Page 58 - Kennemerland VOC ship, 1664 - Published Reports

NAUTICAL ARCHAEOLOGY, 4.2
required in factor analysis demand the use of a computer, it would appear to be most satisfactory if all the data processing, from the recording stage onwards, were com- puterized. Such a system should be designed both to increase the accuracy of the recording and reduce the work demanded of the divers on the sea-bed. One possible programme was developed at Aston University in 1971 for application on the Kennemerland site, al- though it was never actually applied. It required every relevant point to be described by three simple measurements, from which the computer could subsequently produce the required plans, distribution maps, etc. (Forster & Higgs, 1973: 294). Since then, a package known as PLUTARCH has been developed at Keele which is specifically tailored to the requirements of archaeologists (Wilcock,
1974); it is hoped that this can be readily adapted for use on underwater sites. In both of these systems, the improvement is primarily one of accuracy, rather than economy of effort on site, an improvement which is principally due to the fact that they can deal effectively with recording in three dimensions. There should be little difficulty in arranging for the output of such a recording programme to be the input for a programme of analysis, although such analysis could be undertaken quite satisfactorily on the basis of manually
recorded material. It is intended that future work on the Kennemerland site will be conducted with a fully computerised system.
So far, attention has been focused entirely on sites where the remains have been scattered widely over the sea-bed, and no coherent ship’s structure has survived. Mention has also been made of the very different approach required when the ship’s structure is found to be intact. However, there are many sites which are intermediate between these two situations, and which therefore demand special techniques of investigation. Tn princi- pal, the approach should always be one of taking a sample, and analysing it in depth, although the methods of analysis will vary. One common situation is a site where the remains are totally buried under a flat,
featureless sandy bottom, as is the case with the Trinidad Valencera in County Donegal (Martin, 1976). Here, the limits of the site
have been defined by a metal detector survey. which showed a main deposit concentrated inanareaof100by200ft(30.46x60.96m). In the first excavation season in 1974 two areas within this site were opened, acting as ‘windows’onto the whole deposit; the problem was to see if the distributions uncovered could be used to indicate the pattern of the whole deposit. On the basis of the metal detector survey, it was suspected that the deposit centred on a line running diagonally across the site. One of the areas excavated lay near to this line, and proved to be very rich in finds; it was not completely excavated, and so is not susceptible to detailed analysis. The other area, measuring 40 x 50ft (12.19 x 15.24 m), lay in a corner of the site away from the principal diagonal. and was totally excavated; it yielded about 150 artefacts. The question is thus whether the distribution in this area shows a decreasing concentration of artefacts as one moves away from the diagonal; the technique for quanti- fying such a pattern is trend surface analysis, which identifies the principal trend underlying a pattern, and the distortions which interrupt this trend (Cole & King, 1968: 375-9). Elementary exercises in such pattern fitting have shown that the distributions in the excavated area show a very strong trend away from the principal diagonal, so that the
hypothesis, if not confirmed, IS at least not refuted. Of equal interest are the distortions revealed in this exercise, which suggest that very large objects (gun carriage wheels in this case) not only protect fragile objects which might otherwise have perished (e.g. leather- work) but also tend to attract other objects, probably by means of the concretion which they develop. This power of attraction is further demonstrated by the relative emptiness of the zones surrounding these concentrations.
Once the use of computers in analysis in underwater archaeology has been accepted, then there are many other problems which can be tackled more thoroughly. Another type of intermediate site is one on which there is a fair quantity of coherent structure, but in a very distorted condition, and lacking several areas of the ship entirely; a recently investi- gated example is the Dartmouth (Adnams,
1974). By the application of computerized
188