Keith Muckelroy: Methods, Ideas and Maritime Archaeology PDF

Preview

Summary

J Mari Arch (2009) 4:67–82 DOI 10.1007/s11457-009-9045-2 ORIGINAL PAPER Keith Muckelroy: Methods, Ideas and Maritime Archaeology Matthew Harpster Published online: 5 May 2009 Ó Springer Science+Business Media, LLC 2009 Abstract Between his graduation from the Department of Archaeology at Cambridge U...

Description

Accelerat ing t he world's research.

Keith Muckelroy: Methods, Ideas and Maritime Archaeology Matthew Harpster

Cite this paper

Downloaded from Academia.edu 

Get the citation in MLA, APA, or Chicago styles

Related papers

Download a PDF Pack of t he best relat ed papers 

Discovery and Survey and of a 17t h–18t h Cent ury Shipwreck near Drumbeg, NW Scot land: an i… John K McCart hy (2008). Appendix C: Bibliography for Ident ified Dut ch Shipwreck Sit es [Voorcompagnieën & VOC], in T h… Wendy van Duivenvoorde Marble Disc Opht halmoi from T wo Shipwrecks off t he Israeli Coast Ehud Galili

J Mari Arch (2009) 4:67–82 DOI 10.1007/s11457-009-9045-2 ORIGINAL PAPER

Keith Muckelroy: Methods, Ideas and Maritime Archaeology Matthew Harpster

Published online: 5 May 2009 Ó Springer Science+Business Media, LLC 2009

Abstract Between his graduation from the Department of Archaeology at Cambridge University in 1974 and his death in 1980, Keith Muckelroy’s work and ideology were crucial in promoting an alternative research methodology in maritime archaeology. Instead of a particularist or historiographic approach, methods prominent both then and now, Muckelroy’s methodology was grounded in the foundations of the prehistoric archaeology he learned under Grahame Clark and David Clarke at Cambridge, and the basic tenets of New Archaeology maturing in the United States during the 1970s. This paper, which elucidates Muckelroy’s methods and research, is neither a complete biography nor an exhaustive study of his ideas. Although unpublished letters, papers and notes were studied in archives at Cambridge University and the National Maritime Museum, there is still much more to be learned from many of his former colleagues and their memories—only a handful of those individuals were consulted during the creation of this work. Nevertheless, this paper was written in the hope that by understanding Muckelroy’s ideas, and placing them in the larger framework of the discipline of archaeology, maritime archaeologists who are attempting to pursue a variety of approaches may find inspirations, models and, perhaps, questions that still need to be answered. Keywords Muckelroy
Maritime archaeology
Statistical analysis
The New Archaeology
CUUEG

Education and Research Goals During his first year in Jesus College at Cambridge University in 1970, Keith Muckelroy was not a diver, and may have had only a faint impression of the potential of maritime archaeology. Over the next 3 years, however, his interest in the subject certainly blossomed. In 1971, Muckelroy was taught to dive by Jeff Dubery, the diving officer of the Cambridge University Underwater Exploration Group (CUUEG), he became the M. Harpster (&) Eastern Mediterranean University, Famagusta, Mersin-10, Turkey e-mail: [email protected]

123

68

J Mari Arch (2009) 4:67–82

organization’s assistant diving officer in 1972 and, by 1973 and his appointment as their diving officer, he was organizing their third season of work mapping and studying the submerged harbour structures at Apollonia, Libya (CUUEG SOC.92.2, 18; 92.1.1; 92.1.2; 92.13.4; Anonymous 1973, p. 3; Muckelroy 1978, p. viii). Nicholas Flemming, with his friends, started the CUUEG in 1957, and although it was not exclusively an archaeological group—it drew participants from architecture, engineering, and the natural sciences—it maintained an archaeological focus nevertheless (Flemming pers. comm.; CUUEG SOC.92.12, 6). This early relationship between the CUUEG and Muckelroy was particularly important for two reasons. First, as Muckelroy was the organization’s only archaeology student in these years he was thus an ideal member, forging an association that continued well after his graduation. Secondly, his membership in the CUUEG was his introduction to the community of divers throughout other universities in the United Kingdom. Thus, the CUUEG was most likely the vehicle that introduced Muckelroy to divers from Aston University, in Birmingham, who had located and preliminarily surveyed Kennemerland in 1971 (Forster and Higgs 1973, p. 291). Kennemerland was a Dutch East Indiaman which sank off Stoura Stack in the Out Skerries, Shetland Islands, in 1664, and became the archaeological site at the center of Muckelroy’s research. Indeed, Muckelroy participated in the second season of work at Kennemerland in 1973 and co-authored the second interim report on the site published in 1974 (Price and Muckelroy 1974). Following his graduation in June of 1974, Muckelroy established a hectic schedule of researching, publishing and excavating. That summer he moved to St Andrews University, Scotland, to begin his research assistantship under Colin Martin in the newly-established Institute of Maritime Archaeology, and participated in the third season of work at Kennemerland (Martin pers. comm.; Price and Muckelroy 1977; Muckelroy 1978, p. viii). An expedition he jointly organized with Hugh Feilden to survey the harbour works at Salamis, Cyprus, was interrupted and postponed indefinitely by the outbreak of hostilities on the island in July of that year (CUUEG SOC.92.2; Anonymous 1974a). Between 1975 and 1978, Muckelroy published seven journal articles, appearing in The International Journal of Nautical Archaeology, World Archaeology, Britannia and Progress in Underwater Science, he began his role as the Deputy Director of the Mary Rose excavation, and he participated in the fourth excavation season at Kennemerland in 1976. In 1977 in particular, Muckelroy finished his assistantship at St Andrews University, moved to a new position as Archaeologist Diver at the National Maritime Museum’s Archaeological Research Centre, began his doctoral research in October of that year and, in December, finished the manuscript of his book, Maritime Archaeology (Martin and McCarthy pers. comm.; Muckelroy 1978, p. viii, ix). By the opening of 1978 Muckelroy’s qualifications were already well known throughout the United Kingdom and his career, even before earning his doctorate, was well established. He continued his work at Kennemerland with a fifth season in 1978, and was the archaeological director of a CUUEG expedition to investigate a Bronze Age site near Salcombe, Devon, discovered the previous fall (Muckelroy 1978; Muckelroy and Baker 1979, p. 189). His book was published by Cambridge University Press, and was hailed as ‘‘an important book written by a highly respected British maritime archaeologist’’ and representing ‘‘the coming of age of underwater archaeology’’ (Arnold 1980, p. 80; Van Doorninck Jr 1981, p. 228). The years 1979 and 1980 were just as busy, leading a CUUEG expedition to Plitra, Greece, conducting further investigation of a second Bronze Age site near Langdon Bay, Devon, publishing three more articles, and preparing the manuscripts for Archaeology Under Water and Discovering a Historic Wreck (Muckelroy 1980, p.

123

J Mari Arch (2009) 4:67–82

69

100). This progress, Muckelroy’s potential, and his career were cut short on September 8th, 1980, however, when he was killed in a diving accident in Loch Tay. He was 29 years old. What defined Muckelroy as an archaeologist was not merely a prolific list of publications and field work in a short career, but his application of particular archaeological methodologies under water. Key to this identity was an article he published while at St Andrews in 1976, describing an extensive research project investigating new methods of underwater site investigation and interpretation (Muckelroy 1976, p. 281). His plan comprised three key parts. The first part was a flow diagram that depicted five processes through which a ship is transformed from an organized collection of items into the assemblage of artifacts observed on the sea bed (Fig. 1). All five processes, from the wrecking of the vessel to the excavation of the material, represent forces that continually disorganize the ship and its contents; Muckelroy called these forces scrambling devices (Muckelroy 1976, p. 286). Three of these processes also represent forces that remove material from the wreck site as well; these he called extracting filters (Muckelroy 1976, p. 283). By deconstructing and understanding the characteristics of each process, Muckelroy argued that it would be possible to reverse the effects of each hypothetically, reorganize or add artifacts, and thus recreate the original composition of the ship. This diagram, therefore, was an interpretive tool that could guide Muckelroy and other archaeologists through their analyses of underwater wreck sites. As this 1976 article represented this diagram’s first publication, however, this tool was still a hypothesis. As a result, the second part of Muckelroy’s research included determining the individual characteristics of each filter and testing the accuracy of the entire

Fig. 1 Muckelroy’s Flow Diagram depicting the forces that affect a ship as it sinks and becomes an archaeological site on the sea floor. From Muckelroy 1976, Fig. 6

123

70

J Mari Arch (2009) 4:67–82

system. Presumably, if this tool proved valid, and resulted in increased understanding of the vessel prior to its loss, then Muckelroy would pursue the third part of his project. This part, which was likely the ultimate goal of his research, was to determine if the repeated application of this interpretive tool to a number of sites would generate trends in the collated data. Such trends may represent a series of general rules governing the wrecking and deposition processes that could aid in the interpretation of other, more poorly preserved underwater sites.

Muckelroy and the Department of Archaeology at Cambridge University Muckelroy’s 1976 proposal seems to be the first explicit discussion of his project, but its roots reach back to his time in the Department of Archaeology at Cambridge University. Elements of this project appear in his first article about Kennemerland, published while he completed his fourth year. As he wrote, the objectives of their second season were to glean as much information about the material carried by the ship as possible, recreate the processes by which the ship had sunk and disintegrated, and decipher the factors that led to the differing levels of artifact preservation (Price and Muckelroy 1974, p. 258). Indeed, he was already attempting to model the wrecking process in this article: having grounded on Stoura Stack and split in two, Muckelroy proposed that the bow section of Kennemerland was drawn by the current southwards while the stern was pushed by the wind to the north, leaving a trail of cannon as it disintegrated (Price and Muckelroy 1974, pp. 260, 261). The Department of Archaeology at Cambridge, however, was more than just a location at which these ideas were formed. It was, instead, an inspiration for much of his work. Only a year after his graduation, for example, Muckelroy was applying the methods he had learned at Cambridge and was publishing the results. Muckelroy’s second article, ‘‘A systematic approach to the investigation of scattered wreck sites’’ was published in 1975 in The International Journal of Nautical Archaeology and it contains his adaptation of the statistical methods he learned under the tutelage of David Clarke and Robert Chapman (Chapman pers. comm.; Muckelroy 1975a, p. 189). Specifically, Muckelroy demonstrated how matrix analysis and numerical taxonomy may be used as a means of interpreting scattered wreck sites, and how such sites, as a result, are worthy of archaeological investigation (Muckelroy 1975a, pp. 173, 174). Matrix analysis is a statistical process which determines the degree of similarity, correlation, or difference within a set of entities. The entities may be anything in particular, such as animals or plants, but each entity in the set must also be quantified by a common set of discrete characteristics—this is the process of numerical taxonomy. Clarke, both in class and in print, argued that numerical taxonomy was applicable to archaeology because at the elemental level, artifacts could be entities characterized with quantifiable attributes just like a plant or an animal (Clarke 1968, p. 518, 585, Fig. 138). As a result, subjective and intuitive methods of description could be eliminated. For Muckelroy, these statistical tools were appropriate because he defines a scattered wreck site as one lacking coherent hull structure (Muckelroy 1975a, p. 173). Lacking the hull as a reference, it is unknown how scrambled the artifacts on the sea bed may be, and what they may represent. Thus, Muckelroy needed to determine correlations and associations between different artifact classes across the archaeological site. In other words, each artifact class became an entity, and the entities were compared to determine degrees of similarity. To do this, Muckelroy first segregated the artifacts into discrete classes which had no overlap and represented approximately 75% of all the artifacts recovered by that date (Muckelroy

123

J Mari Arch (2009) 4:67–82

71

1975a, p. 176). Second, because he is simply determining similarities between classes, each class was characterized by one of three attributes—absent in a grid square, present in a square, or abundant. The last was applied if the number of artifacts in a class in a square was greater than the average number per square across the entire site (Muckelroy 1975a, p. 178). To analyze the resulting matrix, he compared the interpretations created by the matching coefficient, the coefficient of Jaccard, and the Robinson/Brainerd similarity index. The results, displayed in his Fig. 9, highlight the artifact clusters that form through his analysis and the degrees of similarity among those clusters (Fig. 2) (Muckelroy 1975a, p. 185). In short, his preliminary conclusions demonstrated two things. First, that the clusters of artifact classes in his analysis seemed to match their role on board ship—the largest cluster was the cargo, while the peripheral clusters were equipment, provisions or personal items (Muckelroy 1975a, p. 186). Second, the relative homogeneity of the clusters

Fig. 2 Two sets of attribute clusters that form through Muckelroy’s statistical analysis of the finds from Kennemerland. Groups I, II, III, IVa, IVb and IVc represent Bellarmine flagons, green bottle glass, pewter bottle tops and three classes of clay pipes. Groups V, VIII and IX represent lead shot, bronze sheeting and bronze nails, while VI and XI are personal possessions and bones. From Muckelroy 1975, Fig. 9

123

72

J Mari Arch (2009) 4:67–82

indicated that there was little post-depositional sorting or scrambling of the artifacts on the sea bed (Muckelroy 1975a, p. 186). Initially, his first conclusion seems self-defeating. Muckelroy used complicated statistical analysis to demonstrate what common sense may dictate—that the most numerous and consistent class of artifacts on site is the cargo, while the rarer classes are secondary items. But it should be understood that this study is not an isolated work. It is, instead, part of his larger research plan investigating his flow diagram and this article is his first attempt at testing its components. By undertaking this particular study on Kennemerland, Muckelroy demonstrated that it is possible to determine quantitatively, and not just intuitively, which artifact classes on a scattered wreck site are part of the cargo or the personal items. The fact that his conclusions are similar to an intuitive approach neither invalidates his work, nor buttresses the conclusions of the latter; it merely illustrates that two different methods generated similar results. What is particular about Muckelroy’s statistical approach is that it, unlike an intuitive method, is inherently replicable and testable—it may be applied again and again to numerous wreck sites and the results from each will be equally valid and comparable. As a result, this single study was a first step in Muckelroy’s much larger goal of collating data from a large collection of sites and, through the analysis of that data, determining if any general rules describing site formation under water emerge. His second conclusion is also part of this larger plan. For the first time, Muckelroy is able to argue the degree to which the environmental characteristics surrounding the Kennemerland wreck site affected the scattering and scrambling of its artifacts. In this case, he demonstrated that despite an environment active enough to destroy the hull, the original organization of artifacts within the hull was not greatly altered. The cargo and the personal items, initially segregated while in use, remained segregated on the sea floor. Muckelroy’s 1976 article published in World Archaeology, ‘‘The integration of historical and archaeological data concerning an historic wreck site: the ‘Kennemerland’’’, contains additional adaptations of Clarke’s ideas. The flow diagram presented in this work, for example, is Muckelroy’s version of Clarke’s system presented in his book (Clarke 1968, p. 73, Fig. 11). Whereas Clarke created a general system describing inputs and outputs that affect the creation and continual alteration of an archaeological site, Muckelroy created a system specific for a site underwater. Similarly, the overall focus of this article was to demonstrate the value of integrating archaeological and historical data as equally valid sources of information (Muckelroy 1976, p. 280). Muckelroy’s attempt to join the two, and not subsume archaeology within history, reflects Clarke’s view that archaeology and history are academic equals. Muckelroy’s adoption and adaptation of Clarke’s methods demonstrates the influence he had during the early stages of Muckelroy’s career. Indeed, it also demonstrates how Clarke was hoping to affect the archaeological establishment at large. Clarke’s best-known publication, Analytical Archaeology from 1968, was his attempt to establish a series of scientific and analytical methods of interpreting archaeological data, and an attempt to demonstrate how those methods may be the basis of a central paradigm in archaeology. Essentially, it was a guide for other archaeologists to follow and, in this case, Muckelroy was one of those archaeologists. As a result, Clarke had a comprehensive influence on Muckelroy’s application of statistics and systems theory, and provided him with frameworks within which analyses could be made and ideas tested. But these frameworks lacked substance. It was the work of Grahame Clark, the Disney Chair of Archaeology at Cambridge, instead of David Clarke, who inspired the content of a substantial portion of Muckelroy’s long-term research.

123

J Mari Arch (2009) 4:67–82

73

Clark’s 1939 book, Archaeology and Society, was written to describe the processes of archaeology, its aims, and its social value (Clark 1939, p. viii). Following an introduction to the practice of prehistoric archaeology, the book’s structure is a series of chapters that guide the reader through the various stages of archaeological research, from methods of discovering sites to the reconstruction of culture. The book culminates by addressing archaeology’s role in shaping public perceptions, and its role in education. It was the third chapter of Clark’s book that seems to have impressed itself upon Muckelroy. This chapter, titled ‘‘Preservation’’, investigates the influence of environment on rates of terrestrial site preservation (Clark 1939, p. 50). It discusses the general factors that tend to preserve or destroy organic and inorganic material and, when necessary, provides specific examples from sites around the world. It does not contain any specialized detail or information, and the content does not distinguish this book from other archaeological texts of the era. What may have drawn Muckelroy’s attention is that this chapter could not have been replicated comprehensively in any book discussing maritime archaeology. Wh...