 23, Yacovleff and Muelle, 1932) 37 F i g . 16 D e t a i l of a Paracas turban showing gathering loops (MNAA #1091, photo courtesy of Nobuko K a j i t a n i ) work at the Paracas Gavernas s i t e . The design of the turban repeats double headed serpents inside a complex g r i d ( F i g . 17) ' Two very deteriorated sprang fragments of a single piece with inde-terminate dimensions are also i n the Museo Nacional ( F i g . 18). They have a serpentine g r i d design, s i m i l a r i n organization to the one i n F i g . 13• The o r i g i n of t h i s piece i s unknown but, s t y l i s t i c a l l y and tech n i c a l l y , i t rela t e s closely to fabrics known to have come from the Paracas Gavernas s i t e . This piece seems to be wider than the others mentioned and so may not have been used as a turban although there does seem to be fragments of a looped heading cord and gathering folds at one end. The other two pieces i n the Museo Nacional, Lima for which Nobuko K a j i t a n i supplied information have di f f e r e n t constructional d e t a i l s which suggest they might have been bags. One has complete warp selvedges and a chained hold l i n e . I t i s constructed by fol d i n g across the chained l i n e and sewing the side selvedges together ( F i g . 19) • The design i s horizontal zigzags and the dimensions are 0.12 (x2) M x ).24 M. The other fragment i s also wide i n proportion to i t s length. The seam i n the centre, which joins side selvedges, suggests i t may also have been a bag rather than a turban ( F i g . 20). The design i s a l i t t l e i r r e g u l a r but i t i s also a branching -serpent design, similar to F i g . 14. A long rectangular piece of i n t e r l i n k e d sprang was on display at the J u l i o G. Te l l o Museo, a small museum at the foot of Cerro Colorado,', the h i l l that contains the Paracas Cavernas s i t e : ( F i g . 21). I t i s said to come from the s i t e of Cabeza Larga, a shore s i t e d i r e c t l y by the museum (p.c. Guardian, Te l l o Museo). The r e l a t i v e dimensions of the piece suggest i t might have been another turban, although there are no heading cord loops i n i t s present, post-conservation state. The 39 F i g . 17 D e t a i l of a Paracas turban showing the pattern repeat of a double-headed serpent i n a complex g r i d (MNAA #1091, photo courtesy of Nobuko Ka j i t a n i ) Fig. 18 Two fragments of the same Paracas turban (?) with a serpentine grid design (MNAA Lima 98 and 112, photo courtesy of Nobuko Kajitani) 19 Interlinked sprang tag (?) with zigzag pattern (MNAA Lima 104-, photo courtesy of Nobuko K a j i t a n i ) Interlinked sprang bag (?) with branching serpent design (MNAA Lima 106, photo courtesy of Nobuko Kajitani) Interlinked sprang turban (?) from Cabeza Larg Paracas with branching serpent design (Julio G Museo, no number) branching serpent design d i f f e r s only s l i g h t l y from the one i n F i g . 14. The Ohara Art Gallery i n Kobe, Japan, also has a rectangular piece of int e r l i n k e d sprang from Peru ( F i g . 22) . I t appears to be a turban be-cause of the gathering loops at one end. The dimensions are 1.44 M x 0.45 M (p.c. Nobuko K a j i t a n i ) . Although t h i s piece i s without provenience i t i s closely associated technically and s t y l i s t i c a l l y with the Paracas Gavernas turbans. The design i s a serpentine g r i d having some si m i l a r -i t y to F i g . 13, although, i f there i s a figure i n the i n t e r s t i c e s of the g r i d as i n F i g . 13. i t i s not c l e a r l y discernable i n the photograph. The design i s comparatively large scale. Another group of i n t e r l i n k e d sprang fabrics have come from the l e a Valley s i t e of Ocucaje. M.E. King has reported on a group of "hoods", or head coverings, found on mummies at t h i s s i t e ( I 9 6 5 t p- 304). The sprang hood, which has not been reported from any other s i t e , i s t y p i c a l l y con-structed from one long piece of fa b r i c which i s folded across the trans-verse center and sewn together along one side selvedge and along the warp selvedges. (This method of construction i s different from that of the bags i n F i g . 19 and 20, where both side selvedges are sewn and the warp selvedge i s l e f t unsewn.) On s i t e observers reported that the hoods were sometimes found over the heads and hanging down the backs of mummies (King, 1965» P- 304). Hoods from Ocucaje are made i n woven double-cloth and gauze weave as well as i n t e r l i n k e d sprang. The openwork gauze and sprang hoods are similar i n construction features, while the double cloth hoods are l e f t unseamed along the warp selvedges. There are three hoods, two complete and one fragmentary, of i n t e r -l i n k e d sprang i n the Ocucaje sample worked on by M.E. King. A maroon alpaca hood, with dimensions of O.67O M x O.265 M, has a motif of a F i g . 22 Interlinked sprang turban with serpentine g r i d design (OAG H5, photo by Mr. Tanaka, supplied by Nobuko Kaj i t a n i ) double headed serpent repeated inside a complex l i n e a r g r i d ( F i g . 23) • The motif i s i d e n t i c a l to one on the narrow turban i n F i g . 16 and to another on an openwork tunic of unspecified structure sketched by Rebecca Carrion-Cachot (1930, F i g . 4h) . There i s a s l i g h t error i n King's drawing of t h i s motif (I965, F i g . 78d), as these three openworks do have the same design. The Ocucaje hood has an added multi-coloured fr i n g e . The second complete hood i n King's thesis i s dark blue alpaca with dimensions of 0 .625 M x 0.335 M ( F i g . 24). The structures, construc-t i o n and fin i s h e s are the same as i n F i g . 23• The design i s f a r more elaborate, with three variants of the saucer-eyed "Oculate Being" (Menzel, Rowe and Dawson, 1964, pp. 171 -2 , I96-8 and 239-44). A fragmentary Ocucaje piece of l i g h t blue cotton i n t e r l i n k e d sprang has a design of branching serpents ( F i g . 25), similar to F i g . 14. A photograph of t h i s piece was requested but the museum was unable to com-ply because i t was too f r a g i l e . King i s uncertain whether t h i s f a b r i c was used or intended as a hood (1965, P- 378 ) . Even incomplete, the width measurement of 0.450 M i s more than the other hoods, Also,, i t i s made of cotton while the hoods are made of alpaca. There i s some s l i g h t reason to think i t may have been a mantle or shroud. Two i n t e r l i n k e d sprang hoods not included i n King's study also come from Ocucaje (p.c. Alan Sawyer) . F i g . 26 i s a maroon alpaca hood from the private c o l l e c t i o n of Paul Truel, one of the owners of the Hacienda Ocucaje, where the King sample originated. I t has a branching serpentine design similar to F i g . 14, 20 , 21 and 25- I t resembles the hoods i n King 1s sample except that i t does not have an added fringe. The other hood i s made of gold alpaca with a design of serpent heads set within 4 ? F i g . 23 Interlinked sprang hood from Ocucaje with design of double-headed serpents i n a complex g r i d (AMNH 41.2/5982, photo courtesy of Alan Sawyer) Fig. 24 Interlinked sprang hood from Ocucaje with "Oculate Being" imagery (TM 91.895, Textile Museum photo, courtesy of Ann Rowe) 49 F i g . 25 Branching serpentine design on a fragmentary i n t e r -linked sprang fa b r i c from Ocucaje (after King, 19^5, F i g . 78g), Interlinked sprang hood with branching serpentine design from the c o l l e c t i o n of Paul Truel, Ocucaje an i r r e g u l a r diamond g r i d ( P i g . 27). This hood i s i n the American Museum of Natural History but i t was a much l a t e r acquisition from Mr. Landman than the hoods i n the Ocucaje c o l l e c t i o n studied by M.E. King (p.c. Barbara Conklin, 29/l/82) . Mr. Landman acquired the hood from a New York dealer who regularly handled Peruvian f a b r i c s (p.c. Alan 'X • Sawyer) . The technical and constructional s i m i l a r i t i e s between these ,i two hoods and the others from Ocucaje that i t i s not surprising that they too originated there. Sprang hoods have,,:-so f a r , not been reported from other s i t e s . King's sample includes a fragment, #534, of simple l i n k i n g which she ten t a t i v e l y c a l l s a bag (p. 507). She i d e n t i f i e s the technique as sprang although she notes that no loom ends or centre area are present. I have not been able to study t h i s fragment but a sketch of the piece by Barbara Conklin and a photograph by-Alan Sawyer confirm King's tenta-t i v e position as to what i t i s and how i t was made. I t i s too fragmen--tary to have absolute proof of the sprang technique, yet i t s s i m i l a r i t y . to other complete pieces suggests King i s correct. According to M.E. King (I965, p. 238) and Jane Dwyer ( p . c ) , there are at least three more examples of in t e r l i n k e d sprang i n the Museo Nacional, Lima. King refers to MNAA 13/152 and 89-4-3 and Jane Dwyer showed me a s l i d e of MNAA 12/5486. None of these pieces, nor those that Nobuko K a j i t a n i supplied numbers f o r , could be located by s t a f f when I v i s i t e d the museum i n 1978 because they were i n the process of r e -housing t e x t i l e s i n a new storage f a c i l i t y . Jane Dwyer also had a sl i d e of an i n t e r l i n k e d sprang fabric i n the Truel c o l l e c t i o n (0-9425) but i t could not be located when I was shown that c o l l e c t i o n . MNAA 12/5486 and PT 0-9425 both have Oculate Being iconography, rather than 52 F i g . 2? I n t e r l i n k e d sprang hood with serpent head design (AMNH 41.2/5346, Landman) serpent-derived designs. MNAA 89-4-3 has an elaborate design of birds but King i s not sure the technique i s sprang although the structure i s that of the hoods (1965, P« 24-0) . King does not specify the iconography of MNAA 13/152. The sample of i n t e r l i n k e d sprang fabrics located for t h i s study i s made up primarily of headcloths used i n b u r i a l s . Hammock-shaped turbans are associated with Paracas b u r i a l s while hoods are associated with Ocucaje b u r i a l s . No examples run contrary to t h i s , although some turbans are without secure provenience. Small i n t e r l i n k e d sprang bags are also associated with Paracas and Ocucaje b u r i a l s . Several fabrics are of indeterminate use. Iconography of Interlinked Sprang Openworks The iconography of the i n t e r l i n k e d sprang f a b r i c s f a l l s into several broad categories. The group that uses repeating serpentine patterns i s the largest. Geometric patterns and renditions of the Oculate Being are l e s s numerous. The serpentine patterns can be further grouped. The branching ser-pent design occurs on f i v e f a b r i c s , including hoods, turbans and one possible mantle, from Ocucaje and Paracas s i t e s ( F i g . 28). Minor varia-tions i n the i n t e r i o r open areas and s p a c e - f i l l i n g shapes i n front of the mouth occur within the group._ A l l are depicted with two eyes, a rhomboidal or obtuse-angled head and smooth (non-serrated) outlines i n a configurative arrangement that uses s l i d e symmetry (Shepard, 1948). This pattern appear on other f a b r i c s , l i k e an Early Horizon 9 poncho s h i r t from Ocucaje which also has horizontal bands of twisted elements. The branching serpent design appears to derive from 54 e F i g . 28 Five variations of the branching serpent design on i n t e r l i n k e d sprang fabrics A. Ocucaje (King, 1965, F i g . 78g, #66) B. Ocucaje, c o l l e c t i o n of Paul Truel G. Paracas Gavernas, MNAA #8509 D. Gabeza Larga, JGTM, no number E. MNAA Lima 106. depictions of the three strand braid as a diagrammatic comparison i l l u s -trates (Fig. 29). A further example of an e a r l i e r , painted f a b r i c i l l u s t r a t e s another three strand braid image (F i g . 30). In t h i s instance, the serpent belt of a major Ghavinoid deity i s represented as braided. The interlaced image i s cl e a r l y r e lated to serpentine bodies. There are numerous other examples i n Early Horizon Chavinoid art of fabric structure imagery - l i n k s , braids, twisted strands - which have been grouped together under the single term 'guilloche' ( j . Rowe, I962). In Ghavinoid a r t , the images are cursively rendered and more readable be-cause they are painted on fabric or carved into a surface. The Early Horizon 9 f a b r i c s , l i k e those from Paracas Gavernas and Ocucaje, gener-a l l y are patterned through structural manipulations during fabrication and are therefore more geometric and less easy to read. A single hood (AMNH 41.2/5346) has a serpent head i n a different configurative pattern. I t has the same obtuse-angled head and smooth outline but i t i s repeated i n translation symmetry within the i n t e r -s t i c e s of a roughly diamond g r i d . I t may be s i g n i f i c a n t that the thick-ness of the g r i d l i n e s vary. Because i t would have been easier to make an u t t e r l y regular diamond g r i d i n t h i s structural technique, i t i s sug-gested that the varying thickness of the g r i d l i n e s i s purposeful, that i t may be an attempt to represent another f a b r i c image. In F i g . 31, the diagram shows the relationship between the hood design and the image of a linked structure. Images of l i n k i n g occur on e a r l i e r Ghavinoid f a -b r i c s and on Early Horizon 7 tapestry fabrics from Ghucho (N. Dwyer, 1979) so the survival of the image into Early Horizon 9 i s possible, even expected ( F i g . 32). There are three examples of a broken g r i d pattern where the l i n e s of the g r i d are the bodies of double-headed serpents ( F i g . 33)- Of the F i g . 29 Branching serpent designs and three strand braids A. Discontinuous warp poncho s h i r t with branching serpent design, Ocucaje TM 1959-H-l (photo courtesy of Alan R. Sawyer) B. Derivation of the branching serpent design from the three strand braid Braided serpent belts on a Chavinoid painted t e x t i l e Phase D (Sawyer archive photo, o r i g i n a l l y from Peter Roe; t e x t i l e on loan to Amano Museo, Lima) F i g . 31 Possible derivation of serpent design from a linked structure (design from hood, AMNH 41.2/5346) 0 59 F i g . 32 Images of linked structures on other f a b r i c s A. Painted Ghavinoid f a b r i c , Phase D (photo by A. Rosenschweig, courtesy of A . R . Sawyer) B. Ghucho tapestry, Early Horizon 7 (N. Dwyer, 1979, F i g . 8 ) 60 F i g . 33 Serpentine grids on int e r l i n k e d sprang openworks A. MNAA Lima 98 and 112 B. MNAA #8430 C. OAG H5 three of t h i s type, one example has an obtus-e-angled head and a smooth outline ( A ) . The second one has an obtuse-angled head and a jagged bor-der of small triangles (B) . The t h i r d has an acute-angled head and a serrated outline ( c ) . One of the three, (B), c l e a r l y has another design component (bird?) i n the i n t e r s t i c e s of the g r i d . The other two may also have a s i m i l a r component but the photographs are not clear enough to deter-mine t h i s . I t seems probable that t h i s elegant repeat derives from images of adjacent (and touching) pairs of twisting strands, such as those seen i n a tapestry woven f a b r i c ( F i g . 3^ -A) . The relationship between the t a -pestry version of adjacent twisted strands and the broken g r i d of serpent bodies on the sprang f a b r i c s i s clearer when colour contrast i s eliminated and when the design i s angularized ( F i g . 34-B) . The angularization of the image may be a direct r e s u l t of producing i t by structural patterning. Another broken g r i d pattern besides t h i s one., appears frequently on gauze weave fabrics of t h i s period. I t i s mentioned i n passing because i t s derivation i s very close to the one i n F i g . 34. I t appears that i t also originated i n images of adjacent and touching twisted strands, but strands that alternate i n twist d i r e c t i o n ( F i g . 35) • Interlinked sprang fabrics which were not located f o r t h i s study could very possibly have t h i s pattern as i t appears so frequently on contemporary gauzes. Two examples i n the i n t e r l i n k e d sprang sample have full-bodied, double-headed serpents arranged within a complex g r i d . The heads are obtuse-angled and the outlines are smooth. Though one of these two i s on a narrow f a b r i c and only part of the repeat i s used, a comparison of d e t a i l s shows they are the same pattern. An openwork tunic of unspeci-f i e d structure from Paracas Gavernas also has the same design (Carrion) Cachot, 1931» F i g * 4h). This serpent configuration may be related to images of twisted strands, such as those i n the horizontal bands of the 62 B F i g . 34 Twisted (or twined?) strand image A. c u r v i l i n e a r image on tapestry (photo courtesy of Alan R. Sawyer, o r i g i n a l l y taken by J.B. Bird) B. Derivation of angular g r i d from twisted strands ( a l l S twist) B F i g * 35 Twisted (or twined?) strand image A. Diagram of gauze fabric design with underlying g r i d (O'Neale, 1948, F i g . ?a,b) B. Derivation of angular g r i d from twisted strands (alternating S and Z twist) poncho s h i r t i n F i g . 29- The heads do not overlap and interlock, yet there i s a tenuous relationship to images of twisted strands ( F i g . 36). Besides serpentine patterns, the i n t e r l i n k e d sprang fabrics also have geometric patterns. A small bag i n F i g . 19 simply repeats horizon-t a l zigzags. Although i t does not have overt serpentine c h a r a c t e r i s t i c s , there i s room to suggest that the zigzag i s a more abstract rendering of the undulating serpent body. Serpentine patterns predominate on the i n t e r l i n k e d sprang fabrics and they appear to derive from images of fabric structures. The coupling of serpents and f a b r i c structure images occurs i n the Ghavinoid style on gourds, goldwork, stone and ceramics as well as on fabrics ('J. Rowe, 1962, 1967; P. Roe, 1964-) . Only a few examples on fabrics have been used to i l l u s t r a t e the connection between the more geometric versions on the sprang headdresses and the e a r l i e r , more cursive examples. Although some of the images found i n Ghavinoid art continue, there are several innovations i n style i n the Early Horizon 9 fabrics i n t h i s study. The serpent head i s generally rhomboidal and shown from a dorsal point of view with a s l o t - l i k e mouth. The angularity of serpent motifs and the use of structural patterning d i f f e r s from the c u r v i l i n e a r Ghavinoid serpents. The designs tend to be repeated regularly over a f i e l d i n Early Horizon 9 sprang i n contrast to the more varied and smaller scale use of the serpentine designs i n Ghavinoid examples. The use of multiple serpent heads i n the Early Horizon 9 fabrics i s innovative. In Ghavinoid a r t , the serpentine body i s emphasized, often appearing i n f a b r i c structure images without a head or with only a single head, which may combine f e l i n e and serpent a t t r i b u t e s . In Early Horizon 9 sprang f a b r i c s , serpent heads are used as a device that signals depth or overlap. When two or three strands twist or braid, strands must move behind 65 F i g . 36 Double-headed serpents within a complex g r i d A. P a r t i a l repeat from a narrow sprang fa b r i c (MNAA Lima 106) B. Same design, repeating f u l l y (AMNH M .2/5982) G . Distant relationship with angular version of twisted strand image other strands to make an accurate representation of the structure i n two dimensions. In Figs. 28 and 33» the serpent head i s consistently capped on any strand moving "behind another strand. These s t y l i s t i c innovations suggest that the iconography of the sprang headdresses may not evolve d i r e c t l y out of Ghavinoid precedents. The sprang fabrics are associated with t e x t i l e s and pottery that have considerable non-Ghavinoid iconography. The south coast iconography, referred to as trophy head cult by Alan Sawyer (1966, p. I l l ) , includes the Oculate Being, trophy heads, knives and various jungle creatures l i k e parrots and monkeys as well as the more geometric serpent with the slot-mouth. The trophy head cult imagery may have been imported to the south coast from the i n t e r i o r but i t , too, appears to have some Ghavinoid roots, especially i n the serpent-fabric structure imagery. One i n t e r l i n k e d sprang hood has representations of the Oculate Being. The Oculate Being f i r s t appears i n Early Horizon 8 (Menzel, Rowe and Dawson, 196k) and the saucer-eyed smiling being seems to replace the f i e r c e l y fanged deity of Ghavinoid a r t . I t i s interesting that the Oculate Being shares serpentine a t t r i b u t e s (hair, belt) with the old Ghavinoid gods but sprouts new ones too, as appendages to the face and body. The iconography of the i n t e r l i n k e d sprang f a b r i c s i s a mixture of secondary Ghavin-derived images of serpents and fab r i c structures and a new major mythical being. The predominance of serpent imagery on a sample that i s mostly headdresses may relate to the metaphoric treatment of hai r as serpents, a practice that goes back at least to £he'~"time~' of Chavin>(j. Rowe, I962) . 67 Dating of Interlinked Sprang Openworks Several of the headdresses i n t h i s cohesive group of i n t e r l i n k e d sprang fabrics have a Paracas Gavernas provenience (MNAA #84-30 and 8509, Yacovleff and Muelle, 1932). The Gavernas material, including the struc-t u r a l l y patterned sprang headdresses, correspond to Epoch 9 of the Early Horizon, c. 600-500 B.C. Jane Dwyer, who has made the most detailed study of the Paracas t e x t i l e s , c i t e s stratigraphic evidence and ceramic asso-c i a t i o n to support her placement of the Cavernas material i n Epoch 9 (1979, P- 107). On iconographical and technical grounds, the i n t e r -linked sprang from other s i t e s or unknown s i t e s appears to be contempor-aneous . The sample of i n t e r l i n k e d sprang i s extensive enough to attempt at least some inte r n a l sequencing. In grouping the serpentine patterns f o r the iconography discussion, the features of head shape (obtuse-angled or acute-angled) and outline (smooth or jagged border of small triangles) have been noted. These features are Isolated by Lawrence Dawson i n h i s s t y l i s t i c s e r i a t i o n of painted mummy masks from Early Horizon 9 and 10 (1979). In h i s t r a i t table ( F i g . 37), the obtuse-angled head on ser-pents i s an e a r l i e r t r a i t than the acute-angled head. The jagged border of small triangles on serpent bodies i s a l a t e t r a i t . I f the s t y l i s t i c evolution of s t r u c t u r a l l y patterned f a b r i c s , l i k e i n t e r l i n k e d sprang, p a r a l l e l s the evolution i n painted f a b r i c s , then these t r a i t s can be used f o r ordering t h i s group. Most of the serpent patterned fabrics have the conservative Epoch 9 t r a i t s of the obtuse-angled head and smooth outline - the f i v e i n F i g . 28, the two i n F i g . 36 as well as F i g . 31 and 33A. The single l a t e t r a i t of a jagged border of small triangles appears i n F i g . 33B while both l a t e t r a i t s , the jagged border and the acute-angled S E R I A T I O N OF F E A T U R E S OF C L O T H M U M M Y M A S K S Derived from associations on individual cloths Features Eyes of face framed and shared by two profile felines Dotted circle spots on flanking felines Reserve background used as outline Eccentric pupil in eye Step blocks above and below eyes Nose isolated, not attached to top of face Broad red or purple border or framing band Face only, no body Face with projecting tongue • Obtuse angle at corner of serpent heads Serpent body segmented Facial features mainly in upper half of panel Tongue as long or longer than the nose Serpents looped around the eyes Row of vertical serpents on top of head Body drawn underneath the face Narrow red line borders the cloth Dentate line borders the cloth • Acute angle at corner of serpent heads ^ Small figure inside trunk of body Iv Jagged border of small triangles on serpents Serpent with human face Head with 6-sided outline, much smaller than width of panel (• Legs both turned same way in profile Ocucaje 10 Ocucaje g xxxxxxx xxxxxxx xxxxxxx xxxxxxxxxxxxxxx xxxxxxxxxxxxxxx XXXXXXXAJUUi-KXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX xxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx XXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXX F i g . 37 S t y l i s t i c features of painted mummy masks from lea (L. Dawson, 1979, P« 85) head, appear i n F i g . 3 3 0 • The majority of the i n t e r l i n k e d sprang f a b r i c s with serpentine patterns would then be placed near the beginning of Early Horizon 9 with those represented i n F i g . 33B & G f a l l i n g at the end of that phase. This agrees with Dawson': s dating of a gauze mummy wrapping, which resembles the advanced sprang examples of F i g . 33B & G, to the end of Early Horizon 9 -There i s no clear correlation between garment type or s i t e and con-servative/advanced features. The two advanced examples are turbans and one i s from Paracas Cavernas, the other i s without provenience. The conservative examples include a l l garment types and known s i t e s . One implication of t h i s sequencing which requires further study i s for the two Paracas Gavernas turbans with grave association. F i g . 33B from Tomb 2, Terrace I I I has a l a t e r s t y l i s t i c t r a i t than F i g . 28G from Tomb 3 , Terrace I I I . The single geometrically patterned bag might be considered to be-long with the conservative group of Early Horizon 9 sprang, i f the zigzag i s interpreted as a serpent body with a smooth outline. A photograph of one hood with Oculate Being imagery has been obtained for t h i s study. I have seen photographs of two others among slid e s of Jane Dwyer's but was unable to obtain copies of them for t h i s study. Dawson's ser i a t i o n of painted mummy masks i s o l a t e s many features of the Oculate Being. The design of the sprang hood from Ocucaje ( F i g . 3 8 ) has a number of the more advanced t r a i t s , such as the full-bodied represen-t a t i o n , the small figure inside the trunk of the body, the legs both turned the same way i n p r o f i l e , the acute-angled serpent heads and the jagged border on serpents (Fig. 3 7 ) - Conservative t r a i t s , l i k e obtuse-angled serpent heads and smooth serpent bodies, also appear on the hood. F i g . 38 Three versions of the Oculate Being on an inter-linked sprang hood from Ocucaje (TM 91.895) Dawson fe e l s that the mixture of advanced and conservative t r a i t s accord best with t r a n s i t i o n a l 9-10 (Early Horizon) developments i n the painted masks (1979, P- 101). This hood would then be the l a t e s t i n the t o t a l i n t e r l i n k e d sprang sample. I f features i s o l a t e d for the seriation of painted mummy masks apply to t h i s group, then i n t e r l i n k e d sprang appears to have been used most widely toward the beginning of Early Horizon 9* The conservative serpen-tine patterns are most numerous while the serpent patterns and Oculate Being with more advanced t r a i t s are less numerous. This may indicate a diminishing use of in t e r l i n k e d sprang for openwork fab r i c s toward the end of Early Horizon 9 and t r a n s i t i o n a l to 10, i f t h i s sample i s repre-sentative . Another openwork technique, a combination of gauze and p l a i n weave, i s used for headdresses l i k e the i n t e r l i n k e d sprang ones as well as for many larger cloths that might have been mantles or mummy wrappings (Fig . 39)- The range of iconography on the gauzes i s si m i l a r to that on the i n t e r l i n k e d sprang f a b r i c s . There are conservative Phase 9 serpentine patterns ( j . Dwyer, 1979» Fig * l ) and advanced Phase 9 serpentine patterns (O'Neale, 1942, F i g . 11 and P I . 1; Garrionrpachot, 1931, F i g . 2; Dawson, 1979, F i g . 16). Other gauzes have Oculate Being representations which are t r a n s i t i o n a l ' t o Early Horizon 10 (O'Neale, 1942, F i g . 6; King, 19°5» F i g . 27) and others which are more advanced i n 10 (King, 19^ 5, F i g . j49; B i r d and Bellinger, 1954, PI. 71; two unnumbered pieces from the Truel c o l l e c t i o n i n the Sawyer archive). While, the gauze weave variant ( F i g . 40) and i n t e r l i n k e d sprang are obviously used concurrently, the conservative sprang f a b r i c s are more numerous while the gauze weaves that are t r a n s i -t i o n a l to Phase 10 are more numerous. Also, there are many more large 72 F i g . 39 Gauze mantle from Paracas Cavernas: Oculate Being imagery t r a n s i t i o n a l to Early Horizon 10 (O'Neale, 194-2, F i g . 6) 73 F i g . ko ? n t e r l i n t i l a l n ™™ S t r u c t u r e contemporaneous with interlinked sprang openworks size gauze -weaves. I t appears that i n t e r l i n k e d sprang may be gradually displaced by gauze weaving. Interlinked sprang openworks do not appear i n t h i s sample af t e r the beginning of Early Horizon 10, while t h i s par-t i c u l a r gauze weave continues to be used i n the Early Intermediate Period (O'Neale, 1948, p. 157 and P I . 22b). In extensive experimentation with both the gauze weave variant and i n t e r l i n k e d sprang, I found there was no speed advantage i n gauze weav-ing t h i s variant, even using a heddle to the f u l l extent possible. Sprang s t i l l kept pace by virtue of i t s duplicating action. The major advan-tage of t h i s gauze weave over i n t e r l i n k e d sprang may be that a greater length of warp can be conveniently worked, because the f a b r i c i s worked from one end only, unlike sprang which grows from both ends of the warp simultaneously. The increased emphasis on gauze weave may be related to a greater use of large openworks for mummy wrappings and a lessened use of smaller openworks f o r mummy headdresses. This conjecture i s supported by the greater number of large size gauze weaves that have been published, (an indicator that may not be a r e l i a b l e r e f l e c t i o n of the actual case)'. In any event, i n t e r l i n k e d sprang openworks do not reappear i n the course of Peruvian pre-history, according to my current research. This suggests that the technical i d e n t i f i c a t i o n of i n t e r l i n k e d sprang f a b r i c s with openwork patterns, that are 'orphans' i n Peruvian c o l l e c t i o n s , can date the f a b r i c s to Early Horizon 9, possibly t r a n s i t i o n a l to 10. I n t e r l i n k i n g : Striped Sample I have received information on two i n t e r l i n k e d sprang fabrics that are reputed to be from Paracas. Both are made of dyed yarns and t h e i r only patterning i s an arrangement of stripes i n the warp. 75 One that i s d e f i n i t e l y from Paracas i s i n the Peabody Museum, Harvard,^ (Fig. 4l). I t i s a small bag, 0.17 (x2) M x 0.14 M, and i t i s from a Paracas Necropolis mummy bundle given to the Peabody, Harvard by J u l i o C. Tello and Nelson Rockefeller and opened there. The cotton yarns are Z spun, 2 ply -s \ and the stripe arrangement alternates yellow between brown and green. Though i t i s disintegrated, the in t e r l i n k e d structure and evidence of the sprang technique, i n the form of a central hold l i n e , i s present (p.c. J i l l Mefford, 12/6/81). The garments from the. mummy bundle have both the geometric style and the more c u r v i l i n e a r , "'thick l i n e style of embroidery (p.c. J i l l Mefford, 18/7/82). The style of the garments spans Early Horizon 10 and Early Intermediate Period 1, using Jane Dwyer's s t y l i s t i c s e riation (1971, 1979). This sit u a t i o n m&y indicate that styles of embroidery are more overlapping than the Dwyer study suggested. Some of the garments from the Tello-Rockefeller mummy bundle are published i n W.G. Bennett, 1954, The second example of striped, i n t e r l i n k e d sprang i s i n a private c o l l e c t i o n i n New York (p.c. Daniel Rifken, 29/10/80). He believes i t i s a "Paracas hood". Aside from the s t r i p i n g , i t has no iconography. I f i t i s a hood, i t may be from Ocucaje, as hoods have not been reported from other s i t e s . I have not been able to study t h i s piece so f a r . P l a i n i n t e r l i n k e d sprang with stripes has not been reported i n the l i t e r a t u r e , due to chance and selec t i o n . I t i s quite possible that many more unprepossessing examples l i e unidentified i n museum c o l l e c t i o n s . The Peabody example which i s from a Paracas Necropolis b u r i a l i s the la t e s t occurrence of any kind of in t e r l i n k e d sprang that my research has turned up. Detail of a striped, interlinked sprang bag from Paracas Necropolis (PMH 38.28.30/4114; photo courtesy of Alan R. Sawyer) 77 Interlacing: Small Bags Two small bags made of 2:2 i n t e r l a c i n g were photographed at the Museo Nacional, Lima by Nobuko K a j i t a n i who supplied the pictures i n F i g . 4-2. The two bags r e t a i n Tello's o r i g i n a l numbering but the p r e f i x number 12, according to s t a f f at the Museo Nacional, was a number given to f a b r i c s from Paraoas and does not indicate a pa r t i c u l a r s i t e (p.c. Adriana S o l d i , 27/5/82). A sprang turban described e a r l i e r (Fig. 16) has a 12 p r e f i x to i t s number and i t relates closely to Paracas Cavernas sprang. The p o s s i b i l i t y that these bags are also from the Early Horizon 9 s i t e of Gavernas e x i s t s , but i t i s by no means certain. Both bags are made of dyed alpaca yarns which are arranged i n stripes of several colours. The diamond pattern on the bags i s a simple r e s u l t of obliquely moving threads i n one colour crossing d i f f e r e n t l y coloured threads. The bags have a centr a l l y positioned chained hold l i n e (p.c. Nobuko K a j i t a n i ) and the attendant features of a fabric worked i n the sprang technique. Warp loops are v i s i b l e i n undeteriorated areas around the bag mouth. Surface f l o a t s of the 2:2 interlaced structure l i n e up i n horizontal wales ( F i g . 4-3). A straight forward method of in t e r l a c i n g i n rows against a transverse f a b r i c f e l l by alternating two sequences of thread movements i s s u f f i c i e n t to produce t h i s kind of f a b r i c . Precise d e t a i l s of t h i s technique are given i n Collingwood (1974, p. 185-6) who also points out that i d e n t i c a l f a b r i c can be made i n sprang by repeating only one set of thread movements and reversing the frame or warp support a f t e r each row. The bags are constructed from a rectangle of fabric folded trans-versely i n the area of the hold l i n e and sewn together along the side selvedges. One bag, which i s f i l l e d with small b a l l s of dyed alpaca V Flg' 42 IW° ^ T 1 ^ 6 ^ ? ^ b a g s f r o r a P ^ a c a s A. MNAA Lima 167 B. MNAA Lima 173 (photos courtesy of Nobuko Kajitani) F i g . 4-3 Over two, under two ( 2 : 2 ) interlaced structure of the bags i n F i g . 4-2 yarn, appears to have a cord threaded through the undeteriorated warp loops at the bag mouth. The other bag has a sewn strap of: loosely t w i s t -ed and folded yarns. The presence of these modest l i t t l e bags, which have not been report-ed i n the l i t e r a t u r e , suggests the use of the- sprang technique may extend to other homely items. I t i s quite possible that sprang was used more widely than t h i s paper r e f l e c t s because the sample may be largely com-prised of more elaborate fabrics that have been selected for publication or museum ac q u i s i t i o n . The smallness of t h i s sample and the lack of association make i t d i f f i c u l t to even suggest the temporal extent of t h i s type of interlaced sprang. Interlacing: Bags from the Early Intermediate Period A group of four interlaced sprang bags ;which share many technical features have been located i n various museums. These bags are larger than those i n F i g . 42 and have a decorative seaming and edging s t i t c h . The same colours used i n the body of the bag are generally alternated i n the tubular seam or edging which i s worked i n the cross k n i t loop s t i t c h and the bottom' of the bag i s fringed. These interlaced sprang bags are considered to be roughly contemporaneous because of the finishes and the general form. The features are i l l u s t r a t e d i n the bag i n F i g . 44. The bags are constructed by folding across central unworked warps or a central hold l i n e and sewing the side selvedges and the terminal working edges together. A fringe i s made of the unworked warps or a separate fringe i s added to the bag bottom. Several have carrying straps and several have indications that a snugly wrapped cord near the mouth of the bag was used f o r a closure. Fig. 44 Interlaced sprang tag from the Early Intermediate Period (MNAA 1.3768) The evidence of sprang as the method of working varies from tag to bag. A l l bags are i d e n t i c a l on both faces i n interlaced structure and colour order. One bag has a symmetrical mistake on the two faces (MNAA 1.3768). One has a fringe with looped ends made of the unworked warps which s t i l l j o i n the two bag faces (LM 16.1104-3). One bag has a hold l i n e with symmetrical fa b r i c on either side (MDH 30.19.4-4-5). The fourth bag has no absolute proof of sprang as the technique (MRI, no number) as the hold l i n e area i s covered with s t i t c h i n g . However, i t s structure, form and colours r e l a t e i t so strongly to the group that i t seems u n l i k e l y i t was done i n any other technique. Two of the bags are rectangular products of sprang ( F i g . 44 and 45). The bag i n F i g . 44 has paired threads of black between pairs of a l l other colours. The pairs of threads act as a unit throughout and interlace over four, under four ( F i g . 46A). The second bag has wider diagonal stripes with dark brown alternating between orange, yellow and blue-green (F i g . 45A) . The structure of the bag i s over two, under two i n t e r l a c i n g . (For oblique interlaced sprang techniques, see Collingwood, 1974, p. 184-190.) The fringe i s made from the unworked central warps. The folded threads of the fringe are supertwisted, then r e - p l i e d with an adjacent thread by twisting i n the opposite d i r e c t i o n . The fringe i s lengthened by the addition of colour-matched threads, which are also twisted and r e - p l i e d , through the loops ( F i g . 45B) . The other two bags have a d i s t i n c t i v e shape, with three small pockets at the lo/irer edge. One of the bags (Fi g . 47, published i n d'Harcourt, I962, P I . 54a) has the same 4:4 interlaced structure and almost the same colour order as the bag i n F i g . 44. However, the shape of the bag re-quires a special warp set-up, l i k e the one shown i n F i g . 48. A warp with 83 F i g . 4-5 Interlaced sprang bag from Poroma, Tunga Valley (LH 16-11043) A. Overall view B. Fringe d e t a i l : fringe i s made of unworked sprang warps and lengthened by yarns added through loops (see arrow) 3 F i g . 46 Interlaced structures of sprang bags from the Early Intermediate Period A. Over four, under four (4:4) i n t e r l a c i n g (MNAA 1.3768, MDH 30.19.445 and MRI, no number) B. Over two, under two (2:2) i n t e r l a c i n g (LM 16.11043) Fig. 47 Interlaced sprang bag with three pendant pockets (MDH 30.19.445; photo taken from d'Harcourt, 1962, PI. 54) 86 F i g . 48 Discontinuous warp set-up necessary for the F i g . 47 bag. Brackets indicate separately worked sections and dashes indicate the terminal area of each section. "windows" i s necessary to produce t h i s i r r e g u l a r l y shaped tag. Two scaf-f o l d wefts act l i k e extra loom tars on which warps of different lengths are wound. This discontinuous warping technique has a long history i n Peru, from at least Early Horizon 8 (Garaventa, 1981, p. I67) up to the present. This i s the only example I have found of discontinuous warping which i s used f o r shaping a sprang f a b r i c . Because some warps do not ex-tend the entire length, i t had to be worked i n three sections, which are indicated by brackets i n F i g . 48. Three hold l i n e s , indicated by dashes, would be found, one i n the centre of each separately worked section. In the photograph, one hold l i n e i s v i s i b l e , i n the centre of the main body (F i g . 47). The colour movement i s , as expected, i n mirror-image symmetry on either side of the l i n e . A si m i l a r sized section on the lower bag face would have been worked as a section and would have s i m i l a r l y placed hold l i n e . The t h i r d section i s the pendant pockets. What i s now the bottom of the pockets would have been the terminal area of t h i s sprang section. A hold l i n e , wefted or otherwise secured, would be hidden be-neath the s t i t c h i n g and the applied fringe. Working a sprang warp i n sections i s quite rare although i t has been found i n exceptionally long, intertwined fabrics from Nasca. I t s use here i s a necessity, once the choice of making a shaped warp i s taken. To construct the bag, only the sides of the pockets and the bag proper need to be sewn together a f t e r the completed cloth i s transversely folded. The other bag with three pendant pockets ( F i g . 49) uses a straight-forward warping system with a l l threads of equal length. Work proceeds normally with the interlaced f a b r i c growing from both ends toward the centre. When the pockets are begun, the warp i s divided into three v e r t i -c a l sections and each i s worked independently, making s l i t s between the F i g . 49 Interlaced bag with three pockets shaped by making s l i t s i n the sprang fa b r i c (MRI, no number) v e r t i c a l sections. Although s t r u c t u r a l l y unnecessary, the threads change direction twice just above the pocket section. This i s done by i n t e r -l i n k i n g the threads i n two successive rows, rather than continuing to int e r l a c e . From the ridges i n t h i s area, i t also appears that wefts have been put i n these sheds. This i s not a hold l i n e , because of the lack of symmetry on either side of i t . Only one terminal area and hold l i n e would be necessary. I t would be i n the very centre of the f l a t warp, that i s , under the decorative s t i t c h i n g at the point where the fringe i s attached to the pocket bottoms. The three pocket bag i s a standard kind' of pre- and post-conquest a r t i c l e i n Peru. Ina Van Stan has discussed the technical requirements of s i x pre-conquest woven bags (1969-70). Some of the bags i n her study have pockets at the bottom, others have them positioned part way up the bag face. Bags of the l a t t e r sort, with one to f i v e pockets, continue to be made i n Peru and B o l i v i a today (Cason and Gahlander, 1976, pp. 31, 154-9)• The bags are generally used f o r carrying coca leaves and the pockets are used f o r carrying the lime with which the leaves are chewed (p.c. Market informant, Cuzco, 1978). The bag i n F i g . ,4/7 s t i l l has coca leaves i n i t (d"Harcourt, 1962, p. l6o) and the F i g . 45 bag also has remnants of dried leaves i n i t . Coca use has a long and impor-tant history i n Peru and i t seems l i k e l y that at least some of the inter - 5 laced sprang bags of t h i s s i z e , both with and without pockets, were made expressly f o r coca use i n ancient times as w e l l . The geographical and chronological placement of these fabrics has to be tentative because there are; not many comparative fabrics during the l a s t h a l f of the Early Intermediate Period. Two of the bags are given a gen-er a l provenience of Nasca (d'Harcourt, 1962,, p. 160 and the MNAA catalogue). The bag i n the Museo Regional de l e a had no catalogue information but i t i s probable i t came from the l e a area, l i k e the majority of the c o l l e c t i o n . The Lowie Museum bag (16-1104-3) has a provenience of Poroma, Tunga Valley and i t i s associated with a Nasca 6 ceramic (l6-1104la, LM catalogue) . A Nasca-Ica locale i s indicated f o r a l l the bags so f a r located. The bag that i s associated with the Nasca 6 ceramic does d i f f e r i n colour and interlacement order from the other three which are closely related i n these respects. The colours of the three closely related bags support dating them to the l a t t e r part of the Early Intermediate Period. They use the primary colours plus black and white. F i g . 4-7 also has a l i t t l e green (d'Harcourt, 1962, p. 160) and F i g . 44- has a l i t t l e orange. According to Ann Rowe, the primary colours and black are the favoured colours of t h i s period (1979, P- 117). The decorative s t i t c h i n g (cross-knit loop s t i t c h , Emery, 1966, F i g . 373) that closes the seams and strengthens the edges i s singled out by L i l a O'Neale (1934-, p. 410) for i t s use i n Late Nasca, although i t i s used i n a l l periods. Tentatively, t h i s group of interlaced sprang bags has been assigned to the South Coast area i n the l a s t h a l f of the Early Intermediate P e r i o d , c . 100-550 A.D. I f t h i s i s accurate, then they add to a very small group of f a b r i c s that can be placed i n the same period. Also, t h i s dating would mean that the three pocket bag has an even longer history than indicated by bags with overt Middle Horizon iconography. Technically, t h i s group of bags i s interesting because they i l l u s t r a t e a connection between the technique of sprang and the technique of weaving. Two bags i n the group use methods of shaping more frequently applied to weaving, (discontinuous warping and s l i t s , to produce a r t i c l e s similar to woven ones. Finishes, l i k e warp, fringes with loops and cross-knit looped seams and edges are also common to t h i s group and to woven a r t i c l e s . These bags may represent the f i n a l use of interlaced sprang i n Peruvian pre-history as l a t e r examples have not been.located so f a r . Interlaced Openwork One example of an openwork i n an interlaced sprang variant i s i n the Museo Nacional, Lima (T 252) . I t i s a long narrow fabric i n gold alpaca with slender t i e s attached to the ends. I t s certain use i s unknown but i t s form suggests a belt or headband. The major structure i s an unreported var iant of i n t e r l a c i n g (Fig. 50). Small amounts of 1:1 i n t e r l a c i n g i n h o r i -zontal bands are also present. The hold l i n e i s a horizontal chain of i n -terlooping which suggests t h i s a r t i c l e was worked i n horizontal rows. No information on the provenience or association of t h i s f a b r i c i s available,and, as i t i s a singular example i n form and structure, i t i s not possible to date i t very preci s e l y . Like the other openwork sprang, i t pro bably belongs to the time period between 600 B.C. and 550 A.D. Interlacing: Garment Ties The most numerous items made of interlaced sprang i n archaeological col l e c t i o n s are the t i e s on embroidered garments from Paracas Necropolis and other south coast cemeteries. The t i e s occur i n pairs or fours on wrap-around s h i r t s , aprons and lo i n c l o t h s i n embroidered styles that span Early Horizon 9 and Early Intermediate Period 2. Ususally, the two t i e s that were made on one warp are cut apart and so direct evidence that they were made i n sprang i s l o s t . However, a few unfinished speci-mens , l i k e the wrap-around s k i r t i n F i g . 51• show that sprang was the technique because the two t i e s are s t i l l joined by the centrally un-worked warps. The form (pointed t i p ) , the structure (2:2 interlacing) B F i g . 50 Interlaced sprang openwork (MNAA T252) A. D e t a i l with chained hold l i n e B. Interlaced structure g. 51 Paracas Necropolis s k i r t with interlaced sprang t i e s that have not yet teen cut apart (MNAA 0691. mummy bundle 378, specimen 18) and the positioning of the straps on many of the s k i r t s are very uniform (Bird and Bellinger, 1954, P I . 10). The uniformity suggests that a l l the straps were made i n the same technique. The evenness and control of the i n t e r l a c i n g on large numbers of threads argues that the straps were worked on tensioned or stretched ends, rather than free hanging ends. Unlike the rest of the sprang fabrics i n ancient Peru, these t i e s are worked i n oblique rows, rather than horizontal rows. The triangular point of the t i e and the d i s t o r t i o n just below the heading cord indicate t h i s ( F i g . 52A,B). The interlaced structure i s over two, under two. The outer threads interlace to the centre where they cross each other i n a 1:1 order of interlacement ( F i g . 53)- Near the heading of each t i e , the f i r s t threads l i e i n near horizontal paths, causing the t i e to draw i n . After several inches of work, the threads l i e on oblique paths which are" maintained throughout. I was puzzled f o r a long time by the l o g i s t i c s of spranging such an unweildy length of f a b r i c . The straps are commonly 1.20 M long (Garrion:',Gachot, 1931, P« 81) which means the warp would be over 2.4-0 M. Not even a Boston C e l t i c would have the reach to sprang a f l a t warp that long. In experimenting with the technique, I discovered that the warp could be very conveniently worked i f i t was folded i n h a l f and the two ends were placed'. one above the other and attached to a fi x e d point to-gether. Then, with h a l f of the warp loops (caused by folding the warp) i n each hand, I could tension the warp and make the in t e r l a c i n g with threads i n the top layer ( F i g . 5*0 • A convincing aspect of t h i s method of working i s the effective way each row i s driven into the f a b r i c . By simply spreading my arms, the l a s t row was firmly packed into the top and bottom layer of the fab r i c at the same time and along neatly oblique . 52 Details of s k i r t t i e s A. Heading area of one B. Pointed t i p s of two unworked warps (MNAA 0691) t i e t i e s , s t i l l joined by the F i g . 53 Structure of interlaced sprang t i e s , shown with number of threads greatly reduced: over two, under two i n t e r l a c i n g along oblique working edges Fig. 5k Method of working two ties simultaneously on a folded warp l i n e s . This straight forward method i s quite feasible f o r the t i e s , as i t i s easily e f f e c t i v e over a length of 1.5 M. Furthermore, i t keeps the working area conveniently close to the maker, no matter what length the warp i s . The sprang garment t i e s d i f f e r s i g n i f i c a n t l y from other ancient Peruvian sprang f a b r i c s i n having a oblique working edge. I f e e l that the method just described applies only to fabrics with an oblique work-ing edge, l i k e the t i e s . I t seems probable that these t i e s have an evo-l u t i o n related to braiding on free ends as braids generally have an oblique working edge as w e l l . The other Peruvian sprang fabrics appear to have been b u i l t up i n horizontal rows, l i k e weaving. This technique poses a s l i g h t terminological problem as i t could be described as sprang or loop-end braiding. As two symmetrical fabrics are produced by i n t e r -working a stretched (although not fixed) set of elements, the technique i s included i n a study of sprang. Interlaced sprang t i e s f o r garments were used over a considerable length of time. S k i r t s of the l i n e a r , geometric embroidery style and ones with the co l o u r f u l , c u r v i l i n e a r style (Carrion Cachot, 1931, Fig • 19d. and 19b) are dated to Early Horixon 10 and Early Intermediate Period 2, respectively, by Jane Dwyer's s e r i a t i o n a l features (l979i P« 109 and 113). The s k i r t i n F i g . 51 belongs to the Early Intermediate Period 2 s t y l e , i d e n t i f i e d on the basis of the monkey-like foot of the figure ( J . Dwyer, 1979> P- 115) a n (l other features. Similar wrap-around s k i r t s from Ocucaje (King, 19^ 5i P- 448-9) and from the Cavernas si t e at Paracas (Carrion Cachot, 1931» F i g . 4m) have also been published. These two s k i r t s with straps are probaby Early Hbrizon 9' The design of the Ocucaje s k i r t , "geometric birds i n lozenges" (King, 19&5, F i g . 8a, p.. 72), relates to Early Horizon 9 designs i n gauze and other techniques ( J . Dwyer, 1979, F i g . 1; Yacovleff and Muelle, 1932, F i g . 6). The Cavernas s k i r t i s from the same si t e as gauzes and double cloths dated to Early Horizon 9 by Jane Dwyer (1979, p . 107-8). Another garment type probably has s i m i l a r l y made t i e s . Carrion Cachot i l l u s t r a t e s f i v e waras, or l o i n c l o t h s , from the Necropolis s i t e . They are rectangular or triangular cloths with symmetrical straps at two. cor-ners of one side (1931, F i g . 20). Two of the lo i n c l o t h s have straps that have large t a s s e l additions at the ends, l i k e the ends on several s k i r t t i e s she i l l u s t r a t e s ( F i g . 19d,f). Another garment from Cahuachi, called an apron by O'Neale (1937, p. 191 and P I . XXXIVa), has four straps but she doesn't describe the technique. She does say that two t i e s are one inch longer than the other two. This symmetry i n size between pairs could suggest the sprang technique. Interlaced sprang was used for narrow t i e s between at least Early Horizon 9 and Early Intermediate Period 2 (c. 600 - 200 B.C.). There are examples on wrap-around s k i r t s and l o i n c l o t h s or aprons from a num-ber of south coast s i t e s , most p a r t i c u l a r l y Paracas Necropolis. The tex-t i l e s of the succeeding and. preceding phases are not very f u l l y known so t h i s technique may have had a longer history. Several authors have made suggestions about how the s k i r t s and other garments (mantos, waras, esclavinas and ll a u t o s ) from the Necro-p o l i s mummy bundles were worn i n l i f e (Bird'and'Bellinger, 1954, p. 16 and Carrion Cachot, 1931, P- 81). The^'Museo Nacional, Lima has construc-ted a human model dressed i n a set of matched garments from mummy bundle 421 (Carrion Cachot, 1931, F i g . 13, p. 74). Though the model i s large com-pared to Peruvian median heights, the clothes are s t i l l larger. The elaborateness and the p r i s t i n e quality of most of the embroidered garments 100 suggest they may have been made expressly f o r b u r i a l . The large size of the interred garments suggests they may have been made for a larger, human substitute, that i s , the mummy bundle i t s e l f which i s constructed i n a way that imitates a human, complete with a false head. Although these garments may not have been worn i n l i f e , t h e i r design may well derive from functional clothing. Intertwining: Openworks Intertwining i s the most specialized and complex structure worked i n the sprang technique. In ancient Peru, i t appears to have been used long-er and more extensively than i n t e r l i n k i n g or i n t e r l a c i n g and to have been used for several different types of f a b r i c . One fa b r i c type i s a very elegant and stable openwork, generally made of monochromatic alpaca with designs that are most frequently geometric (F i g . 55). The most common intertwined structure of openwork fabrics i s shown i n F i g . 5^ . A diagonally moving pa i r of threads encloses one thread of a p a i r moving on the other diagonal. Each pa i r i s twisted before repeating the interworking with the p a i r at the next intersection. The openwork contrast i s often produced by coupling t h i s structure with p a i r s that have been given multiple twists ( F i g . 5&B)• The proof that sprang was the technique i s the presence of a hold-l i n e and mirror-image symmetry of structure and design on either side of i t . Many fa b r i c s are fragmentary and r e t a i n only a small area of the hold l i n e ( F i g . 57)- Even more fragmentary pieces have no i n t r i n s i c proof of the sprang technique. Both the fabrics with direct evidence of the sprang technique and those that are now too fragmentary to have the proof use the same l i m i t e d range of structures and designs. Because of D e t a i l of an intertwined sprang openwork (from a MNAA s l i d e , no number available) 102 Fi'g. 56 Intertwined sprang diagram A. Most common intertwined structure B. Twisted pa i r s combined with the structure i n A F i g . 57 Fragmentary intertwined sprang openwork with a small section of the hold l i n e , Gahuachi, Nasca Valley, AMNH 41 .0/5400 (photo courtesy of Barbara Conklin) 104 the homogeneity of the group, i t i s suspected that a l l intertwined frag-ments of t h i s general type were made i n the sprang technique. The conser-vatism i n ancient Peruvian techniques as well as the great advantage i n doing intertwining i n sprang (the duplicating capacity and the f i x e d order-ing of tensioned threads) support the contention that even0 the fragments were products of the sprang technique. For these reasons, intertwined fragments without absolute evidence of the sprang technique but which re l a t e closely to fabrics that have the evidence w i l l be included i n the sample of intertwined sprang openworks. Several fragments of intertwining which were too incomplete to con-t a i n direct proof of the sprang technique have been published and they are included i n t h i s study. D'Harcourt c l a s s i f i e s Plate 54b under " P l a i t i n g of Yarns with the Lower Ends Le f t .Free" (a non-sprang technique), yet he also acknowledges that " t h i s method was also used i n p l a i t e d speci-mens i n which the yarns were held fast at t h e i r extremities" (l n 62, p.?4-6). The l a t t e r quote amounts to a description of the sprang technique. 0'Neale describes the intertwined piece she twice published as "twine p l a i t i n g " , a term she seems to use f o r a l l fabrics with diagonally twining p a i r s , irrespective of whether they were done on f i x e d ends (sprang) or free ends (O'Neale and Kroeber, 1930, PI. 7a, O'Neale, 1937, P L LIV). Two unfinished sprang warps i n the Lowie Museum, Berkeley give some insight into the manner of working intertwined sprang. Both have con-tinuously wound warps with heading cords inserted i n the f i r s t sheds at either end of the warp. One warp has only the heading cords and the other has a few inches of f a b r i c at both ends. The more advanced warp has c l e a r l y been worked i n horizontal rows from a shed with alternate threads ra i s e d and lowered, as the termination of work and the shed cord indicate ( F i g . 58) • I t seems most plausible that the -warp was stretched f l a t between bars or supports and horizontal rows of interworking were carried to both ends of the warp. Collingwood describes i n d e t a i l a method for t h i s way of working (1974, p. 209-10). The unfinished fabric was probably prepared f o r b u r i a l by removing the re-usable equipment and tying the bunched ends with cord. I t i s l i k e l y loom bars and possibly shed s t i c k s , for carrying the rows of intertwining to the far end of the f a b r i c , were used. The dimensions of the piece, 1.10 M x 0.50 M suggest some auxi-l i a r y tools were used. The precise uses of the intertwined cloths are not known. There are, however, at l e a s t two sizes represented among those with i n t a c t edges. The largest ( F i g . 59) has a width of 0.88 M and an o r i g i n a l length of at leas t 4.00 M, calculated from the placement of hold l i n e s . Other pieces, l i k e F i g . 57 and examples i n the Museo Nacional, Lima and the Peabody Museum, Harvard are about a metre wide but fragmentary i n length. Cloths of t h i s size may have been wrappings f o r mummies. The regular deteriora-t i o n i n the F i g . 59 fabric i s consistent with being used as a shroud. There are also narrow fabrics of varying lengths from 1.10 M (Fi g . 58 and LM 16-10154) to 3.20 M (OM C50) . ;.The size suggests they may have been headdresses. Although none of the fabrics examined i n museums were found on bodies, a photograph (Ubbelohde-Doering, I967, 191) shows two trophy heads from a Potrero grave, near Cahuachi. The upper head appears to have an intertwined cloth wrapped around the head. The knob of excess fabric was placed on the forehead, although the entire headdress has slipped down over the eyes. The other intertwined fabrics are of indeterminate size due to disintegration. I t i s worth noting that the openwork cloths of F i g . 5 8 Unfinished intertwined sprang warp A. LM 16.10153, Ghanguillo, Ingenio Valley B. De t a i l of transverse working edge and shed cord F i g . 59 Large intertwined sprang fabric that may have teen a wrapping for a mummy, MM 1954 W605 97492 Early Horizon 9 were also used as headdresses and wrappings for mummies. Most of the intertwined openworks i n t h i s group are monochrome wool, probably alpaca. White, various shades of yellow, red and green are present. There i s a small range of mostly geometric designs that follow the structural diagonals of the intertwining p a i r s . The geometric de-signs of the intertwined sample are summarized i n F i g . 60A-G . Diamonds and zigzags are the primary motifs and they are frequently .combined i n the same piece. One of the designs (G) which has been published i s re-ferred to as "serpentiform" (Tello and Xesspe, 1979, P- 126). I agree with t h i s interpretation but include i t with geometric designs on the ar b i t r a r y grounds that there are no eyes. Three examples of more representational motifs appear i n the sample as well ( F i g . 6l). One i s c l e a r l y a serpent head design (A) while another i s a quadruped with a t a i l ( f e l i n e or monkey?). The f i n a l de-sign resembles some representations of trophy heads with hair stream-ing out behind. To see t h i s design as a trophy head, i t i s necessary to accept the convention that the outline; i s i n p r o f i l e and the eyes are f r o n t a l (C) . / In the range of designs on intertwined openworks, presented i n F i g . ,60 and 61, the numbers are weighted more heavily to-ward the geometric designs. Although diamonds and zigzags are described as geometric motifs, I do not consider them to be without meaning. The zigzag i n Peruvian ar t frequently has a serpent hea !| or a feline-serpent head attached to i t . Diamonds often f i l l the interspace between two opposing zigzags and t h i s combination of markings frequently covers the body of an overt serpent representation, l i k e the Paracas braid i n F i g . 62. This braid has a whole series of serpent heads attached at i n t e r v a l s along the length 109 F i g . 60 Geometric designs on intertwined sprang fabrics A. F i g . 55j TM I960.1213Ay FI 7113; PMH, no number available B. MM 1954 ¥605 97492 G. LM 16-10153 D. AMNH 41.0/5400 E. LM 4-8843 F. LM 4-8537 G. MNAA 12/6545 110 fc c F i g . 61 Figurative designs on intertwined sprang fabrics A. Serpent heads OAG C52 B. Quadruped ( f e l i n e or monkey?) MNAA 06005 and MNAA 02293 G. Trophy head? MNAA 02293 Ill Fig. 62 Serpent braid with geometric markings A. MNAA 01875, Paracas Necropolis mummy bundle 38 specimen 43 J ' B. Detail of diamonds and zigzags on braid of the zigzag and diamond braid. This i s , by no means, an i s o l a t e d example. For instance, the feline-serpent e f f i g y vessels of Pashash have a sim i l a r range of geometric designs along the coiled, serpentine bodies of the vessels (Grieder, 1978, F i g . 57» PP» 195-7)• The zigzag may r e f e r to the undulating aspect of a snake and both the diamond and zigzag seem to refe r to the markings on snakes. The structures of most of the geometrically patterned intertwined f a b r i c s are l i k e those shown i n F i g . 55A,B. However, a number of minor variants i n structure were found. The fragment i n F i g . 63A has another structure as w e l l , i n the small dark diamonds that make up several zigzags This structure diverges from the usual one i n the selective omission of twists between some interworkings and the addition of a twist between the groups of interworkings ( i e . the small diamonds), as shown i n the diagram i n F i g . 63B. Another variant occurs i n two large fabrics with figurative designs ( F i g . 64A , B). The same p r i n c i p l e s of omitting some twists between i n t e r -secting pairs and adding a twist between groups of intersections are used. Only the precise placement of twists d i f f e r ( F i g . 64c) . In these fabrics no twisted p a i r s are used. Three other openworks'; from the Ohara Art Gallery i n Kobe, Japan (C50, G52, D33) may have the same structure but the. available photographs are not quite detailed enough for certain i d e n t i f i -cation . A f i n a l variant of intertwining produces the open areas i n a frag-ment that was studied at the Textile Museum (Fig. 65A). The open areas simply double the threads involved i n each interworking and then separate each interworking with an extra t w i s t . So, two pairs on one diagonal intertwine with two pairs on the other diagonal. Although two pairs act B F i g . 63 Intertwined fragment with a variant structure, i n addition to the usual structures i n F i g . 56 A. LM 4.8537 B. Diagram of variant form of intertwining that appears i n the small diamonds (see arrow) c F i g . 64 Intertwined fabrics with a variant structure; no twisted pairs used A. MNAA 02293 B. MNAA 06005 G. Diagram showing two types of intertwining used for contrast i n A and B F i g . 65 Intertwined f a b r i c with a variant structure; no twisted pairs used A. TM i966.99.35 B. Diagram showing two types of intertwining used for contrast i n A l i k e one pair at the intersection, each thread twists only with i t s o r i -g i n a l partner i n between intersections ( F i g . 65B) . Very few intertwined openworks have been published, yet fragments are encountered not infrequently i n museum co l l e c t i o n s . I t i s quite pos-si b l e that there are more variants i n structure than the ones just described. An extraordinary piece of openwork intertwining does deserve special attention. The large, red fragment i n F i g . 59 has unmistakable signs that an unusual method was used for fabricating an inconveniently large f a b r i c . The ov e r - a l l design of horizontal zigzags i s interrupted by three h o r i -zontal rows of diamonds. A p a r t i a l row of diamonds can be found at each fragmentary end (Fi g . 66). The design and the structure (direction of twining twist and the diagonal d i r e c t i o n of the surface pair) are mirror-imaged at the exact, transverse centre of these diamonds. At one end, a fragment of a wefted hold l i n e i s s t i l l i n p o s i t i o n . There must have been a sim i l a r hold l i n e at the other end to have kept the section from unravel-l i n g but only the course for i t now remains'. The mirror-imaging of design and structure indicate that both of these now fragmentary rows of dia-monds were terminal areas of construction and that the f a b r i c must have been worked i n at least two separate sections. The t h i r d row of diamonds i n the approximate centre of the fragment i s , on close examination, quite d i f f e r e n t . S t r u c t u r a l l y , there i s no change of twining twist at the centre of the diamonds and the uppermost p a i r continues on the same diagonal (upper l e f t to lower r i g h t ) across the entire diamond ( F i g . 67). This could not be an area of terminal construction because the structure i s not i n mirror-image. Rather, the central row of diamonds i n F i g . 67 i s the l i n e along which the two separately worked sections meet. D F i g . 66 Large size intertwined sprang f a b r i c worked i n two sections (MM 1954- W605 97492) A. D e t a i l of terminal area of construction with hold l i n e B. D e t a i l of second terminal area of construction: hold l i n e no longer i n place F i g . 67 D e t a i l showing where two separately worked sections abutt (MM 1954 ¥605 974-92) A diagrammatic reconstruction of the whole fabric ( F i g . 68) shows that the o r i g i n a l f a b r i c must have been almost twice as long as i t pre-sently i s . The present length i s indicated by the heavy i r r e g u l a r l i n e s . The brackets ( l e f t ) enclose the separately worked sections. The row of diamonds i n the centre of each bracketed section i s the area of terminal construction ( s o l i d l i n e s ) . The dotted l i n e shows where the two sections abutt. O r i g i n a l l y , the cloth must have extended an equal distance i n each direction from the areas of terminal construction. Calculations based on measurements taken from the fragment indicate that the cloth was o r i g i n a l l y about 4.00 M long and was worked i n 2.00 M sections. There i s a p o s s i b i -l i t y that the cloth was even longer, that i t was worked i n three or four sections, but there i s only evidence of two sections. A study of the twist of the two sections suggests the warp was worked from both ends. I f c r a f t habits are fixed, and i t i s suggested that they are, then the maker worked one section from one end with the S twist i n t e r -twining nearer the body and then worked the second section from the other end of the warp, again with the S twist intertwining nearer the body. Done i n t h i s way, the worker's hands make the same set of motions for both sections. No insight into the suspension and tensioning of the sprang warp was gained from t h i s piece which has fragmentary warp selvedges. The length and width seem p r o h i b i t i v e l y large f o r a body-tensioned loom. A horizon-t a l staked-out loom i s a p o s s i b i l i t y , although each section i s about two metres long and some system of carrying sheds (on sticks?) beyond the maker's reach would need to be employed. Since studying t h i s piece, I have encountered two more fragmentary fabrics that were worked i n at least two sections (MNAA 06005 and MNAA 02293) and have seen s l i d e s of another 120 / 1 / V V V V V V V „ A A A / W v V ^ w y y y y y y ^ A A A A A A A A V W V W W XMXMMXMX . vyvwyyy A A A A A A A A ^ v A 0 v W v \ A V \ A A A 7 V V M F i g . 68 Schematic drawing of the reconstructed f a b r i c i n Fig« 59 • The brackets indicate the separately worked sections and the s o l i d l i n e s the areas of terminal construction. The heavy, ir r e g u l a r l i n e s show the present extent of the fragmentary f a b r i c . The oblique l i n e s at the r i g h t show the direction of the twining t w i s t . ^ 121 i n the Peabody Museum, Harvard (no number a v a i l a b l e ) . The same problem of fragmentary warp ends occurred i n the two I was able to study. The knowledge of pre-Columbian looms i s s t i l l very incomplete and many woven fabrics of incredible length from the Necropolis burials s t i l l perplex researchers. James Vreeland (1977) has reviewed the evidence f o r suggesting a v e r t i c a l loom was in'use and, while h i s findings were not conclusive, i t may be another p o s s i b i l i t y . For now, the exact method of making huge fabrics remains another puzzling accomplishment of the ancient Peruvians. The intertwined openworks that have provenience are mostly from the valleys of the South Coast: Changuillo, Ingenio Valley (LM 16-10153 and LM 16-10154), Nasca Valley (AMNH 41.05400), C a c a t i l l a , Nasca Valley (LM 4-8537), Nasca Valley (LM 4-8843, LM 4-8844). The only intertwined openwork said to be from outside of the south coastal area i s one i l l u s -trated i n d'Harcourt (1962, P I . 54b) which i s said to come from the central coast region. Early Intermediate Period t e x t i l e s from the central coast are incompletely known at present, which may account f o r the single a t t r i b u t i o n of an intertwined openwork to a central coast l o c a l e . I t i s d i f f i c u l t to assess the time period of the intertwined sprang openworks. A photograph of a yellow intertwined d e t a i l ( F i g . 55) was pur-chased at the Museo Nacional, Lima and i t had a Paracas Cavernas Culture l a b e l (Early Horizon 9). Another f a b r i c from Paracas Cavernas, described as "calada", or openwork, technique, i s sketched i n the recently released book, Paracas, I I Parte: Cavernas y Necropolis (j.C. Tello and T.M. Xesspe, 1979). Although the sketch i s not absolutely precise, i t does appear to be intertwining (F i g . 69). The twisting p a i r s , the oblique emphasis and the "serpentiforme" design (p. 126) do support the i d e n t i f i c a t i o n as 123 intertwining. O'Neale places a G a c a t i l l a fragment (F i g . 63, LM 4-8537) i n Early Nasca ( f i r s t part of the Early Intermediate Period) on the basis of s t y l i s t i c s i m i l a r i t y but not on associated pottery (O'Neale and Kroeber, 1930, p. 25, Footnote 9) • The other intertwined fabrics i n t h i s study have no ceramic association either. Some s l i g h t , i n d i r e c t evidence f o r dating comes from Nasca ceramic designs from the middle of the Early Inter-mediate Period. A few ceramics i n the Lowie Museum (4-8572, 4-8583, 4-8751» 4-8745) have bands of t e x t i l e designs that resemble intertwined f a b r i c s but they are very generalized. The trophy head, i l l u s t r a t e d by Ubbelohde-Doering (1967, 191), that appears to have an intertwined head-dress i s dated to that period between the "Old Nazca Culture arid Coastal Tiahuanaco" (p. 142). In Rowe's chronology, t h i s would be approximately Middle Horizon 1, c. 600 A.D. I f various i d e n t i f i c a t i o n s can be r e l i e d on' intertwined sprang openworks may have been made from Early Horizon 9 u n t i l Middle Horizon 1 (600 B.C. - 600 A.D.) . Within the sample of intertwined sprang openworks, two groups can be separated on the basis of structure: those that use twisted pairs and the basic intertwined structure ( F i g . 56) and those that use two va r i e t i e s of intertwining ( F i g . 64, 65) f o r the openwork contrast. Those that belong to the f i r s t group have geometric designs ( F i g . 60). A l l three fabrics with fi g u r a t i v e designs belong to the second group ( F i g . 6 l ) . There i s too l i t t l e information on grave context to be certain that any time d i f f e r -ence i s represented by the differences between the groups i n structure and iconography. However, several of the geometrically patterned f a b r i c s that use twisted pairs do have e a r l i e r associations (Fig. 69, MNAA 12/6545; F i g . 63, LM 4-8537)- The fabrics that use two var i e t i e s of intertwining for openwork contrast (and do not use twisted pairs) have intertwined 1 2 4 structures that relate more closely to the intertwined sprang tassels of the Middle Horizon which w i l l be described i n the next chapter. I t i s possible that the less common v a r i e t i e s of intertwining with the fi g u r a -t i v e designs are l a t e r as they provide a smoother t r a n s i t i o n , s t r u c t u r a l l y , to the figu r a t i v e tassels of Middle Horizon IB. Summary The most frequent and varied use of the sprang technique occurs be-tween Early Horizon 9 and the beginning of the Middle Horizon. The three basic structures of i n t e r l i n k i n g , i n t e r l a c i n g and intertwining are found primarily i n the south coastal area. The geographical d i s t r i b u t i o n could be f a r wider but the fab r i c s of the highlands and many parts of the coast are simply not known because conditions f o r f a b r i c preservation are so p a r t i c u l a r . From the present sample, which may be skewed toward the more showy f a b r i c s , i t appears that i n t e r l i n k i n g was used early i n t h i s time span while i n t e r l a c i n g and intertwining were used fo r longer and for several v a r i e t i e s of f a b r i c s . Interlacing appears mostly i n firmly worked fa b r i c s , l i k e garment t i e s and bags, that are unpatterned except for colour s t r i p i n g . However, one example of an interlaced openwork was located. I n t e r l i n k i n g and intertwining appear almost exclusively i n openwork f a b r i c s . The s i z e , the material (alpaca) and the iconography of the openwork sprang fabrics suggest that these were high status f a b r i c s and t h e i r probable use as headdresses and shrouds supports t h i s . The predominance of serpentine iconography i n the form of fabric structure designs on the i n t e r l i n k i n g and i n the form of zigzag/diamond motifs on the intertwining i s an interesting example of the layering or underlining of meaning. Serpents, i n early Chavinoid art on stone, are the metaphor for natural body coverings, l i k e human h a i r , animal p e l t s and b i r d feathers ( J . Rowe, I967, Figs. 15, 17 and 21). Throughout the Chavinoid phases, the serpent i s also associated with images of f a b r i c a -ted body coverings, l i k e belts and headdresses, and the images of f a b r i c •structures themselves. In early Horizon 9, f a b r i c s that are s t r u c t u r a l l y patterned are especially favoured and many are adorned with serpentine patterns (Carrion Cachot, 1931). I t i s as i f the twisting movements of the threads portray the same meaning as the serpentine structures that appear as designs. That t h i s moiling mass of serpents then wraps the heads and bodies of the dead seems to trebly enforce the association of serpent iconography with body covering. Serpents are natural analogues to body coverings i n several ways. The sloughing of an i n t a c t skin i s an impressive event i n nature which p a r a l l e l s the exterior, or covering, aspect of serpent imagery i n the a r t . The l i n e a r and f l e x i b l e body of the serpent i s a physical analogue to the h a i r , fur and feathers of natural body coverings. The serpent i s also physically analogous to the f i b r e s and the threads of fabricated coverings. Not only animal sources of f i b r e s , but the cotton plant as well (Cordy-Collins, 1979, F i g . 3) a r e given serpentine aspects i n Chavinoid-; „ a r t . Fabrics are related to serpents i n both t h e i r body covering r o l e , and as interworkings of l i n e a r , f l e x i b l e elements. The most convincing i l l u s t r a t i o n of the connection between serpents and fabrics i s i n the "macro" images of fa b r i c structures rendered as serpent bodies. In these designs, the structure or the very substance of fabric i s given an icon i c significance through i t s association with the image of serpent. The ser-pent has a wider set of references than body coverings i n Peruvian a r t . Nevertheless, the association with fabrics i s very apt as the metaphor works on a number of l e v e l s simultaneously. The transluscence of the openwork headdresses and mantles i n t h i s study may relate to another as-pect of serpents. Like the transparent skin sloughed by a serpent i n the process of growth, these semi-sheer fabrics enshroud the mummy i n the transforming surroundings of the tomb. The stimulus f o r the extraordinary range and beauty of ancient Peruvian f a b r i c s may well have been that the substance and process of fabric making were as i n t e g r a l to the r i t u a l and art as the images. In the narrow terms of t h i s paper, that may explain why there i s such a di v e r s i t y of sprang openworks i n the nascent, then f l o u r i s h i n g t e x t i l e a r t of Paracas and Nasca. The fourescence of structural explorations i n fabric making on the south coast does continue b r i e f l y into the Middle Horizon, along with other Nazcoid c u l t u r a l elements. The most i n t r i c a t e and flamboyant sprang fa b r i c s date to approximately Middle Horizon IB (p.c. Lawrence Dawson) and are the subject of the next chapter. FOOTNOTE Wallace i s not i n complete agreement with Rowe, Menzek and Dawson (1964). He dates t h i s f a b r i c , through ceramic association to Ocucaje 7 (upper valley) but feels i t i s contemporaneous with Ocucaje 8 (lower valley) i n t h e i r system. He sees t h e i r temporal d i s t i n c t i o n between Ocucaje 7 and 8 as a contemporaneous geographical d i s t i n c t i o n , i . e . upper valley and lower va l l e y , respectively (p.c. Dwight Wallace, 21/12/79) • CHAPTER IV THE MIDDLE HORIZON Tassels In the l a s t throes of the south coast Nasca culture, the a r t style on ceramics and fabrics becomes f r e n e t i c a l l y a c t i v e . Figures are not i s o l a t e d i n space but appear to be ensnared i n t e n d r i l s of rays and volutes. Often the figure i s almost l o s t i n the encroaching mass of r e p e t i t i v e secondary motifs. Fabric making also goes into convulsive complexities and the tangle of interpenetrating sets of threads coin-cides, perhaps echoes, the highly worked style of the images. A number of south coast fabrics from near the beginning of the Middle Horizon have continued to defy technical analysis. One notable group i s a series of bell-shaped tassels which are usually made i n intertwined sprang ( F i g . 70). The patterning of the tassels most often uses the double cloth p r i n c i p l e of interpenetrating two layers of f a b r i c of contrasting colours although one example of t r i p l e cloth ( j . B i r d , p.c. and 1963b, p. 58) and one example of quadruple cloth (d'Harcourt, I962, P I . 58) have been reported. These extraordinary tassels have fascinated other writers and they have been published a number of times (Lehmann, 1924, P I . 1; d'Harcourt, 1962, P I . 57, 58; O'Neale and-Kroeber, 1930, P I . 21; B i r d , 1963b, P I . 10; Kanegafuchi, I956, Vol. 1, No. 10 and Vol. 8, No. 76; Lothrop, Foshag and Mahler, 1957, P L CXLII; A. Rowe, 1973, No. 38, p. 4 and Collingwood, 1974, P L 43). However, accompanying information on structure has been incomplete as the denseness of the cloth has impe-ded analysis. The iconography has been and continues to be less than 129 A. F i g . 70 Intertwined sprang t a s s e l , Middle Horizon IB A. MM 1954 WCh. 5 599 B. Figure on warp cover G. Figure on ta s s e l body, repeated twelve times well understood. Even the use to which the tassels were put i s a matter of conjecture. My research adds to the e x i s t i n g information on technique and, i n p a r t i c u l a r , provides an analysis of the structures of the-intertwined sprang tas s e l s . Also, by drawing together a large group of tassels which are technically homogenous, the record of the iconography of the s t y l e i s enlarged. The more recent research of others on the style and iconography i s applied, to some extent, to the sample f o r dating. The association of the tassels with other a r t i f a c t s or with a mummy has not been reported. Almost a l l the tassels located i n museums appear to be the re s u l t of huaquero (grave looter) a c t i v i t y so that even proven-ience i s rare. However, one t a s s e l or fragment (Fig. 71) i n the F i e l d Museum comes from Kroeber's work at Cahuachi i n 1925 and 1926. I t i s un-clear whether t h i s piece was surface gathered or whether i t came from an excavation with or without associated pottery (Kroeber and O'Neale, 1930, p. 24). Kroeber and 0'Neale assigned i t to the Epigonal s t y l e . A l e t t e r to the F i e l d Museum regarding s p e c i f i c associations has remained unan-swered. The tassels i n the Musee del'Homme (d'Harcourt, 1962, P I . 57, 58) are given a provenience of Nazca (p. 161) but no information on the source of the a t t r i b u t i o n i s given. Although provenience information i s scarce, none contradicts a south coast l o c a l e , which i s supported by s t y l i s t i c and iconographic a f f i n i t i e s to t e x t i l e s of secure provenience (O'Neale and Kroeber, 1930, P I . 16, Trancas and PI. 19, G a c a t i l l a ) . As no tassels are reported to have been found i n s i t u , the manner i n which they were worn i s unclear. Id e n t i c a l and nearly i d e n t i c a l p a i r s of tassels are i n museum coll e c t i o n s ( F i g . 72). Some pairs are s t i l l joined by a long, t h i n strap (TM 1959.11.3 and MDH 34-145-2). Some single 131 F i g . 71 Intertwined sprang tassel from Cahuachi A. FM 171134; photo taken from O'Neale and Kroeber, 1930, PI. 21. B. Figure on t a s s e l body, minus the head which i s not shown i n the photograph 132 F i g . 72 P a i r of tassels A. TM 91.537, photo courtesy of Alan R. Sawyer B. Figure on the tassel body, repeated four and one h a l f times 133 tassels have remnants of sewing thread that suggest they, too, were o r i -g i n a l l y joined to a si m i l a r strap. The pairing and collocation of joined parts suggests that the sprang parts hung, t a s s e l - l i k e , from the long cord which may have been wrapped around the head or even the waist. Their con-struction and suspension makes i t clear they were not used as containers as the opening f a l l s downward, away from the suspension point. A rare, modelled scene from the south coast area i n the l a t t e r part of the Early Intermediate Period shows i n d e t a i l elaborate headdresses and simple clothes on Nasca style male and female figures (Tello, 1931, P i g ' 1 »3>4-,5i6,) . Although no parts of the headdresses shown correspond very closely to the tassels, the emphasis on the headdress suggests a stronger l i k e l i h o o d they were incorporated there rather than over the austerely simple tunics and l o i n c l o t h s that cloth the figures. The tassels have been c a l l e d "neck coverings" by d'Harcourt (1962, p. 82). Although he didn't explain why, he obviously f e l t they were worn on the head and hanging down over the neck. The style of the majority of the tassels has been i d e n t i f i e d as the Ghakipampa B style (p.c. Lawrence Dawson). The Ghakipampa B style i s the l o c a l , secular style of the Ayacucho-Huari area (Highlands) i n Epoch IB of the Middle Horizon (Menzel, 1964, p. 68). The Ghakipampa style has a very d i s t i n c t i v e appearance but i t s orig i n s , extent and evolution, p a r t i c u l a r l y i n t e x t i l e s , i s incompletely understood. In describing the style for ceramics, Menzel says i t draws on themes from the Nasca t r a d i t i o n , both Nasca 7 and 8, and related fea-tures of the Huarpa style (the Ayacucho highland style just p r i o r to the Middle Horizon) as well as new influences from the Nasca 9 style (which i s r e s t r i c t e d to the coast during Epoch l' of the Middle Horizon). She also says there are innovating features present whose antecedents are not now known (1964, p. 10). Although the tassels are believed to have come from the coast, t h e i r style i s a highland one which has a coastal (Nasca) s t r a i n i n i t . The cu l t u r a l a c t i v i t i e s that resulted i n t h i s s i t u a t i o n , according to Menzel's interpretation, were ones of strong interaction between the coastal (Nasca) and the l o c a l highland (Huarpa) cultures during the late stages of the Early Intermediate Period. Even as Nasca's power waned i n the Middle Horizon, the re l a t i o n s between Nasca and the highlands remained close, both areas influencing and being influenced by the other. This situ a t i o n explains why Nasca 7, 8 and 9 themes are present i n the Ghakipampa B style and why the Ghakipampa B style i s found on coastal t e x t i l e s and pottery i n Middle Horizon IB, c. .600 A.D. Menzel has com-pared Nasca to ancient Greece, i n that Nasca seems to have enjoyed a pri v i l e g e d position i n the new Middle Horizon empire, even after i t s own power dwindled (1964, p. 66-68). The Ghakipampa B style i n t e x t i l e s has been p a r t i a l l y described by Ann Rowe;, (1979)• Although no t e x t i l e s of t h i s style have been reported with ceramic associations, she has i d e n t i f i e d a number of fabrics of t h i s style through iconographic comparisons,..to.(ceramics. She expanded the sample through technical comparisons of f a b r i c s with i d e n t i f i a b l e icono-graphy. She i d e n t i f i e d fabrics i n various structures, including one of the tassels i n t h i s study, which bear a wide range of presently l i t t l e understood iconography. Her work has provided a broad view of what the Chakipampa B style i n t e x t i l e s looks l i k e , but she r e a d i l y admits that secure r e l a t i v e dating of the f a b r i c s awaits the discovery of ones with ceramic association (1979, P« 117)-Thirty-nine hell-shaped tassels (nine p a i r s i n the group) have been located i n publications, museum collections and through correspondence with other researchers who generously added to the sample. The majority are made i n intertwined sprang, but even those that are not are included i n order to more completely discuss t h i s d i s t i n c t i v e type of adornment. The sample has been divided into four categories on the basis of structural differences. When t h i s d i v i s i o n was made, i t became clear that the categories were s t y l i s t i c a l l y d i s t i n c t as w e l l . Part of the style difference might be a t t r i b u t e d to the l i m i t a t i o n s imposed by the different structures. However, i t seems l i k e l y that the differences i n both structure and style are responses to c u l t u r a l differences, either temporal or geographic. Category 1 The largest group (21 out of 39), which might be considered the "standard" type, i s put i n the f i r s t category. The group i s technically and s t y l i s t i c a l l y quite homogenous although colour choice and s p e c i f i c iconography vary. Two layers of f a b r i c made from contrasting sets of threads interpenetrate to produce figure against ground i n an a l l - o v e r , maze-like style of patterning ( F i g . 73)• The back side of the two i n t e r -penetrating layers has the same design with colours i n reversed positions. The technique i s undoubtedly sprang because the central, unworked warps s t i l l j o i n the two r e f l e c t i v e l y symmetrical pieces which are worked i n a variant of the intertwined structure. The Ohara Museum has an un-finished tassel which shows the f l a t sprang warp (Fig. 74) . The f i n i s h e d a r t i c l e i s constructed by transverse folding across the unworked warps and seaming the side selvedges of the two pieces together. The warp 136 F i g . 73 Category 1 t a s s e l A. DO #509; photo taken from Lothrop, Foshag and Mahler, 1957, #3^ 5 B. Figure on warp cover C. Double-headed figure with misplaced legs (faces and d i g i t s circled) F i g . 74 Unfinished t a s s e l i l l u s t r a t i n g the technique of sprang was used, OM P I 6 3 (photo Mr. Tanaka, courtesy of Nobuko K a j i t a n i ) selvedges, although l y i n g together, are unseamed. The unworked warps are generally covered with a veneer of cross-knit looping. The structure of the tassels d i f f e r s somewhat from the intertwined structure of the openwork fabrics described e a r l i e r . Not only are the tassels made of two interconnected layers of fabric (double cloth) but the actual intertwining also d i f f e r s . The basic structure has pairs of threads twining on both diagonals but each p a i r encloses one thread of two pairs on the opposite diagonal between twining t w i s t s . F i g . 75 shows the structure i n a single layer of f a b r i c . The p r i n c i p l e of omitting twists between intersections of pairs i s the same as i n the variants of intertwined openworks (Figs. 63, 64, 65) but the precise placement of twists d i f f e r s . F i g . 76 shows both layers, using two contrasting sets of threads. The top h a l f of the diagram shows horizontal colour changes (A B A i n the top layer and B A B i n the bottom l a y e r ) . The lower h a l f of the diagram shows v e r t i -c a l colour changes (A B A i n the top layer, B A B i n the bottom layer, reading across the diagram from l e f t to r i g h t ) . Only one structure of intertwining i s used but the joins between v e r t i c a l colour areas i n the lower h a l f of the diagram are very s p e c i f i c and have the appearance of an i r r e g u l a r i t y . V e r t i c a l l y adjacent colour areas are joined by l i n k i n g one thread from each colour area at the boundary. In v e r t i c a l colour areas, threads move diagonally only to the edge of the colour area be-fore changing directions. The diagram i s shown stretched out so that the structure and the joins i n the bottom cloth can be seen. In actual-i t y , only the threads of the top cloth are v i s i b l e ( F i g . 77)' The simul-taneous working of the two layers i s shown i n the r e p l i c a t i o n photograph (F i g . 78). V e r t i c a l l y adjacent colour areas are being worked i n the ,Fig. 75 Structure of Category 1 tassels: an intertwined variant where each pair encloses one thread of two pairs on the opposite diagonal Category 1 tassels: the structure i n two, interpene-t r a t i n g layers with v e r t i c a l j o i n s . The top h a l f of the diagram shows horizontal colour change (A B A -set A i s grey-black and set B i s white). The lower h a l f of the diagram shows v e r t i c a l colour areas and the l i n k s that j o i n them (A B A). A Fig.. ?7 Category 1 t a s s e l d e t a i l : only the top layer i s v i s i b l e A. MM 1914 7-31 51 B. Diagram of top layer only, showing threads i n ver-t i c a l colour areas Replicating the intertwined sprang structure on two interpenetrating layers that i s used i n the Category 1 tassels (photo taken by Alan R. Sawyer) intertwined variety of Category 1 tassels. Because of the complexity of the structure and the huge number of threads used (over 750 i n 0.18 M), every t a s s e l i s not worked i n precisely t h i s way throughout. Deviations or "errors" are f a i r l y numerous although the basic structure, as d i a -grammed, i s mostly adhered to. Several subsidiary structures appear i n the tassels of Category 1 . The s o l i d end borders are generally worked i n two separate layers of i n t e r l a c i n g . Single threads usually interlace obliquely, over(;two, . . under two, to form horizontal r i b s (see Collingwood, 1974, p. 187-8 for the technique) . The border near the top of the t a s s e l i s also usu-a l l y interlaced i n two separate layers, but using groups of threads, either four or eight, as a single u n i t . Two tassels i n Category 1 have lower end borders i n an openwork variation of intertwining, rather than the usual oblique i n t e r l a c i n g (Kanegafuchi, 1956, Vol. 1, No... 10; BME IVC 345)'. As the main body of Category 2 tassels have t h i s struc-ture, i t w i l l be described i n that section. Most of the Category 1 tassels have cross-knit looping covering the unworked warps (AMNH 4l.2/843a may have a woven-top; FM 171134 does not show the top; Kanegafuchi, 1956, Vol. 1, No. 10 and OAG PI63 have no covering). Designs i n the cross-knit looping of a l l the others are worked by carrying along contrasting threads under the background looping and substituting them where required for colour change ( F i g . 79). A small range of colours i s shared by eighteen of the twenty-one tassels (the colours of three tassels cannot be determined). The colours are yellow (including gold), green (including blue-green) and red (including pink and red-brown). Almost a l l two colour combinations are present: yellow on red (6), yellow on green (4), red on green (4) 144 F i g . 79 D e t a i l of cross-knit looped warp cover A. MM 1954 ¥ Gh. 5 599 B. Figure on warp cover red on yellow (3) and green on yellow ( l ) . Yellow i s used frequently but no colour combination dominates. The yellow on red group i s more numer-ous only because i t contains two p a i r s . The f a i r l y extensive use- of green i s noteworthy. Nasca t e x t i l e s from the middle of the Early Inter-mediate Period and from the beginning of the Middle Horizon, about which Ann Rowe writes, tend to use the three primary colours with black outlines. Green i s usually confined to accent areas (1979, P« 117) • The consistent factor i n Category 1 tassels i s the use of two colours - one colour appears as figure and the other appears as ground. There does not seem to be any correlation between the colour combination and the iconography. The l i m i t a t i o n of one figure and one ground colour influences the maze-like style i n Category 1 t a s s e l s . Chakipampa B style t e x t i l e s i n other techniques, where more colours can be introduced without incredible d i f f i c u l t y , generally have more discrete figures which are heavily out-l i n e d and set against a background of a different colour (H .U . Doering, 1967, 183 and O'Neale and Kroeber, 1930, P I . 16). Ann Rowe has noted a tendency of some Chakipampa B designs to f i t rectangular spaces very t i g h t l y , to the extent that the outlining becomes synonomous with the background space (1979, P> 117-8). This i s carried to the extreme i n Category 1 tassels with the s k e l e t a l l i n e s of the figure sketched i n one colour and surrounded by equal size l i n e s of the other colour. A l l space i s f i l l e d by l i n e a r l y described motifs, outlined by the other colour ( F i g . 8Q\). Locating the face and forearm i s the f i r s t step to d i s t i n -guishing the figure i n t h i s d i f f i c u l t style ( F i g . 80B). The d i g i t s on hands and sometimes on feet are terminated i n small squares. Eliminating background motifs and the halo of rays and volutes that surround the f i -gure also makes them more readable. The figure s t i l l can remain elusive B F i g . 80 Maze-like style of patterning i n Category 1 tassels A. drawn from TM 91 .536 B. s i m p l i f i c a t i o n of figure i n A because body parts are sometimes not shown i n ' r e a l i s t i c r elationships. The maze-like style i s further influenced by the complex structure. An analysis of the structure shows that, although threads move on oblique paths, colour changes are st r u c t u r a l along horizontal and v e r t i c a l l i n e s (Fig. 76). Diagonal and curved areas are made by combining the movements for horizontal and v e r t i c a l colour changes. A standard modular size f o r colour areas i s adhered to generally. A horizontal colour area i s usu-a l l y two rows wide and a v e r t i c a l colour area i s generally two i n t e r -workings wide (eight threads of each colour) . Diagonal l i n e s have ser-rated edges as they are made up of modular blocks (Fig. 81). Because the l i n e s that make up the figure and the background are s t r u c t u r a l l y l i m i t e d to the same s i z e , the o v e r - a l l effect i s vibratory. This adds to the d i f f i c u l t y i n seeing the f i g u r e . The style of the looped warp covers i s closer to the Ghakipampa B style of t e x t i l e s i n other techniques, l i k e tapestry (O'Neale and Kroeber, 1930, P I . 16), discontinuous warp and weft (H .U. Doering, 1967, I83) and the "tapestry medallion" t e x t i l e s (A. Rowe, 1979)' Outlining i s somewhat reduced i n scale and, although the figures tend towards f i l l i n g a rectangu-l a r space quite closely, the figure and background space are i n different colours ( F i g . 82). More colours, usually three to s i x , are used. The predominant' background colours, where they can be determined, are red or red-brown (7)» green-blue (3) or yellow (3)' The only correlation to the colours on the tassel bodies i s that a l l the yellow on green tassels and some of the yellow on red tassels have a red background i n the looped area. From t h i s sample, a favoured but by no means standard combination would be a yellow on green tassel body with a predominantly red looped top ( F i g . 83). Fig. 81 Category 1 tassel with diagonal lines built up of rectilinear modules A. FI 13024 (photo taken from d'Harcourt, 1962. P I . 57) B. Figure on tassel body - face and digits circled 149 F i g . 82 Tassel warp cover with more c l e a r l y d i f f e r e n t i a t e d figure A. DO B-509 (photo from Lothrop, Foshag and Mahler, 1957, #3^ 5 B. Figure on warp cover - head and d i g i t s c i r c l e d 150 B F i g . 83 A favoured colour combination i n Category 1 tassels -yellow and green tassel body with a predominantly red warp cover A. TM 91.536 (photo by J . B i r d , courtesy of Alan R. Sawyer) B. Figure on t a s s e l body - head and d i g i t s c i r c l e d Two aspects of the Category 1 ta s s e l s t y l e , r e p e t i t i o n and viewpoint, correlate quite closely. Most tassels show figures from p r o f i l e view-points and a l l p r o f i l e creatures repeat i n horizontal rows. They may he repeated one to s i x times and p a r t i a l repeats, truncated due to lack of space, are not uncommon. Some tassels are divided into two v e r t i c a l r e g i s -ters as well as a number of horizontal rows. Often, though not always, p r o f i l e figures face i n opposite directions i n alternate rows (Fi g . 84). Other p r o f i l e figures appear on the tassels i n Figs. 70-73 an<3- 81-83• One tas s e l figure that repeats i n horizontal rows has a p r o f i l e head but a dorsal view of the body (Fi g . 85A'). The combination of these two viewpoints i s not uncommon i n other Chakipampa B style t e x t i l e s , a l -though i t i s not found i n the Nasca style of the Early Intermediate Period.(Fig. 85B) . A single, central image appeals on a few of the tassels but i t may be rendered from one of several viewpoints. F i g . 86 shows a creature from the dorsal point of view, with appendages extended and head thrown upwards. This positioning i s also used for the "stinger animal", one of the few motifs s p e c i f i c a l l y i d e n t i f i e d as Chakipampa B (see O'Neale and Kroeber, 1930, P I . 16 for such a stinger animal, A. Rowe, 1979, c," p. 117). F i g . 86 i s not necessarily a stinger animal as i t lacks the triangular t a i l . Some single, central images are presented from a fr o n t a l viewpoint ( F i g . 87). Two other tassels (Fig. 74 and AMNH 41.2/843a) are simi l a r to F i g . 87 and seem to be primarily a face, elaborated with face and head ornaments. Only one f r o n t a l view figure i n the sample i s repeated (CAI 1955.1795 i n Lehmann, 1924, P I . l ) . Apart from t h i s one exception, p r o f i l e viewpoints are used for figures that are repeated and f r o n t a l or dorsal viewpoints are used for single images. 152 F i g . 84 P r o f i l e figures are repeated figures i n Category 1 tassels A. MM, Pentland Collection, no number B. Figure on ta s s e l body - head and d i g i t s c i r c l e d 153 F i g . 8 5 A repeating figure with a p r o f i l e head and a dorsal view of the body A. Drawn from a Category 1 t a s s e l , BME IVC 3 ^ 5 B. Figure presented from the same dual viewpoint on a contemporaneous interlocking warp t e x t i l e , LM 8 5 3 8 a (photo courtesy of Alan R• Sawyer) 154 F i g . 86 Single, central image shown from a dorsal viewpoint A. MM 1914 7.31 .51 B. Drawn out figure - head and d i g i t s c i r c l e d G. Simplified version of the same figure (flanking f i gures, top and bottom, are p r o f i l e ) 155 B F i g . 87 Single image shown from a f r o n t a l viewpoint A. photo taken from Kanegafuchi, 1956, Vol. 1, #10 B. Drawing of face with elaborate ornaments and attachments - face i s c i r c l e d Two tassels have scrambled images i n the central areas coupled with repeating images. The scrambled figures have missing and misplaced body-parts and are shown from several viewpoints simultaneously (CAI 1955 1793 and 1794). This practice i s common i n Chakipampa B style t e x t i l e s and many of the images are, at present, v i r t u a l l y indescipherable (see F i g . 1, A. Rowe, 1979, and F i g . 3, Conklin, 1970). The scrambling of figures may derive from Nasca style precedents as Roark reports the scrambling of anatomical d e t a i l s i n Nasca 5 ceramics (1965, P- 26). A l l the viewpoints from which figures are depicted i n the tassels -p r o f i l e , f r o n t a l , dorsal, combined p r o f i l e and dorsal, and multiple -are present i n Chakipampa B style t e x t i l e s of other techniques i d e n t i f i e d by Ann Rowe (1979)• Except for the combined p r o f i l e head and dorsal body viewpoint, they are also present i n Nasca 5 and 6 ceramics (Roark, 1965)« The combined p r o f i l e head and dorsal body viewpoint may be a s p e c i f i c a l l y highland contribution to the Chakipampa B s t y l e . The consistent correlation of p r o f i l e figures with repetitions possibly r e f l e c t s a convention of t h i s and other art styles for repre-senting sub-major or secondary figures. The "angels" on the Gateway of the Sun at Tiahuanaco are shown i n p r o f i l e , are repeated i n registers and flank the central, f r o n t a l and obviously more important "split-eye god". Four of the f i v e tassels with symmetrical figures ( f r o n t a l or dorsal viewpoint) are single and central figures on the f i e l d , perhaps in d i c a -t i n g these figures are r e l a t i v e l y more important. Another s t y l i s t i c feature that may distinguish between major and sub-major themes i n the Category 1 tassels i s the use of space f i l l i n g motifs. Most are angular ray and volute forms and appear on heads, t a i l s , appended to bodies, emanating from mouths and generally f i l l i n g any exterior or i n t e r i o r space. Disembodied heads are less frequently used i n a si m i l a r way. In general, s o l i t a r y figures have the greatest number of these ( F i g . 74, 86 and 87). Scrambled figures also have a great number. Repeated, p r o f i l e figures usually have a reduced number of these additions. Several p r o f i l e figures which are repeated many times on one tassel have no space f i l l i n g motifs ( F i g . 70 and 84). The copious use of rays and volutes occurs i n the Nasca s t y l e , especially Nasca 6 when major themes were treated i n the proliferous style (Roark, 1965 > P« 59)-Technically and s t y l i s t i c a l l y , the tassels of Category 1 show some continuities with the e a r l i e r Nasca t r a d i t i o n . The structure of i n t e r -twining i n the technique of sprang i s , so f a r , only known from the Paracas-Nasca fabrics of the south coast. Certain s t y l i s t i c trends from middle Nasca, l i k e o u t l i n i n g , reducing background space and p r o l i f e r a t i n g ray and volute motifs, can be seen as precursors to the style of Category 1 t a s s e l s . Other aspects of s t y l e , l i k e the p a r t i c u l a r combined p r o f i l e and dorsal viewpoint and the extensive use of the colour green, do not seem to have t h e i r origins i n the Nasca s t y l e . Like much of the icono-graphy, they may be s p e c i f i c Highland contributions to the style of the Category 1 tassels. Category IB Two pairs of tassels share most of the features of Category 1 tassels but, because they are s l i g h t l y elaborated, they have been placed i n a sub-category of 1. They have the same intertwined sprang structure (Fig.- 76) and the same l i n e a r style of patterning. However, both pa i r s incorporate four colours into the body of the tassels. The well-preserved pair of tassels with a cord joining them ( F i g . 88) has a central panel of white on blue flanked by narrower v e r t i c a l panels 158 F i g . 8 8 P a i r of Category IB tassels with four colours A. TM 1959.11.3, photo taken from Collingwood, 1974, F i g . 43 B. Drawing of standing, p r o f i l e figure i n the central panel - head and d i g i t s c i r c l e d of yellow on red (Collingwood, 1974, caption to F i g . 43). The greater number of colours used i n t h i s p a i r i s the simple r e s u l t of warping the two outer sections i n yellow and red and the inner section i n blue and white. Aside from t h i s elaboration i n warping, the main and subsidiary structures appear, from a study of the photograph,, to be i d e n t i c a l to those of Category 1. The other pair introduces four colours into the tassel body i n a way that staggers a f a b r i c analyst. Four complete sets of warps i n con-tra s t i n g colours are superimposed and a l l sets are consistently i n t e r -worked i n the same structure as the tassels of Category 1 ( F i g . 76) • Two sets at a time are worked i n double clo t h , then the two double cloths completely change positions, moving to the opposite face. The blue and green sets of threads that make up one double cloth remain together throughout as do the yellow and red sets of the other c l o t h . At two points, the two double cloths exchange positions, producing the horizon-t a l colour change from green on blue to yellow on red and back to green on blue (Fig/189). I t i s suggested that a f t e r one pattern repeat on the upper two sets i s worked, the loom can be f l i p p e d , bringing the other two sets onto the upper surface i n a convenient position for working. Exchanging the position of the two double cloths remains a tedious, thread by thread undertaking. In experimenting with the technique, I have become convinced that the other obvious alternative (Of'working a l l four sets simultaneously from the same shed i s beyond human ca p a b i l i t y . . . at least mine. This extraordinary p a i r of tassels i s smaller than most (L. 8 l / 2 " xW. 5", d'Harcourt, 1962, p. l6l) and they are constructed d i f f e r e n t l y . Each ta s s e l i s made from h a l f of the sprang warp and the cut ends of the Fig. 89 Pair of Category IB tassels with four colours -quadruple layer intertwined sprang A. MDH 34-14-5-2, photo taken from d'Harcourt, 1962, PI. 58 B. Drawing of figure, rotated ninety degrees to upright position - head and digits circled c e n t r a l l y separated warp are concealed under a woven, rather than looped, warp cover. A hook design i n b i f o l d r o t a t i o n a l symmetry i s embroidered on the cover. The introduction of more colours into the Category IB tassels divides the f i e l d , boldly and simply, into areas. The i n i t i a l effect of these two pairs i s different from the two colour tassels where the whole f i e l d i s uniformly vibratory. The figures are a l i t t l e easier to read because of the d i v i s i o n of space. Within each two colour area of the Category IB tassels, the l i n e a r patterning s t i l l resembles that of Category 1 t a s s e l s . Category 2 Category 2 i s a small group of f i v e tassels (one p a i r i n the group) . Like Category 1, they are made i n intertwined sprang but i t i s an open-work variety of intertwining. They have a very different appearance to Category 1 tassels as they use large scale colour contrasts and simple motifs. I t seems possible that t h i s small group of aberrant tassels may be prototypical versions of the more standardized tassels of Category 1. The openwork structure and the s i m p l i c i t y of the motifs appear to be more closely connected with the openwork fa b r i c s ( F i g . 64, 65) discussed i n the l a s t chapter. This can only be a conjecture because of the lack of associations for both groups of f a b r i c s . I have only examined one of these tassels ( F i g . 90) as the others are i n Japanese museums. Nobuko K a j i t a n i kindly supplied the s l i d e s and the i d e n t i f i c a t i o n of the sprang technique for the examples i n Japan. The structure of the F i g . 90 t a s s e l i s a s p e c i f i c kind of intertwining that combines twisted pairs (to create the open areas) and intertwining 1 6 2 F i g . 90 Category 2 tassel - two interpenetrating layers of openwork intertwining A. MAI #14/2845 B. Branching hook design p a i r s enclosing one thread of one p a i r and one thread of each of two pairs (to create the denser diamonds) . The basic structure i n a single layer of fa b r i c i s shown i n F i g . 91. The colour contrast i s produced by interpenetrating two layers of fa b r i c (double c l o t h ) . The diagram i n F i g . 92 shows the threads of both layers and how they interpenetrate to produce the design. I t can be seen that colour changes take place along diagonal l i n e s and that groups of threads (four on each diagonal) move from top to bottom layer together. The branching hook design i s composed t o t a l l y of diagonals i n accordance with the r e s t r a i n t s of t h i s structure. There are no v e r t i c a l joins because there are no v e r t i c a l colour areas. This t a s s e l uses four colours i n a striped warp arrangement, as the Category IB tassel i n F i g . 88 does. The t a s s e l i n F i g . 93 appears to have the same structure although the photograph i s not quite detailed enough for certain i d e n t i f i c a t i o n . The design i s a series of hooks joined i n a horizontal l i n e . The back-ground area i s i d e n t i c a l and in t e r l o c k i n g . The colour change takes place along oblique l i n e s as groups of threads i n the openwork structure change positions from top layer to bottom layer. A p a i r of tassels from Kanebo, Osaka has a pattern of simple h o r i -zontal l i n e s ( F i g . 9*0 . The tassels appear to be larger than most and there i s a well preserved strap connecting the two. An image of a trophy head with h a i r streaming backwards i s repeated along most of i t s length. The p a i r of tassels i s probably at least p a r t i a l l y made i n the i n t e r -twined variant of F i g . 92 as the s t r i a t i o n s i n the l i g h t e r colour sug-gest i t i s openwork. The f i n a l example ( F i g . 95) i s aberrant i n ,many ways but i t i s closest to the small group of Category 2 t a s s e l s . I t appears, from i t s sheerness, F i g . 91 Intertwined structure of the Category 2 tassel i n • F i g . 90, shown i n a single layer of f a b r i c F i g . 92 totsrtwined structure of the Category 2 tassel l a y e r s ' o f f a o r i r 1 1 6 **•*" A 8 F i g . 93 C a t e g o r y 2 t a s s e l - o p e n w o r k i n t e r t w i n i n g i n t w o i n t e r p e n e t r a t i n g l a y e r s A ' K a j i S n i P h ° t 0 ^ T a n a k a ' C O U r t e s y o f N o b ^ ° B. D e s i g n o f h o o k s F i g . 94 P a i r of Category 2 tassels with horizontal s t r i p e s , Kanebo, Osaka, No number (photo courtesy of Nobuko K a j i t a n i ) F i g . 95 An aberrant t a s s e l with embroidered designs, grouped with Category 2 tassels, OAG K13 (photo courtesy of N©buko Ka j i t a n i ) to tie an openwork variety of intertwining but i t may be only one layer. The bold figure of an eyed creature with three s p i r a l s for a body i s repeated twice on each side of the t a s s e l . Uniquely, the design i s em-broidered on the surface, rather than s t r u c t u r a l l y made. Generally, Category 2 tassels have large scale designs, l i k e hooks, s p i r a l s and bands produced i n an openwork variety of intertwining i n double cl o t h . The structure i s l e s s complex than Category 1 t a s s e l structures because no v e r t i c a l joins are needed for the simple designs that follow oblique or horizontal l i n e s . This technical aspect relates them more closely to the openwork mantles of the previous chapter. The colour range i s similar to that of Category 1 but the warp covers are unusual. Only one uses cross-knit looping ( F i g . 93). One i s woven (Fi g . 90) and the p a i r from Osaka appears to have braided warp covers (F i g . 94). The aberrant example appears to be fragmentary i n t h i s area ( F i g . 95). The shapes and proportions also vary from the homogenous group i n Category 1 and from each other. Although the technical variety might be explained by differences i n l o c a l e , they could also be e a r l i e r versions than the more standardized tassels i n Category 1. Category 3 The t h i r d category of tassels i s a small and highly homogenous group of four. Although no two are so i d e n t i c a l that they could be considered a p a i r , they are amazingly a l i k e . Red and yellow are the predominant colours ( F i g . 96). Space i s divided by jagged horizontal bands i n con-t r a s t i n g structures which are equal and inter l o c k . The tassels are made of two interpenetrating layers of f a b r i c . One layer i s warp-patterned weaving with substitution. The other layer i s a combination 170 F i g . 96 Category 3 t a s s e l - two interpenetrating layers of weaving and oblique twining and i n t e r l a c i n g A. MM 1931 .11.23.13 B. B i r d (?) design on woven warp cover C. Frontal human (?) figure repeated on the woven fa b r i c (arrows indicate s t r i p of oblique i n t e r l a c i n g below the warp cover) of oblique twining and i n t e r l a c i n g . This p a r t i c u l a r structure i s dia-grammed i n F i g . 97. Unlike most Category 1 and 2 tas s e l s , they have a thickness of only one double cloth layer. Although the structure can be analyzed, the technique remains a mystery at t h i s point. I t i s barely possible that the oblique twining and i n t e r l a c i n g was done i n sprang. This assumes that each tassel i s one-half of the o r i g i n a l warp which was cut •"apart through the central, unworked warps. The other p o s s i b i l i t i e s f o r making the obliquely twined/ interlaced fabric are free-end braiding and looped-end braiding. Com-bining any of these techniques with warp-patterned weaving i n double cloth has i t s d i f f i c u l t i e s but perhaps free-end braiding has the fewest impediments as warp take-up does not need to be calculated closely. Several technical features are mildly suggestive that i t was the sprang technique that was combined with weaving i n making these t a s s e l s . At the lower end of the t a s s e l , both sets of warps have heading wefts (Fig. 98). This i s the standard way of preparing a continuously wound warp f o r weaving or sprang and i t suggests there might have been a simi-l a r treatment at the other end of the warp. I f t h i s was the case, then a l l the warps were f i x e d and sprang was the technique of the obliquely twined/interlaced f a b r i c . More tension control i s possible on f i x e d end warps and these tassels are uniformly and evenly worked, suggesting they were worked under tension. La s t l y , the area just below the. warp covers on the woven fabric i s obliquely interlaced, using groups of threads (see arrows, F i g . 96). Tassels i n other categories also have a: s t r i p of oblique i n t e r l a c i n g on grouped threads i n t h i s area. A minor, but possibly s i g n i f i c a n t , d e t a i l concerning t h i s interlaced s t r i p was pointed out by Noemi Speiser. In the Category 3 tassels that F i g . 97 Diagram of the obliquely twined and interlaced structure used i n Category 3 tassels ( i n t e r l a c i n g threads are paired) g. 9 8 Inside a Category 3 tassel: the warps of both layers of f a b r i c are mounted on heading cords (MM I 9 3 i . l l . 2 3 . i 5 ) have i n t e r l a c i n g below the warp covers, the r i b s , or l i n e s of f l o a t s , are v e r t i c a l . In a l l the other tassels with t h i s feature, the r i b s are horizontal. Noemi Speiser suggests that the v e r t i c a l r i b s are more easily produced i n a non-sprang method. I t might be possible that the obliquely twined/interlaced f a b r i c was made on stretched but not f i x e d warp ends, using a technique that resembles the one used for the t i e s on the wrap-around s k i r t s ( F i g . 54). There i s s t i l l the unresolved problem of how t h i s technique could be -synchronized with the woven f a -b r i c . In Category 3 tassels, the space i s divided simply using the contrast i n colour and structure, of large scale, horizontal motifs that i n t e r l o c k . The f a b r i c i s highly patterned within these simple d i v i s i o n s . The twined/interlaced f a b r i c has a structural design of concentric diamonds which are off-set i n successive rows. The woven fabric uses the substi-tution of contrasting warps f o r the designs. Frontal, human (?) f i -gures wearing headdresses appear on two tassels (Fig. 96 and MM I 9 3 i . l l . 2 3 . i 5 ) . P r o f i l e birds appear on the other two tassels ( F i g . 99 and MM 1931.11.23.14). The woven structure allows more figurative d e t a i l and more varied colour use than the intertwined sprang of Category 1 and 2 t a s s e l s . Usually, f i v e colours are used i n the woven fabric and three i n the twined/interlaced one. The favoured colours are red and yellow (several shades of each) with accents of blue and white. A yellowish green i s occasionally used. There are two styles of warp covers i n the Category 3 group. Two have cross-knit looped covers with exactly the same design (Fig. 99 and 100). Two others have supplementary weft-patterned covers (Fig. 96 and 101) which also have the same design of a b i r d - l i k e creature with 99 Category 3 t a s s e l A. DO B510, photo taken from Lothrop, Foshag and Mahler, 1957, #3^ 6 B. Figure on t a s s e l cover C. Figure of b i r d repeated on woven section, rotated ninety degrees (heads c i r c l e d ) rPjTj'i i miitijt tm U l l l l H l l l l l h i l i l I.I P i g . 100 Category 3 t a s s e l A. MM 1931.11.23.15 B. Figure on warp cover C. Frontal human (?) repeated on woven section (heads cir c l e d ) 177 F i g . 101 Category 3 t a s s e l - warp cover i s woven with supple-mentary warp patterning A . MM 1931.11 .23.14 B. B i r d (?) figure on warp cover C. Figure of b i r d repeated on woven sections, rotated ninety degrees (heads c i r c l e d ) a p r o f i l e head and a dorsal viewpoint of the "body. Category 3 tassels relate to the previous groups i n shape and some de t a i l s of f i n i s h i n g . They deviate most strongly i n using a woven struc-ture for one of the layers of the double c l o t h . The structural and s t y l i s t i c s i m i l a r i t y within t h i s small group of tassels i s so great that i t i s tempting to think that they were made by the same workshop. Although none of them are i d e n t i c a l , they a l l share iconography with at least one other t a s s e l . Category 4 There are f i v e tassels (two pairs) i n Category 4-. They are made of two interpenetrating layers of woven f a b r i c . They do not use the tech-nique of sprang at a l l , except perhaps for a small section of oblique i n t e r l a c i n g just below the warp cover ( F i g . 102, see arrows). Like Category 3 tassels, these are only one thickness of double cloth. The presence of two i d e n t i c a l pairs i n the group suggests that two tassels were made onf'.one warp which was cut apart i n the middle. There i s no r e a l problem to weaving two tassels on the same warp. Andean weav-ing t r a d i t i o n a l l y uses a continuously wound warp to produce four selvedge f a b r i c . The area of terminal weaving i s generally set away from the warp selvedges, indicating that some weaving was done from both ends of the warp. In the case of the Category 4 f a b r i c s , a complete tassel could have been woven at each end of the warp. When the tassels were complete, a small area of oblique interlaced sprang could have been done on the remaining central warps, producing the structure that i s v i s i b l e just below the looped warp covers. The design organization i s quite standard within the group. The major s p a t i a l d i v i s i o n s are v e r t i c a l . Two types of patterning alternate 179 F i g . 102 Category 4- t a s s e l made of two interpenetrating layers of woven fab r i c A. CAI 1955 1791 B, photo Chicago Art I n s t i t u t e B. Figure on warp cover C. Woven figure repeated i n three r e g i s t e r s (head and d i g i t s c i r c l e d ) i n f i v e r e g i s t e r s . Although the major s p a t i a l d ivisions are bold, each area i s re-divided with small scale patterning (Fig. 103) • This approach to design i s shared with the Category 3 t a s s e l s . The colours are r i c h and varied with strong contrast between major motifs. The range of colours i s reds, yellows, blue and white. Similar colours are used on the warp covers. A l l the warp covers are constructed i n cross-knit looping. Their designs vary from p r o f i l e heads, repeated eight times ( F i g . 103), to f a m i l i a r Chakipampa B style creatures who are repeated four times (Fi g . 102) to a single and singular creature ( F i g . 104). Most tassels i n a l l categories have looped warp covers and these vary only i n being s l i g h t l y larger and more elaborate iconographicall/y. In summary, Category 4 tassels share an over-all shape, the double cloth construction and the looped warp cover with tassels of other cate-gories. In structure and technique, they vary widely from the i n t e r -twined sprang of Category 1 and 2. They relate most closely to the Category 3 tassels which are p a r t i a l l y woven. The design organization and colour range are also closer to Category 3« Iconography The iconography of the Chakipampa B style has not been f u l l y des-cribed. Dorothy Menzel does i d e n t i f y several of the more distenct f i -gures found on ceramics i n the a r t i c l e "Style and Time i n the Middle Horizon" (1964). These figures - the "Ayacucho serpent", the "humped animal" and the "stinger animal" (p. 11-15) - may be among the figures on the tassels but the scant, often p a r t i a l i l l u s t r a t i o n s on sherds i n her references do not provide d e f i n i t i v e sets of characteristics f o r 181 103 Category 4 t a s s e l A. AMNH 41.2/5347 B. Head on warp cover, repeated eight times C. Woven figure repeated i n three r e g i s t e r s C. S c r o l l motif repeated i n two reg i s t e r s 182 Fig. 104 Category 4 tassel A. CAI 1955 1792 B, photo Chicago Art Institute a. Creature on warp cover G' YZen« fY?f i : i l a m a ? ) r e P e a t e d ^ three registers (head and digits circled) 183 f o r each creature. The r i g i d l y geometric renderings on the tassels, the variations i n viewpoint and the amount of de t a i l i n g also make matching ceramic and t e x t i l e iconography d i f f i c u l t . A serpent-like figure appears on only one tassel ( F i g . 86) and, although i t i s double-headed l i k e some of the Ayacucho serpents, the heads are p r o f i l e , humanoid ones ( F i g . 105A) rather than toothed, whiskered heads shown from a dorsal viewpoint (Menzel, 1964, p. 15). The only other creature with a long body has many legs and a human-like head and i s , at most, a distant r e l a t i v e ( F i g . 105B) . The humped animal, which appears i n the Nasca 9 s t y l e as well as the Ghakipampa A and B styles (1964, p. 28), varies considerably i n the ceramic references given by Menzel (pp. 11, 15, 28, 35)' I t i s l i k e l y that at least two tassels, and possibly more, have a variation of the humped animal ( F i g . 106A,B). Although the backs are not curved, both share a number of general characteristics with the humped animal, i n c l u -ding posture and gesture. The stinger animal i s described as a ventrally extended animal with ray appendages, a triangular t a i l and an elongated 'stinger' i n front (Menzel, 1964, p. l l ) . Two tassels have ventrally extended animals. F i g . 107A i s the most clear-cut representation of the stinger animal i n the tassel sample, with i t s tongue-like stinger that projects from i t s mouth. Though F i g . 107B i s v e n t r a l l y extended, i t has neither t a i l nor stinger and may be a different creature. Ann Rowe (1979, Note 13, p- 123) has pointed out a clear, tapestry representation of the stinger animal (O'Neale and Kroeber, 1931, PI- 16) . The diagonal l i n e s emanating from the muzzle relate t h i s creature to several t a s s e l figures. One tassel body has what may be the front h a l f or head of the stinger animal (F i g . 108A) . 184 6 •Fig. 105 Tassel figures that may r e l a t e to the "Ayacucho serpent" A. MM 1914 7.31 .51 B. TM 91.537 1 (heads and d i g i t s c i r c l e d ) 185 107 Ventrally extended t a s s e l figures A. OM P 163 - clearest representation of the stinger animal i n the t a s s e l sample (face, d i g i t s and stinger circled) B. MM 1914.7.31'51 - ventrally extended creature with rays but which may not be the stinger animal (face and d i g i t s c i r c l e d ) F i g . 108 Tassel figures r e l a t i n g to the head of the stinger animal, as represented i n O'Neale and Kroeber, 1930, P I . 16. A. drawn from Kanegafuchi, 1956, Vol. 1, #10 B. warp cover, AMNH kl .2/84-3A (Diagonal l i n e s emanating from muzzles are ci r c l e d ) 188 The woven warp cover of another tassel has a f r o n t a l view of a ' " goggle-eyed face with diagonal l i n e s emanating from the muzzle ( F i g . 108B). The looped warp covers on the p a i r of tassels i n F i g . 83 have s i m i l a r faces but the photograph i s too i n d i s t i n c t f o r a close comparison. Another feature of the stinger animal i s the.triangular t a i l . While not exactly triangular, two tassels have figures with barbed t a i l s ( F i g . 109A.B). None of the above can be cer t a i n l y i d e n t i f i e d with the stinger animal which may i t s e l f be a composite of attributes which can be separated and recombined i n d i f f e r e n t ways. Perhaps the variety of t r a i t - s h a r i n g f i -gures i n the Ghakipampa B style arises from a s t y l i s t i c practice of com-positing figures from discrete elements of independent meaning. For instance, F i g . 109B has a barbed t a i l l i k e the stinger animal and the posture, stance and general features of the humped animal. Ann Rowe has noted a small f i l l e r element resembling a bird's head which appears on a variety of Chakipampa B t e x t i l e s (1979, P« 121). This element often occurs inside the body of zoomorphs (O'Neale and Kroeber, 1930, P I . 19; A. Rowe, 1979, Figs. 8, 10) or attached to arms, legs and beaks (d'Harcourt, 1962,-P1. 4; H.U. Doering, 1967, 183) or as an i s o l a t e d motif (A. Rowe, 1979, P« 121 and F i g . 1, p. 114; O'Neale and Kroeber, 1930, P I . 18; H.U. Doering, I 9 6 7 , I83). The representations vary from having just an eye and mouth to having d e t a i l s of headdress, ear and hai r included. This element occurs a number of times on the ta s s e l bodies and on the warp covers. The ventrally extended figure,' (Fig . llOA) has an eared version complete with hair set inside i t s body. I t also has eared versions with and without headdresses attached to elbows and f i l l i n g i n exterior spaces. One warp cover of a Category 4 t a s s e l has a version with headdress inside i t s body (Fi g . HOB) and another has F i g . 109 Tassel figures with barbed t a i l s which may relate to the triangular t a i l of the stinger animal A. FM 171134, drawn from O'Neale and Kroeber, 1930, P I . 21 B. drawn from Kanegafuchi, 1956, Vol. 8, #76 (barbed t a i l c i r c l e d ) A F i g . 110 Various representations of t h e ^ I f d head' element which i s probably a trophy head motif A. MM I914.7.3I.5I -trophy heads below arms B. CAT. 1955 1791 - warp cover, trophy head inside body G. AMNH 41.2.5347 - warp cover, trophy head repeated eight times t). TM I959.II.3 - trophy head appended at waist E. 0M. PI 63 - trophy.heads'appended to stinger and chin (trophy heads circled) i s o l a t e d eared versions with headdresses repeated eight times on the warp cover ( F i g . HOC) . A Category IB tassel with an upright, p r o f i l e human has an abbreviated version attached to the belt (Fig. HOD). The 'stinger animal' ( F i g . HOE) has heads terminating the stinger and appended to the chin. The overlapping uses of variously detailed versions suggest they a l l refer to the same theme. The most detailed versions (ear, h a i r , headdress) suggest that the theme i s a disembodied head. Some of the uses, l i k e the attachment to appendages or b e l t , are very reminiscent of trophy head uses i n the Nasca t r a d i t i o n of the Early Intermediate Period (Roark, 1965, p. 23, 27). Menzel writes that the trophy head i s a theme of the Chaki-pampa sty l e which i s introduced from the Nasca drainage (1964, p. 11, 29) and Roark describes a similar range of v a r i a b i l i t i e s i n trophy head depic-t i o n i n the Nasca 5 period (1965, p. 27). I t seems quite possible that a l l the versions of t h i s f i l l e r element and independent motif are slack-jawed trophy heads. The trophy head theme 'may be elaborated i n two tassels which have a s i m i l a r design. Both tassels have dominant eared heads attached to two rudimentary appendages (F i g . 111A,B). Both also have abstract emanations from the mouth area. That t h i s figure appears twice i n t h i s l i m i t e d sample suggests i t i s a standard theme. I t i s possible i t derives from the so-called "full-bodied trophy head", a sub-major theme which Roark finds i n Nasca 6 ceramics (1965, p- 45). A looped warp cover ( F i g . 86) may also have trophy heads animated by the addition of legs. The trophy head motif has a long history i n ancient Peru. There are some depictions i n Chavinoid art (Roe, 1974, Feature 60, F i g . 4b, F i g . 28e) but there are a great many more on the South Coast from Early Horizon 9 onwards. Although i t i s not possible to make a d e f i n i t i v e interpretation of trophy heads, Alan Sawyer has suggested that they are Tassel figures that may be "full-bodied" trophy heads A. MDH No. 34.14-5.2, drawn from d'Harcourt, 1962, P I . 58 B. photo from T e l l o , 1959, F i g . 139, p. 306 (the head i s the primary motif; a speech s c r o l l and two abbreviated appendages are attached to each head) part of a f e r t i l i t y cult iconography, that they were "a ceremonial means of gathering the l i f e - or soul-fource of enemies to be used f o r the bene-f i t of the c o l l e c t o r group" (1966, p. 122). The trophy head continues to be a frequent element i n Nasca and Ghakipampa B style images. A number of p r o f i l e creatures [share the gesture of an outstretched appendage .(Fig. 112). This gesture may express the idea of a trophy head. In e a r l i e r phases of Nasca a r t , figures with outstretched appen-dages often grasp trophy heads or hanks of h a i r , which appear to be shorthand symbols for trophy heads (Sawyer, 1966, F i g . 205, 207, 213, 215)• The extended arm may be a s t i l l more condensed rendition of the same idea. Two images on the same bottle of k i l l e r whale d e i t i e s with extended arm are the same (Sawyer, 1966, F i g . 205) except one grasps a trophy head and the other i s empty-handed. In the tassel figures, there are varying degrees of s i m i l a r i t y among the group and some are, nor.'-doubt, variations on the same theme, perhaps the humped animal. F i g . 112 shows s i x r e l a t e d figures which may contain an i m p l i c i t reference to the trophy head theme. F i g . 112D i s found on the warp covers of two tass e l s . Three more elaborate figures with outstretched forelegs are very i n d i v i d u a l ( F i g . 113) and do not relate closely to the others i n F i g . 112. However, they may also contain a reference to the trophy head idea. A p a r t i c u l a r element that i s used more frequently than trophy heads needs to be described. The ray motif i n four-rpart r a d i a l symmetry i s a f i l l e r element with Nasca antecedents (Menzel, 1964-, p. 28). I t s renderings and used i n the tassel sample are variable. I t can have curved or angular rays emanating from a focal point that i s a c i r c l e , a square, intersecting l i n e s or a single l i n e . In the lower one-third 194 F i g . 1 1 2 Zoomorphs with outstretched foreleg, possibly a l l r e l a t e d to the humped animal and implying trophy head A. MM 1 9 5 4 ¥ Gh5 599 B. MM, Pentland Collection, no number C. drawn from Kanegafuchi, 1 9 5 6 , Vol. 8, # 7 6 "D. DO B-509 and Kanegafuchi, I 9 5 6 , V o l . 8, # 7 6 (warp covers) E. GAI 1955 1 7 9 3 B, warp cover F. CAP 1955 1791 B, warp cover (outstretched foreleg c i r c l e d ) 195 F i g . 1 1 3 Three zoomorphs with outstretched forelegs which possibly imply trophy head A. FM I 7 I I 3 4 , drawn from O'Neale and Kroeber, 1 9 3 0 , P I . 21 B. FI 1 3 0 2 4 , drawn from d'Harcourt, 1962, P I . 5 7 C. -CAI 1955.1792B, warp cover (outstretched foreleg c i r c l e d ) of a badly deteriorated t a s s e l ( F i g . 114), a curved ray with c i r c u l a r center i s flanked by angular ray motifs that are arranged around two intersecting l i n e s . This motif often terminates extensions from mouths, from heads and headdresses ( F i g . 115) and from bodies and t a i l s . Any l i n e a r appendage, excepting arms and legs, seems a candidate for such a terminal. This ray motif occurs as an isol a t e d design on other Ghakipampa B t e x t i l e s (A. Rowe, 1979. F i g . 1, p. 114) although i t does not occur alone on the t a s s e l s . On one t a s s e l , a distorted version i s used to f i l l an i n t e r i o r body space ( F i g . 116). The headdresses which are worn by over h a l f of the tassel figures a l l take the form of one-half of a ray motif i n r a d i a l sym-metry centred on two intersecting l i n e s . Some figures are outlined with halves of the ray motif ( F i g . 117)• The ray motif i n r a d i a l symmetry i s the most frequently repeated design element i n the tassel sample, although trophy heads are also re-peated to a lesser degree. The sum of the minor uses of the ray motif amounts to what Roark terms " p r o l i f e r a t i o n " i n the Nasca style (1965, P- 2) . The most frequent use of the p r o l i f e r a t e d motifs i n the Nasca style i s extensions to the gold forehead ornaments and facemasks which become i n -creasingly complex during the l a t t e r h a l f of the Early Intermediate Period. The actual gold ornaments from Paracas and Nasca often have embossed serpent terminals (Sawyer, I960, P I . I l l , IV; Moseley, 1978, P. XXXI and F i g . 44). I t appears that the ray and volute may be l a t e r , more abstract motifs f o r designating the serpent projections from masks. Serpentine appendages, attached to various parts of the body, particu-l a r l y the mouth, are frequently depicted i n the e a r l i e r art of Paracas and Nasca. The ray and volute are used s i m i l a r l y i n the Ghakipampa B s t y l e , suggesting that rays and volutes have the same group of associations 197 F i g . 114 Varied ray motifs A. AMNH 4l.2/843a B. Angular ray motif arranged around two intersecting l i n e s G. Curved ray motif with c i r c u l a r centre 198 F i g . 115 Ray motifs terminating extensions from mouths, heads and headdresses A. TM i 9 5 9.ll .3 B. OM P I63 (rays are c i r c l e d ) Fig,. 116 Distorted variant of a ray motif f i l l i n g an i n t e r i o r body space (FI 13024, drawn from d'Harcourt, 1962, P I . 57) (ray i s cir c l e d ) 200 F i g . 117 Ray motifs as outlines (CAI 1955-1792B, warp cover) that serpents had i n the e a r l i e r styles of Paracas and Nasca. In the Ghakipampa B style of the tassels, the ray motif and trophy head are used i n varied, multiple ways and i t i s l i k e l y that both the images and the p r o l i f e r a t e d uses derive from Nasca antecedents'-'.1 Three tassels appear to have faces that are engulfed i n the p r o l i f e r a t e d motifs of rays and volutes ( F i g . 74, 87 and 114) that derive from the gold face masks and headdresses. F i g . 74 also has numerous trophy heads. Another class of elements i n the tassel sample i s the geometric. Except for the ray and volute forms, most Category 1 tassels do not have any. One t a s s e l i n Category IB has a 'S% or hook i n b i f o l d , r o t a t i o n a l symmetry, which i s embroidered on the warp cover ( F i g . 89). In Category 2, two tassels have large scale hook designs, one i n s l i d e symmetry (Fi g . 118A) and the other i n translation and b i f o l d , r o t a t i o n a l symmetry ( F i g . 118B) . The large scale space d i v i s i o n on Category 3 tassels may be a further abstraction of F i g . 118B, l i m i t e d by the r e c t i l i n e a r combination of struc-tures i n those tassels ( F i g . 118c). In Category 4, two pairs of tassels have large scale, angular hooks i n b i f o l d , r o t a t i o n a l symmetry ( F i g . 118D) and the f i n a l Category 4 t a s s e l has a series of l e s s angular hooks ( F i g . 118E) i n the same configuration as F i g . 118B. The Category 3 tas-sels a l l have designs of outlined diamonds which are s t r u c t u r a l l y produced i n oblique twining and i n t e r l a c i n g . The shapes, the symmetries and the interlocking of figure and ground i n the hook designs of F i g . 118;-' are reminiscent of the fabric structure images of Paracas Cavernas which were discussed i n Chapter 3 of t h i s paper. I suspect that the branching hook motif of F i g . 118A derives from a three strand braid which also has the branching configuration of slide symmetry. The hooks i n F i g . 118B, E have the same symmetries and configurations as the twisted or twined strands. 202 F i g . 118 Geometric motifs i n the tassel sample A. MAI 14/2845 B. 0M Gl6 G. A l l Category 3 tassels - MM 1931.11.23.13, 14 & 15 and DO B-510 D. CAI 1955.1791 and 1792, A and B E. AMNH 41.2/5347 A few figures on the tassels are depicted more n a t u r a l i s t i c a l l y . One figure appears to be an ungulate, possibly a llama (Fig. 119A) . A convoluted ray appendage i s attached to i t s mouth. Two pairs of Category 4 tassels have long rlegged, eared animals which are possibly llamas ( F i g . 119B) The warp-patterned weaving technique changes the rendering so d r a s t i c a l l y that i t i s d i f f i c u l t to compare F i g . 119A and B with any certainty. The Category 3 tassels have simple b i r d - l i k e and humanoid figures on the ta s s e l bodies ( F i g . 120A.B). Two tassels have clearer depictions of upright, p r o f i l e figures that appear to be ess e n t i a l l y human. These figures have considerable ana-tomical veracity compared to the animated trophy heads of F i g . 111. One squat figure carries a s t a f f ( F i g . 121) which i s similar to the fea-thered s t a f f s of Nasca 7 (p.c. L. Dawson) pots ( F i g . 122). The other figure also appears to have a hand clenched around a pointed s t a f f ( F i g . 123). Several features of t h i s figure seem to be drawn from the Tiahuanaco-related ceremonial s t y l e , a practice i n the Chakipampa B style that i s l i m i t e d to design d e t a i l s (Menzel, 1964," p. 14). Both the divided eye and the ray attached to the mouth (speech s c r o l l ? ) are features of the Tiahuanaca-related style of Conchopata (p. 20-21). Another possible human i s i n a horizontal or f l o a t i n g p o s i t i o n ( F i g . 124A) . This figure also has a ray attached to the mouth and an eye marking. A f i n a l example, also horizontal, i s very schematic but i t also has a divided eye and a projection from the mouth ( F i g . 124B) . . The warp covers of two Category 3 tassels have Tiahuanaco-related creatures with t o o t h - f i l l e d maws (Fig . 125). The prominent teeth,, the hooked eye marking, the elaborate headdress and a possible wing element on i t s back re l a t e i t loosely to variants of the Tiahuanaco s t a f f bearers, 204 F i g . 119 Tassel figures that appear to "be ungulates, possibly llamas A. CAI 1955'1795, warp cover, drawn from Lehmann, 1924, P I . 1) B. CAI 1955.1791 and 1792, A and B (head and feet circled) 205 F i g . 120 Warp-patterned figures on Category 3 tassels A. b i r d - l i k e figure - DO B-510 and MM 1931.11.23.14 B. human-like figure - MM 1931.11.23.13 and .15 (heads are c i r c l e d ) A g. 121 Tassel with a s t a f f holding human (upside down i n the upper hal f ) and a scrambled figure A. GAI 1955 1793 B, photo Chicago Art I n s t i t u t e B. human figure holding a s t a f f which i s at the r i g h t ; figure i s rotated 180 degrees (head and d i g i t s c i r c l e d ) g. 122 Nasca 7 pot with a warrior carrying a feathered s t a f f (LM 4-8951) 208 F i g . 120 P r o f i l e human carrying a s t a f f , along the l e f t side TM 1959.11.3 (head and d i g i t s are c i r c l e d ) B F i g . 12% Two horizontal figures that may "be humanoid A. TM 91.536 B. BME IVG 34-5 (head and d i g i t s c i r c l e d ) F i g . 125 Warp cover figure possibly related to Tiahuanaco-Wari staff-bearers (MM 1931.11.23.15 and DO B-510) such as those on o f f i c i a l Wari tunics (Sawyer, 19^3, F i g . 3, 6). In some tass e l s , zoomorphic and humanoid attributes are combined i n the same figure. The representations of humanoid heads (and trophy heads) generally have a nose that protrudes beyond the plane of the face. Several t a s s e l figures have a human-like head attached to a non-human body (F i g . 126A, B). One tassel has an anatomically scrambled body and two faces, one human and one non-human ( F i g . 126c). Several tassels have figures that are t o t a l l y scrambled. Heads and legs can be found but not i n n a t u r a l i s t i c relationships. Faces and disjointed legs appear i n the lower t h i r d of the tassel body i n F i g . 12?. These scrambled d e t a i l s are exactly repeated i n the upper t h i r d . The central section remains quite indescipherable. F i g . 121 also has a scrambled figure i n the lower h a l f . In the Ghakipampa B s t y l e , the lack of clear cut iconography and the practices of scrambling figures and overwhelming them with p r o l i f e r -ated motifs suggests there i s not a pantheon of major d e i t i e s . The f i -gures are composed of and elaborated by elements which may have indepen-dent meanings. In fact, some Chakipampa B t e x t i l e s , l i k e the tapestry medallion ones described by Ann Rowe, are primarily decorated with a catalogue of elements (1979, F i g . l ) which can occur i n conjunction with diverse figures. The complex interplay of culture forces which Menzel postulates f o r the'period of the Chakipampa B style (1964, p. 10, 66-68) may be ref l e c t e d i n the art style which i s more a st i t c h i n g together of conventions and ideas from diverse sources than a presentation of d i s -t i n c t mythical beings. Chonological Implications of the Tassel C l a s s i f i c a t i o n The c l a s s i f i c a t i o n of thirty-nine tassels into four categories 212 F i g . 126 Zoomorphs with humanoid heads A. MM 1914.7.31.51 B. TM 91-537 G. DO B-509 - humanoid head on the r i g h t and zoomor-phic head i s on the l e f t (heads and d i g i t s c i r c l e d ) Fig. 127 on the basis of structure i s reinforced by differences i n design organ-i z a t i o n , style and iconography. The l i m i t e d and undocumented nature of the sample does not allow any insight into contemporaneous regional s t y l e s but i t i s suggested that Category 1 and 2 tassels are e a r l i e r than Category 3 and 4. The largest group, Category 1, was described f i r s t as they appear to be the standard type. Technically, Category 2 tassels have the closest a f f i n i t y with f a b r i c s of the Early Intermediate Period as they continue to use openwork variations of intertwined sprang. Category 1 tassels may have evolved s l i g h t l y l a t e r than Category 2 tassels as they use a very complex and specialized structure f o r v e r t i c a l colour j o i n s . This v e r t i c a l j o i n i s necessary for the devastatingly detailed patterns of the Chakipampa B s t y l e . Category 2 tassels remain somewhat problematic. The group i s too small and too diverse to be sure they represent a temporal and not a regional difference. Intertwined sprang has a s i g n i f i c a n t presence among Nasca f a b r i c s . I t seems reasonable to think that the double cloth variants of Category 1 and 2 tassels develop out of the intertwined openworks of that period. The o v e r - a l l patterning and f i l l i n g of a l l background space continues some Nascoid propensities. The iconographic range of zoomorphs, humans and combinations share some characteristics with the standard iconography of the Chakipampa B st y l e , although few are 'dead ringers'. Certain Nasca derived motifs, l i k e the ray and trophy head, and the s t y l i s t i c practice of p r o l i f e r a t i n g motifs are strongly i n evidence. A few elements derived from the ceremonial or Conchopata s t y l e , l i k e the divided eye and the ray attached to the mouth, are also present. The l i m i t e d use of colours and the use of one of those colours as both outline and background r e l a t e to trends observable i n middle Nasca t e x t i l e s (Bennett, 1954, F i g . 72). A l l i n a l l , Category 1 and 2 tassels relate more strongly to the practices of the Early Intermediate Period than the other categories of t a s s e l s . Category 3 and 4 tassels have a change i n structure and technique. Both introduce warp-patterned weaving into the body of the t a s s e l . ; _r : Category 3 tassels may re t a i n the use of sprang for one of the layers but t h i s cannot be ascertained. Category 4 tassels may have a narrow area of interlaced sprang just below the warp covers. Space di v i s i o n s are bold and large scale geometric motifs, l i k e hooks, are generally present. Within the broad s p a t i a l d e f i n i t i o n s , smaller motifs are r e -peated, varying i n orientation and colour placement. Space i s not cram-med with maze-like patterning but ordered through major contrasts and re-lie v e d through minor contrasts. Figurative motifs are unelaborated representations of birds, humans and llamas (?) . The colour range of Category 3 and 4 tassels i s larger and more varied. Many shades of gold and red with strong accents of blue are used. The underlying s p a t i a l or-ganization of v e r t i c a l and horizontal r e g i s t e r s and the colour use are closer to the sty l e of Wari tunics which date primarily to Middle Horizon 2. The tassel warp covers are a l i t t l e more homogenous from Category 1 through 4. The warp covers of Category 3 and 4 continue to have the Chakipampa B iconography and, i n general, the cross-knit looped structure. In f a c t , they become larger i n size and the representations become more elaborate. I t i s as i f the iconography on the Category 1 tassel bodies i s forced upward, into t h i s secondary cL-3T@cl} E L S the main part of the tas s e l gives way to change i n s t y l e , technique and iconography. 216 The tassels seem to have been a south coast specialty, perhaps a badge of o f f i c e or status. The largest, most standardized group, Category 1, dates to Middle Horizon IB (p.c. Lawrence Dawson). The number of tassels located f o r t h i s study tapers o f f as changes i n technique, style and iconography tend toward Tiahuanaco-Wari influences. Tassels of Category 3 and 4 may extend into Middle Horizon 2. This tapering off i n number, coupled with a change i n style and technique, suggests the supplanting of one culture by another, a change supported by ceramic chronology (Menzel, 1964). So f a r as my present research can determine, the technique of sprang f a l l s into disuse at t h i s point, as the south coast Nasca culture comes under the sway of the highland Wari culture. FOOTNOTE Since I completed writing t h i s section, Alan Sawyer has located three more tassels i n the private c o l l e c t i o n of Anton Roeckl of Munich. Each tassel o r i g i n a l l y had a mate but they have been sold to a New York' c o l l e c t o r . The tassels share a l l the characteristics of Category 1. CHAPTER V Summary and Conclusions A search of the l i t e r a t u r e and a number of museum collections has turned up a considerable body of fa b r i c s manufactured i n the sprang technique. The sprang technique has a v e r i f i e d existence from Early Horizon 7 through Middle Horixon IB (c. 700 B.C. - 650 A.D.). The report of preceramic sprang*from Asia (Engel, I963) i s considered to be i n error. The geographic area over which sprang fa b r i c s have a confirmed appearance i s almost t o t a l l y l i m i t e d to the South Coast area. Many pieces without provenience can be t i e d s t y l i s t i c a l l y to ones with ai' confirmed South Coast provenience. D'Harcourt (1962, P I . 54B) does c i t e a Central Coast locale for one fragment of intertwining. The e a r l i e s t sprang fa b r i c located for t h i s study i s an i n t e r l i n k e d band from C e r r i l l o s . I t i s the sole f a b r i c of i t s type. I t i s expec-ted that there are many others but that they .have not been collected or commented on because of t h e i r p l a i n appearance. Interlinked sprang i s p a r t i c u l a r l y used f o r headdresses, most of which can be dated to Early Horizon 9 on the basis of a comparable style to excavated examples with ceramic association. During Early Horizon 9» there i s a great emphasis on openwork fabrics with structural patterning and i n t e r l i n k e d sprang, along with gauze weaving and knotting, appears to be a favoured technique. The c u l t u r a l importance of fabric making i s demonstrated i n t h i s group as the designs themselves are de-r i v e d from images associated with f a b r i c making. Interlinked sprang seems to be primarily associated with the people who made incised, resin-painted ceramics l i k e those found at Ocucaje and Paracas Gavernas. A single example of i n t e r l i n k i n g , a striped bag, i s associated with a mummy bundle from Paracas Necropolis. Interlaced sprang i s used for garment t i e s f o r l o i n c l o t h s and s k i r t s during a period that may span Early Horizon 9 and Early Interme-diate Period 2. An unusual method of warp suspension i s indicated by -the obliquely worked rows. The warp was probably folded i n h a l f trans-versely and then manipulated from the loops formed i n the f o l d i n g . This variant method i s indicated only f o r the t i e s as a l l other sprang fabrics appear to be worked i n horizontal rows. These d i s t i n c t i v e interlaced t i e s are, so f a r as I presently know, associated only with embroidered garments of the type found at Paracas Necropolis and some less opulently embroidered ones from other south coast s i t e s . Several different types of interlaced sprang bags appear to date between Early Horizon 9 and the l a t t e r part of the Early Intermediate Period. They are few i n number and unspecialized i n structure and design. No conclusions can be drawn from the very l i m i t e d sample. One group of s i m i l a r l y finished, interlaced bags appears to have been made expressly f o r carrying coca leaves and lime. Two of the bags are shaped during the making to have three pockets, a p a r t i c u l a r bag shape that remains associated with coca use to the present day. Intertwined sprang appears to have continued i n use longer than i n t e r l i n k i n g and i n t e r l a c i n g . The few associations indicate i t i s con-nected with the groups who made incised , r e s i n painted ceramics i n the l a s t stages of the Early Horizon and slip-decorated pottery during the Early Intermediate Period. This i s not surprising as there i s no abrupt change from one ceramic type to the other at several s t r a t i g r a p h i c a l l y excavated s i t e s (Sawyer, 1966, p. 96). The intertwined openworks are bea u t i f u l l y executed and some are so large that they were made i n two sections. The fab r i c sizes indicate both headdresses and shrouds were made of intertwined sprang. The most unusual sprang fabrics are a series of double cloth tassel The majority exhibit the Chakipampa B style which i s the l a s t pahase of the Nasca st y l e i n Middle Horizon IB. Within the tassel sample, i t i s possible to follow a change i n technique, which i s p a r a l l e l e d by a chang i n style and iconography. The f i r s t two categories continue the Nasca technique of sprang and the intertwined structure. Some s t y l i s t i c con-ventions also relate these groups to the Nasca style although the icono-graphy draws on a number of highlard (pre-Wari) themes. The t h i r d and four t h fcategories of tassels use warp-patterned weaving, a highly f a -voured technique among the highland Wari. The change i n technique i s para l l e l e d by changes i n colour, design organization and iconography which are tending toward the Wari style of the Middle Horizon. The categorization of the tassels, which i s based on structure and technique appears to be as diagnostic of change as the style and iconography. Like other culture t r a i t s that have been used to construct chrono-logies, the technique and structures of sprang'are l i m i t e d temporally and geographically. Various types of sprang f a b r i c s are s p e c i f i c to certain phases of cultures and can be associated with pottery types and styles i n other t e x t i l e s . The occurrence of sprang fabrics i n ancient Peru i s summarized i n Table 2. The technique of sprang appears to f a l l into disuse about the time the highland Wari culture spreads to the coast i n Middle Horizon 2. In the course of examining a number of museum co l l e c t i o n s , I have found SOUTH COAST OCCURRENCE OF THE SPRANG TECHNIQUE IN THE THESIS SAMPLE MIDDLE HORIZON EARLY INTER-MEDIATE PERIOD EARLY HORIZON 110 9 8 7 6 5 -900 A.DT "550 A . D T A.D, B.C. •400 B.CT •1000 B.Cf w 0) w fl) M nj « CD US* w O -P HH B S5 CD n3 faO M u @ bp s s @ § E H ft ! S O I — I V — o M E H ( D g s 55 -P TABLE I I only two fab r i c s of l a t e r periods that relate to the sprang fabrics of t h i s study. Neither are true sprang as they both have wefts. They do make use of the sprang p r i n c i p l e of reciprocal•shedding which allows symmetrical f a b r i c to be b u i l t from both ends of the warp toward the central, terminal area. One i s a small bag of two-strand warp twining from Huarato, Acari Valley (LM l 6 - 1 1 0 9 2 d ) which was found with a fabric with Late Intermediate Period c h a r a c t e r i s t i c s . The other sprang-like f a b r i c i s a Late Horizon bag (BMS 3:124-73) made i n t r i - a x i a l i n t e r l a c i n g Like I s l interlaced sprang, the threads move obliquely on two axes. The horizontal weft, placed i n the shed of each row, l i e s on the t h i r d a x i s . The lack of true sprang f a b r i c s i n the l a t e periods underlines how sp e c i f i c techniques are to the originating culture. The Wari and subsequent groups who peopled the south coast favoured quite different techniques of fa b r i c making and these supplanted many of the methods used i n Paracas and Nasca f a b r i c s . Technical studies such as t h i s one, which uses a large number of unassociated f a b r i c s , are not capable of revealing the entire picture of Peruvian culture h i s t o r y . However, there i s a considerable applica-t i o n for t h i s information. At least a general chronological and geo-graphic placement of 'orphan' sprang fabrics i n museum coll e c t i o n s should now be possible from no more than a technical analysis. Also, i t i s possible to suggest some in t e r n a l sequencing, based on an evo-l u t i o n i n technique and structure, when the sample of specialized f a -b r i c s i s large enough. Iconographic comparisons have been used to c l a r i f y the c u l t u r a l associations of various parts of the sprang sample. Now the technical groupings, p a r t i c u l a r l y of complex fabrics l i k e the tas s e l s , can be used to expand the iconographic repertory of a culture phase. Technical 'association' i n the case of complex structures and esoteric techniques i s as certain a diagnostic as many s t y l i s t i c c r i t e r i a i n present use. As a t e x t i l e typology by culture continues to develop, studies such as t h i s one, which follows one technique and i t s variant structures, may be useful i n c l a r i f y i n g contacts, movements and contin-u i t i e s between groups. The presence of the sprang technique on the South Coast from 700 B.C. to 650 A.D. suggests a strand of continuity i n t h i s time period. A cataclysmic culture change a f t e r Middle Horizon IB i s indicated by the t o t a l disappearance of the sprang technique. FOOTNOTE 1 The term, t r i - a x i a l i n t e r l a c i n g , was suggested hy Peter Collingwood. BIBLIOGRAPHY Bennett, Wendell C. 1 9 5 4 Ancient Arts of the Andes. Museum of Modern Art, New York. Bennett, Wendell and Junius B i r d 1 9 4 - 9 Andean Culture History. American Museum of Natural History Handbook Series, No. 1 5 , New York. B i r d , Junius B. 1 9 5 2 "Before Heddles Were Invented." Handweaver and Craftsman, Vol. 3 , No. 3 , Summer. 1 9 6 3 a ^Preceramic Art from Huaca P r i e t a , Chicama Valley." Nawpa Pacha, 1 , pp. 2 9 - 3 4 . I n s t i t u t e of Andean Studies, Berkeley. 1 9 6 3 b "Technology and Art i n Peruvian Tex t i l e s . " IN Technique  and Personality, by M. Mead, J . B i r d and H. Himmelheber. The Museum of Primitive Art Lecture Series, No. 3 -New York. 1 9 6 8 "The Use of Computers i n the Analysis of Textile Data: S p e c i f i c a l l y Archaeological Fabrics from Peru." IN Computers  and Their P o t e n t i a l Applications i n Museums, pp. 1 2 7 - 1 4 - 5 . Arno Press, New York. B i r d , Junius and Louisa Bellinger 1 9 5 4 Paracas Fabrics and Nazca Needlework, 3 r d C. B.C. - 3 r d C. A.D. The Textile Museum, Washington, D.C. Bir d , Junius and Joy Mahler 1 9 5 1 - 2 "America's Oldest Cotton Fabrics." American Fabrics, Vol. 7 3 , No. 2 0 , pp. 7 3 - 7 7 . New York. Cason, Marjorie and Adele Cahlander 1 9 7 6 The Art of Bol i v i a n Highland Weaving . Watson-Guptill Publications, New York. Carrion-Cachot, Rebecca 1 9 3 1 "La Indumentaria en l a Antigua Cultura de Paracas." Wira Kocha, Vol. 1 , No. 1 , Lima Peru. Collingwood, Peter 1 9 7 4 The Techniques of Sprang. Faber and Faber, London. Conklin, William J . 1 9 7 0 "Peruvian Te x t i l e Fragment from the Beginning of the Middle Horizon." Textile Museum Journal, Vol. I l l , No. 1 , pp. 1 5 - 2 4 . The Textile Museum, Washington. BIBLIOGRAPHY, con't Gonklin, William J . 1 9 7 1 "Chavin Textiles and the Origins of Peruvian Weaving." Text i l e Museum Journal, Vol. I l l , No. 2 , Dec. The Textile Museum, Washington, D.C. 1 9 7 5 a "An Introduction to South American Archaeological Textiles with Emphasis on Materials and Techniques of Peruvian Tapestry." IN Archaeological Textiles; 1 9 7 4 Proceedings of  the Irene Emery Roundtable on Museum Textiles (ed. P a t r i c i a L. Fiske), pp. 7 7 - 9 2 . Textile Museum, Washington, D.C. 1 9 7 8 '^The Revolutionary Weaving Inventions of the Early Horizon." Nawpa Pacha, 1 6 , pp. 1 - 1 2 . I n s t i t u t e of Andean Studies, Berkeley. 1 9 7 9 "Moche Tex t i l e Structures." IN The Junius B. B i r d "vV Pre-Columbian Textile Conference (eds. Ann P. Rowe, Elizabeth P. Benson, Anne-Louise Schaffer), pp. 1 6 5 - 1 8 4 . The Textile Museum and Dumbarton Oak's. Cordy-Collins, Alana 1 9 7 9 "Cotton and the Staff God: Analysis of an Ancient Chavin T e x t i l e . " IN The Junius B. B i r d Pre-Columbian Textile Conference (eds. Ann P. Rowe, Elizabeth B. Benson, Anne-Louise Schaffer), pp. 5 1 - 6 0 . The T e x t i l e Museum and Dumbarton Oaks. Dawson, Lawrence E. 1 9 7 9 "Painted Cloth Mummy Masks of l e a , Peru." IN The Junius B.  B i r d Pre-Columbian Te x t i l e Conference (eds. Ann P. Rowe, Elizabeth P. Benson, Anne-Louise Schaffer), pp. 83-104. The Textile Museum and Dumbarton Oaks. Dwyer, Edward B. 1 9 7 9 "Early Horizon Tapestry from South Coastal Peru." IN The  Junius B. B i r d Pre-Columbian Textile Conference (eds. Arm P. Rowe, Elizabeth P. Benson, Anne-Louise Schaffer), pp. 6l-82. The Textile Museum and Dumbarton Oaks. P. Chronology and Iconography of Late Paracas and Early Nasca Textile Design. Doctoral Dissertation, University of C a l i f o r n i a , Berkeley, 1 9 7 1 . "The Chronology and Iconography of Paracas Style T e x t i l e s . " IN The Junius B. B i r d Pre-Columbian Textile Conference (eds. Ann P. Rowe, Elizabeth P. Benson, Anne-Louise Schaffer), pp. 1 0 5 - 1 2 8 . The Textile Museum and Dumbarton Oaks. Dwyer, Jane n .d. 1979 Emery, Irene I 9 6 6 The Primary Structures of Fabrics: An I l l u s t r a t e d C l a s s i f i c a t i o n . The Tex t i l e Museum, Washington, D.C. BIBLIOGRAPHY, con't Engel, Frederic 1963 "A Preceramic Settlement on the Central Coast of Peru: Asia, Unit 1." Transactions of the American Philosophical  Society, n.s., Vol. 53, Part I I I . Philadelphia. Garaventa, Donna 1981 "A Discontinuous Warp and Weft Textile of Early Horizon Date." Nawpa Pacha, 1 9 , pp. I 6 7 - I 7 6 . I n s t i t u t e of Andean Studies, Berkeley. Grieder, Terence 1978 The Art and Archaeology of Pashash. University of '• Texas Press, Austin and London. Harcourt, Raoul d' 1 9 6 2 Textiles of Ancient Peru and Their Techniques, edited by Grace G. Denny and Carolyn Osborne. University of Washington Press, Seattle. Kanegafuchi Boseki Kabushiki Kaisha 1956 Textiles of Pre-Inca, from Burying Grounds i n Peru. Kanegafuchi Spinning Company, Osaka. King, Mary E. n.d. Te x t i l e s and Basketry of the Paracas Period, l e a Valley, Peru Doctoral Dissertation, University of Arizona, Lapiner, Alan 1 9 7 6 Pre-Columbian Art of South America. Harry N. Abrams. Lehmann, Walter and H. Ubbelohde-Doering 1924 The Art of Old Peru. Verlag Ernst Wasmuth, Tubingen. Lothrop, Samuel D., W.F. Foshag and Joy Mahler 1957 Pre-Columbian A r t . Phaidon Press, New York. Menzel, Dorothy 1 9 6 4 "Style and Time i n the Middle Horizon." Nawpa Pacha, 2, pp. 2 - 1 0 5 . I n s t i t u t e of Andean Studies, Berkeley. Menzel, Dorothy, John Rowe and Lawrence Dawson 1964 "The Paracas Pottery of l e a : A Study i n Style and Time." The University of C a l i f o r n i a Publications i n American  Archaeology and Ethnology, V o l . 5 0 . University of C a l i f o r n i a Berkeley. Moseley, Michael E. 1 9 7 5 The Maritime Foundations of Andean C i v i l i z a t i o n . Cummings Archaeology Series, Cummings Publishing Company, Menlo Park. 1978 Peru's Golden Treasures. F i e l d Museum of Natural History, Chicago. 228 BIBLIOGRAPHY, con't O'Neale, L i l a 1934 "Peruvian Needleknitting." American Anthropologist, n.s. 36> pp. 405-4-30. 1937 "Archaeological Explorations i n Peru, Part I I I , Textiles of of the Early Nazca Period." F i e l d Museum of Natural History,  Anthropology Memoirs, Vol. 2, No. 3- Chicago 194-2 "Textile Periods i n Ancient Peru, I I : Paracas Cavernas and the Grand Necropolis." University of C a l i f o r n i a Publications  i n American Archaeology and Ethnology, Vol. 39, No. 2, pp. 143-202. University of C a l i f o r n i a , Berkeley. O'Neale, L i l a and Bonnie Jean Clark 1948 "Textile Periods i n Ancient Peru, I l l s The Gauze Weaves." University of C a l i f o r n i a Publications i n American Archaeology and Ethnology, Vol. 40, No. 4 . University of C a l i f o r n i a , Berkeley. O'Neale, L i l a and A .L. Kroeber 1930 "Textile Periods i n Ancient Peru." University of C a l i f o r n i a  Publications i n American Archaeology and Ethnology, Vol. 28, No. 2, pp. 23-56. University of C a l i f o r n i a , Berkeley. Roark, Richard P . „ 1965 "From Monumental to P r o l i f e r o u s i n Nasca Pottery." Nawpa Pacha, 3t PP- 1-92. I n s t i t u t e of Andean Studies, Berkeley. Roe, Peter ^ 1974 "A Further Exploration of the Rowe Chavin Seriation and i t s Implications f o r North Central Coast Chronology." Studies  i n Pre-Columbian Art and Archaeology, No. 13. Dumbarton Oaks, Washington. Rowe, Ann P. 1973 "A Heritage of Colour: T e x t i l e Traditions of the South Coast of Peru." Catalogue, T e x t i l e Museum. Washington, D.C. 1977 Warp-Patterned Weaves of the Andes. The Textile Museum, Washington, D.C. 1979 "A Late Nasca-Derived T e x t i l e with Tapestry Medallions." B u l l e t i n of the Detroit I n s t i t u t e of Arts, Vol. 57, No. 3, pp. 114-123. Detroit I n s t i t u t e of A r t s , Detroit. 1980 ^ T e x t i l e s from the B u r i a l Platform of Las Avispas at Chanchan. Nawpa Pacha, 18, pp. 81-148. I n s t i t u t e of Andean Studies, Berkeley. Rowe, John H. 1962a "Stages and Periods i n Archaeological Interpretation." Southwestern Journal of Anthropology, Vol. 18, No. 1, pp. 40-54. BIBLIOGRAPHY, con't 1 9 6 2 b Chavin Art.: An Inquiry Into I t s Form and Meaning. The Museum of Prim i t i v e A r t , New York. I 9 6 7 "Form and Meaning i n Chavin Art." IN Peruvian Archaeology: Selected Readings (eds. J . Rowe and D. Menzel), pp. 7 2 - 1 0 3 . Peek Publications, Palo A l t o . Rowe, John H. and Dorothy Menzel, eds. 1967 Peruvian Archaeology: Selected Readings. Peek Publications, Palo A l t o . Sawyer, Alan R . i 9 6 0 "Paracas Necropolis Headdress and Face Ornaments." Workshop Notes, Paper No. 2 1 , May. Textile Museum, Washington, D.C. I 9 6 3 " T i a h u a n a c G Tapestry Design." Textile Museum Journal, Vol. 1 , No. 2 , pp„ 2 7 - 3 8 . Textile Museum, Washington, D.C. 1 9 6 6 Ancient Peruvian Ceramics: The Nathan Cummings Collection . The Metropolitan Museum of Art. Seiler-Baldinger, Annemarie 1 9 7 7 "General Introduction to the Literature on South American Ethnographic Textiles since 1 9 5 0 . " IN Ethnographic Textiles  of the Western Hemisphere: 1 9 7 6 Proceedings of the Irene  Emery Roundtable on Museum Textiles (ed. P a t r i c i a L. Fiske), pp. 1 7 - 3 4 . The Te x t i l e Museum, Washington, D.C. Shepard, Anna 0 . 1 9 4 8 "The Symmetry of Abstract Design with Special Reference to Ceramic Decoration." Contributions to American Anthropology  and History, No. 17, PE- 2 1 0 - 2 9 2 . T e l l o , J u l i o C. 1 9 3 1 "Un Modelo de Escenografia P l a s t i c a en e l Arte Antiguo Peruano." Wira Kocha, Vol. 1 , No. 1 , pp. 8 7 - 1 1 2 , Lima. 1 9 5 9 Paracas: Primera Parte . Publicacion del Projecto 8 b del Programa 1 9 4 1 - 4 2 de The In s t i t u t e of Andean Research of New York. Empresa Grctfica T. Scheuch, S.A., Lima. T e l l o , J u l i o C. y T. Mejia Xesspe 1 9 7 9 Paracas, I I Parte: Cavernas y Necropolis. Universidad Nacional Mayor de San Marcos y de The In s t i t u t e of Andean Research de Nueva York. Ubbelohde-Doering, Heinrich 1 9 6 7 On the Royal Highways of the Incas. New York. 2 3 0 BIBLIOGRAPHY, con't Van Stan, Ina 1 9 6 9 - 7 0 ^ S i x Bags with Woven Pockets from Pre-Columbian Peru." Nawpa Pacha, 7 - 8 , pp. 1 7 - 2 8 . I n s t i t u t e of Andean Studies, Berkeley. Vreeland, James 1 9 7 7 "The V e r t i c a l Loom i n the Andes, Past and Present." IN Looms and Their Products: 1 9 7 6 Proceedings of the Irene • Emery Roundtable on Museum Textiles (ed. P a t r i c i a L. Fiske), pp. Wallace, Dwight T. 1 9 6 2 " C e r r i l l o s , an. Early Paracas Site i n l e a , Peru." American  Antiquity,. Vol. 2 7 , No. 3 , pp. 3 0 3 - 3 1 4 . 1 9 7 5 "An Analysis of Weaving Patterns: Examples from the Early Periods i n Peru." IN Archaeological Textiles: 1 9 7 4  Proceedings of the Irene Emery Roundtable on Museum Textiles (ed. P a t r i c i a L. Fiske), pp. 1 0 1 - 1 1 6 . The Textile Museum, Washington, D.C. 1 9 7 9 "The Process of Weaving Development on the Peruvian Coast." IN The Junius B. B i r d Pre-Columbian Textile Conference (eds. Ann P. Rowe, Elizabeth P. Benson, Anne-Louise Schaffer), .pp. 2 7 - 5 0 . The Textile Museum and Dumbarton Oaks. Weitlaner-Johnson, Irmgard n.d. Twine P l a i t i n g : A H i s t o r i c a l , Technical and Comparative Study. M.A. Thesis. University of C a l i f o r n i a , Berkeley, 1 9 5 0 . Yacovleff, Eugenio and Jorge C. Muelle 1 9 3 2 "Una Exploracion en Cerro Colorado." Revista del Museo  Nacional, Vol. 1 , No. 2 , pp. 3 1 - 1 0 2 . Lima, Peru. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0095501/manifest

Comment

Related Items