Lost in translation ? A comment on the excavation report

A substantial proportion of archaeology's primary data collection takes place through excavation and surface survey. Writing-up and presenting these field observations and experiences are translations (Shanks 1992: 79). The fieldwork database becomes the raw material of hard-copy texts, notably text books, synthetic books and articles, and site reports. Of these, the site report in particular has reached a floruit of insipid, formulaic stagnation. This brief comment raises the issue of whether it is necessary, or inevitable, that most site reports be presented in such a boring, uninspired way. It could, for instance, be proffered that such reports are not meant to be interesting as an integral whole but are primarily collations of archives. In this scenario it is what can be extracted from reports, with particular questions and specialist interests in mind, that is interesting (e.g. Morris' 1994 synthesis of first millennium BC ceramic production and exchange using information gathered from numerous site pottery reports).


Introduction
"Everyone who has dug up anything knows the excitement ... Why is it, then that publication of that pattern in a site report is a more wearisome business ?"(Tilley 1989: 279) "A lot of people are fascinated by excavating the past ... But it appears in report form; the experience is lost and hardly evoked."(Shanks 1992: 183).
"There seems to be presented a choice: write poems, novels, paint watercolours -subjective fictions; or do archaeology -concerned with the past itself."(Shanks 1992: 12).
A substantial proportion of archaeology's primary data collection takes place through excavation and surface survey.Writing-up and presenting these field observations and experiences are translations (Shanks 1992: 79).The fieldwork database becomes the raw material of hard-copy texts, notably text books, synthetic books and articles, and site reports.Of these, the site report in particular has reached a floruit of insipid, formulaic stagnation.This brief comment raises the issue of whether it is necessary, or inevitable, that most site reports be presented in such a boring, uninspired way.It could, for instance, be proffered that such reports are not meant to be interesting as an integral whole but are primarily collations of archives.In this scenario it is what can be extracted from reports, with particular questions and specialist interests in mind, that is interesting (e.g.Morris' 1994 synthesis of first millennium BC ceramic production and exchange using information gathered from numerous site pottery reports).
What follows touches upon the history and percepts behind the current situation.At the same time I wish to explore some of the possibilites for change, particularly in the realm of 'everyday' field archaeology and its publication.The discussion has a specivity, given that my examples are based on British, and particularly prehistoric fieldwork.

Turning from passive to active
"Cutting XIII to the north-west of the gateway brought to light several interesting features."(Wilson 1939: 206).
We are at a transition point in our consideration of how archaeological reports and texts Papers from the Institute of Archaeology 10 (1999): 1-8 should be written.Fieldwork reports have remained in much the same format for at least 40 years (Hamilton 1996).The spirit of antiquarian archaeological tracts is currently being evoked as providing an experience of site, place and the interpretative process that is overtly absent from our present tradition of the depersonalized site report.Hodder (1996: 263-273) has noted how 18th and early 19th century excavation accounts give a story of investigation which is fixed in time and place and which is documented as a sequence of events of discovery which follow the progress of the excavation.As a result, these accounts have an engaging tension as they follow the process of discovery, and record the ideas and field tactics which were rejected or adopted as work proceeded.In these antiquarian texts the 'I' of the author is omnipresent (e.g.Pitt-Rivers 1894), and the reader knows that the site could have potentially yielded a different story if dug differently.
The site reports of today are the outcome of a process which began towards the end of the 19th century.An increasingly passivity of writing style developed.The effect of this passivity was to imply that the data spoke for itself, rather than being the outcome of the individual field actions and the perceptions of the investigators.A world was constructed where inanimate tools and trenches rather than people seemingly produced the results.Data was increasingly categorized into units of study which formed the structural framework of the report.Information became ordered by features and artefact categories, rather than by the sequence of excavation.By the 1960s the excavation report was wholly in its present form.Its format was entrenched by the authority of standard texts which laid out the principles of publication (Alexander 1970;Grinsell 1974).The style of the site report went hand in hand with the ethos of processual archaeology which in part has been characterized as involving rationality, not emotion, and a detached expression of procedures rather than an involved, personalized experience of the chosen methods of analysis.
From the 1980s there has been a quirky and continuing trend to soften this depersonalized sterility by providing thought-provoking asides via the addition of literary quotes fronting the text, or at the beginning of each section or chapter.This fashion crops up in text books (Bradley 1984), thematic books (Bradley 1990), articles (Kinnes 1981), and some reports (Barrett, Bradley and Green 1991).It is individualistic in that it characterizes the work of specific authors.Similarly, by the mid 1970s, the titles of some fieldwork and excavation reports mark incipient change in the idea of what a site report should be all about.In these, the title advertises that the work is being seen as part of, or contributing to, a wider whole.This greater entity might be, for instance, a landscape region, a major site category, or a more generalized socio-economic interpretative framework (e.g.Cunliffe 1976;Drewett 1982;Wilkinson 1988).
On the surface, the formal remit within which English Heritage-supported, and developer-supported project work currently functions has created more researchorientated fieldwork.Project designs, and publication guidelines require the scale of analysis to be commensurate with the assessed significance of the data (Cunliffe 1983;Department of Environment 1975;English Heritage 1991).In spite of this, most reports that get to hard copy publication essentially look and read the same due to their standardization of layout, and graphics.

Textual regeneration
Alongside the above observations, it is important to note some of the several British publications of the last decade which have more conspicuously published fieldwork in thematic, and unorthodox formats.Examples include: i) Fleming's (1988) popular synthesis of his fieldwork on the Bronze Age Dartmoor Reaves (particularly Chapter 6 on excavation).The latter gives a sequential history of discovering and changing approaches to excavating the Venford reave and Holne Moor houses; ii) Pitts' and Roberts' Fairweather Eden (1998) which traces the discoveries from ten years of excavation at the Boxgrove Palaeolithic site, and is "part textbook, part racy narrative" (Times review published on the book jacket); iii) the explicit use of a daily excavation log or diary to generate an excavation narrative in the 'excavation biography' of the very first ten days of the Bodmin Moor Leskernick excavations of a Bronze Age settlement and ritual complex (Bender, Hamilton and Tilley 1997); and iv) the publication of fieldwork on the prehistory of Cranbourne Chase in two complementary, but contrastingly different, volumes.The glossier of the two volumes (Barrett, Bradley and Green 1991) meshes the reports of the excavated sites into wider themes relating to the placement of sites and monuments in the Cranbourne Chase landscape through time.The specialists reports on bone, pottery, environmental data and the like are decanted into a separate, companion volume (Barrett, Bradley and Hall 1991).This specialist volume is innovative in its own right.Individual categories of finds from the various excavated sites are each considered collectively (rather than by site) against the trajectory of time change (e.g.Cleal 1991).Each specialist report is fronted by a statement of the aims and issues to be addressed, re-emphasizing their thematic quality and wider applicability.

Debate, dialogue, and diaries
"The final report of an excavation is only a story of what people did." (Shanks 1992: 130).
Today's site reports are the outcome of the work of numerous contributors and specialists, and yet there is rarely any debate or uncertainty in the text.Debate needs to be a more upfront, and standard part, of the write-up.Possibly the publication of transcriptions of actual dialogues would be a suitable medium for textually reproducing these discourses.Bender (1999), for example, has used dialogues as an effective means of presenting the complex and diverse knowledges and interests which focus on the site of Stonehenge.
The publication of personal, project diaries and trench diaries has been advocated as a medium for expressing sequential and personalized thoughts about site and place, and the changing human factors which affect the process of field interpretation and decision making.These diaries are curiously more rivetting to read than the empirical descriptions of the site data (Bender, Hamilton and Tilley 1997).Their role is to give insights into the processes of knowledge recovery and interpretation in the field and beyond.They provide ethnographies of excavations.Such diaries create a huge database in their own right.They can be selectively manipulated into hard copy site reports (Bender, Hamilton and Tilley 1997) but the only way to give an unfiltered account is to publish them in full via Web sites (Hodder 1999: 121; the Çatalhöyük web site on http://catal.arch.cam.ac.uk/catal/catal.html; the Leskernick web site on http://www.ucl.ac.uk/leskernick; the Fishbourne Roman villa diaries on http://www.sussexpast.co.uk).
The formal keeping and archiving of personal diaries belongs to a rarefied world which is most workable within the timetable and closed communities of research excavations.It is not easily viable in the tight schedules of developer-generated contract excavations, and the often home-based contexts of individuals working on them.Perhaps here, the primary recording stage is the means of documenting some of the fluidity and realities of recognising, extracting and interpreting field data.Context sheets should have sections which encourage debate and comment on the evolving strategies and changing interpretations which form part of any fieldwork.The Leskernick excavations, for example have context sheets with prompt boxes for giving the interpretation of features/contexts both before and after their excavation.They also request the percentage of confidence in each interpretation to be stated.These can be seen as devices to produce documentation which allows the final site report to be more open ended and more reflective of the real process of data extraction and reconstitution.

The graphics
"He handed me the trowel and said 'what do you think'...The reasoning process was a physical one ... The micromorphologist on site however spent much of her life 'looking'.She could not feel the layers in her glass slides."(Hodder 1999: 10-11).
At one end of the scale of graphic representation, site installation 'art' and different forms and types of image representation are being used to explore and represent the construction, and understanding, of meaning in archaeological sites (Bender, Hamilton and Tilley in press; see also the Leskernick web site).At the other end of the spectrum, the imagery of site reports remains wholly uninspiring and neutral.The conventions for plans, sections and artefacts in site reports are highly codified (Adkins and Adkins 1989;Griffiths, Jenner and Wilson 1990) and produce bland, unarresting images (Hamilton 1996(Hamilton , 1997)).
Graphic imagery is a significant vehicle for giving a sense of the look, touch and atmosphere of field data.In site reports, graphic imagery is under-exploited as means of enabling thought.Complex ideas can be brought together by the juxtaposition of images, for instance by using collage and montage (Shanks 1991;Bender, Hamilton and Tilley in press).The tactile (e.g.texture representation), visual (e.g.light and dark colouring, and shape), and experiential (e.g.atmospheric conditions) nature of site locales, features and finds all fall within the realm of graphic imagery.There is a case for putting the individual back into graphics.There is a certain pleasure in recognising the drawer and/ or organizer of a graphic layout (the equivalent of the 'I' in the text).In using bland, anonymous graphics the particular premises and interests that the graphics are inevitably in part an outcome and expression of are lost, or subsumed.Examples of the single person pronoun in graphics include Gurd's remarkable 'textural' drawings of the prehistoric pottery produced particularly in Sussex Archaeological Collections from 1915-1937(Goddard 1997), and Ashbee's (1970) distinctive drawings of barrow plans and sections.
Graphics can provide tableaux for bringing together diverse empirical, and contextual information.The current mutual exclusivity of text and graphic representation in site reports is debilitating.The animation and annotation of graphics should be more fully explored by the mixing of text, different media of graphic representation, and different scales of reproduction together on the same page.The ease with which various types of graphics and text can be mixed via CAD means that it is possible to focus data presentation around graphics at the site report level.For instance a page of pottery drawings could include: blow ups of technological and decorative details (line drawn or photographs); lettering giving colour codes, fabric types and context attributions; graphs giving details of mineral inclusions; and notes, for example, about difficult to interpret pottery forms (Hamilton 1996(Hamilton , 1997;;Macpherson Grant 1991).This self-evidently would provide richer and more thought-provoking information than either pure text (artefact descriptions) or a series of simple line drawings.Graphics can also be importantly used to recontextualize finds.The tradition of publishing all the pottery, metalwork, and so on by type in separate blocks of text and illustration (i.e.specialist reports) separates finds that originally cohabited, making it difficult to make interpretative linkages (see the Westhampnett Iron Age cemetery site report, which successfully publishes features together with their recovered finds; Fitzpatrick 1997).

Conclusion
Until excavation reports are no longer wholly familiar and standardized, they never will make exciting, or indeed thought-provoking reading.Equally, the full expression of a lyrical reflexivity involving dialogue and doubt is a luxury for the research projects of academia.Any attempt to successfully change or enrich the essential nature of the site report needs to take strategic account of a market dominated, and structured, by developer archaeology.One way forward may be to use graphics as a more central medium for expressing and juxtaposing inter-connected and multivocal information.Another important remit is to generate formal space for dialogue at the level of context recording, which subsequently can be reflected in elements of the final report.
The publication of field data will be more dynamic and more inspiring if our use of site reports to produce syntheses explicitly show and emphasize the links between these two tiers of knowledge production.The format of the site report has been characterized as analogous to describing in detail the ingredients of a cake without providing a recipe for baking the cake (Tilley 1989).Part of the issue is therefore who should be thinking up recipes and baking the cakes.Pragmatically, we have to accept that some site reports will never become cakes, and a few will become gateaux.A more fundamental and immediate requirement is that 'everyday' site reports better facilitate and inspire the invention of recipes.