What archaeologists do: The site report & what it means to excavate a hard drive

Colleen Morgan and I are wrapping up the first chapter of MAD-P (Media Archaeology Drive Project), an experiment in extending archaeological method into the systematised analysis of media objects. This project began as a provocation — intended to prompt reflection (both within and beyond the discipline) on the place of archaeology in the wider media and cultural studies landscape. That provocation has exposed, we think, an obvious gap between what we do as anthropologists and what we could do, and the space that archaeology might occupy in variously exploring the past, exposing the present and anticipating or shaping the future. Our modest excavation of an abandoned hard drive hints at what happens when the taken-for-granted aspects of media products are subject to step-by-step archaeological recording. Such an investigative process draws your eye immediately to both the material and the discursive, to the layered nature of each, and to the impossibly entangled and slippery interconnections amongst them. The individual material constituents of the artifact, their assemblage, the labour behind their composition, and their various manifestations in both computer code and in complex virtual spaces are made obvious. Indeed, as discussed below, the entire concept of an artifact is destablised in such work. From our perspective, the productivity of such a project should not be underestimated in terms of its potential both to critique the past and to speculate about possible futures.

To facilitate MAD-P as a whole, Colleen prepared context sheets, using as a model those employed at the Neolithic site of Çatalhöyük. We recorded by hand and photographed or screenshot all elements of our process. We also kept an associated set of notes — perhaps the equivalent of a field diary, but logged electronically and as a combined output, weaving together observations that we’d made in dialogue with one another. Following the excavation, Colleen set about writing our archive report, a structured review of our field site, findings and interpretations, which we present here.

What is an archive report?

Following archaeological excavations, it is generally expected that the primary investigation team writes a preliminary archive report. This report details the stratigraphy of the excavation in stratigraphic order, from the earliest activity until the latest–what could be considered a “bottom-up” approach. Along with stratigraphic details, we discuss any notable finds, and provide an initial phasing for the site. By phasing, we mean that we group archaeologically-ascertained events together, such as major building events, fires, or architectural changes. When sites contain a multitude of archaeologically-identifiable events, or “contexts”, this endeavor requires a mastery of stratigraphic understanding. To hone such mastery, we typically draw upon a Harris Matrix, a graphic representation of stratigraphic relationships. In the case of the Media Archaeology Drive Project, we’ve limited the excavations of our found hard drive to an extent that the stratigraphy is simple and major phasing is unnecessary. Still, in the spirit of archaeological media archaeology, we present our excavation as a site report.

These reports are usually articulated in coded language, primarily only comprehensible to experts and written in the passive tense. There is much to be critiqued about both the style and the legacy of such reporting, and we note with some despair the lack of progress over the years in rethinking its dimensions (although a not insignificant number of scholars have commented on, and indeed, creatively experimented with it; e.g., see Beranek 2008, Mickel 2012, Praetzellis et al. 1998 in the journal Historical Archaeology) . Where site reports prove useful, we’d argue, is in instilling care for process and interpretation. They can work as a meaningful pedagogical strategy, aiding in thinking through the fit between disparate data gathered during archaeological investigation. They can be used to reflexively review intellectual processes during excavation and to reevaluate interpretations after the fact. They can provide a record for future researchers to understand what has been systematically destroyed through excavation.

MAD-P Context Sheet
MAD-P Context Sheet [unless otherwise noted, all photos in this blog post are by the authors, Colleen Morgan & Sara Perry]
For these reasons, we have invested in our own MAD-P report, reproduced below. For reference, the numbers in parentheses or square-brackets identify each “context”, which we recognise as archaeological events. Parentheses identify positive events and square-brackets identify negative events, or “cuts.” Each of these contexts has been fully described using a standard context sheet, drawn by hand, and photographed. Cardinal directions were somewhat arbitrary, but for descriptive purposes, north is always at the top of the photograph. Measurements were generally rounded to the closest millimeter.

Sara measuring for MAD-P
Sara measuring for MAD-P

MAD-P Archive Report

The Media Archaeology Drive Project centers on the excavation of the contents and material of a Samsung hard drive, produced in Korea in 2004, and subsequently bought by the Department of Archaeology at the University of York. The drive is 100mm x 150mm and 30mm in depth. It is silver, and has technical specifications and other identifying text on a white label on its exterior. As it is a portable device, other contextual information is variable, but the bulk of the excavation work was performed by Colleen Morgan and myself in my office in the Department of Archaeology at the University of York (UK). As we excavated in the confines of the office, the weather was not a factor (as is typically the case)–but it was sunny and lovely outside. As Colleen notes, she found herself wishing that she could be in the open with a shovel, instead of in an office with a dozen tiny screwdrivers. I thought the opposite.


Context 18 - Image 1617
Context 18 – Image 1617

At the heart of the hard drive was a circuit board (18), green (5G 4/6 on the Munsell color chart) measuring 100mm x 90mm x 5mm. It was bumpy to touch, and contained metal inclusions printed on the board. On the east side of the board was a black interface. The circuit board (18) was covered by a small piece of black (10YR 2/1) cloth (17), placed between the circuit board and the plastic body of the hard drive (16). This cloth conformed to the outline of the circuit board on the north, west, and south sides (90mm x 70mm x 1mm), and was spongy and smooth.

The black (10YR 2/1) plastic body of the hard drive (100mm x 150mm x 20mm) lay on top of the black cloth (17) and provided one of the main structural elements of the hard drive–many of the elements of the drive were attached to this body. There was a circular cut [15] in the plastic body, through which a silver metal spindle (13) interacted with the circuit board (18).

Context 16, 14, 13, 12, 11 in Image 1606
Context 16, 14, 13, 12, 11 in Image 1606

Also set into the black plastic body (16) was an actuator in the north-east corner (12), with a ribbon that allowed interaction with the reader arm in the south-east corner (11). The reader arm was sandwiched between two metal plates, at depth (14) mounted the reader arm onto the black plastic body, and (10) which kept the reader arm in place. The plates were relatively uniform, both measuring 50mm x 50mm x 2mm. A silver platter (9), 100mm x 100mm x 1mm was mounted onto the spindle to the western extent of the hard drive and kept in place by a washer and four screws. All of these contexts were secured by the silver metal case, (8).


The second phase explored by our MAD-P team was the user interface installed on the hard drive. This was a mistake, as there was an intervening phase that was not investigated, but this will be addressed in the discussion below.

The excavation of the user interface allows for some reconsideration of understanding the hard drive as a stratigraphic sequence. Our method, “drilling down” through the folder structure of the user interface mimics our expectations of archaeological excavation, by moving down or deeper into the folder structure. Yet at “depth” the icon that represents the goal of our excavation might be temporally older or younger than the folder that contains it; the very presence of the icon has changed the temporal signature of the folder. Further complicating this excavation was the concept of depth as applied to a user interface. To record depth, MAD-P decided to use the “doubleclick” (DC) as a unit of measurement (see below for critique of this decision). For the purposes of this investigation, we’ll sidestep this spatial and temporal snarl and treat the folder structure as a stratigraphic sequence.

Screenshot of David Byrne music file
Screenshot of David Byrne music file

At depth, a music file (7), appeared as a 150mm x 150mm window, black with a yellow line, blue button, and a small rainbow with a face. This window provided a visual representation of an auditory event, a song “Like Humans Do” from the David Byrne album “Look into the Eyeball.” Further discussion of this context (7) is appended to the stratigraphic report (below). This file was contained within a blue and white icon (6), a 15mm x 5mm x 1DC, with an eighth note (quaver) and two frames of movie reel, and the label “music.asx”. This icon was one of several contained in the My Music folder (5), which were labelled “sample.music,” “music.bmp,” and “music.wma.” These were not given context numbers in the interest of simplicity, but a further investigation would have included them within our sequence. The “My Music” folder (5) 20mm x 5mm x 1DC had a time stamp of 02/08/2010, the latest in our sequence, which could provide our terminus post quem (TPQ), date after which the hard drive had stopped being used. The My Music folder (5) was yellow, black and white, with a beamed eighth note, in contrast to the single eighth note apparent in the music.asx icon (6).

Screen shot of Shared Documents > My Music folder
Screen shot of Shared Documents > My Music folder

The folder structure that contained the My Music folder (5) was relatively simple, with Shared Documents, (4) a yellow 30mm x 5mm x 1DC icon contained within “All Users” (3), contained within “Documents and Settings” (2). Also contained within “Documents and Settings” were the folders “Heather” and “Michael.” These folders were password protected and we did not investigate them, as outlined in our research design ethics statement. Casual enquiry in the department did not reveal the identity of these individuals and they remain unidentified. The top level icon, Local Disk (E:) (1) was gray, 25mm x 5mm x 0DC, and appeared on the desktop of my computer.

MAD-P Harris Matrix
MAD-P Harris Matrix


The formalized strategy employed during our MAD-P excavation led to several unexpected problems and insights that may be productive for future research. Perhaps the biggest omission in the project is represented by the break in the project’s Harris Matrix. There is nothing connecting Phase I and Phase II of the excavations because we did not excavate the code that connects the hard drive with the user interface. This would have added considerable depth and complexity to our analysis, and is a priority for future investigations.

Secondly, the anchors of archaeological investigation–temporality and spatial distribution–were slippery and indistinct during the excavation. The decision to measure depth in double-clicks added some coherence to the idea of folder stratigraphy, but it is untested as a relative measure for evaluating the overall folder hierarchy and would require more investigation. As noted in discussion of our process (e.g., with @adreinhard), the legibility and longevity of the DC as a unit of measurement is debatable.

Thirdly, again with reference to measurement, and as noted by designer @iankirkpatrick, in the future icons should be measured in pixel width, rather than in actual mm.  As each screen will obviously have bigger or smaller pixels, the icon will be different in size depending on the screen & its resolution.  The only really consistent measurement of icons, then, is pixel width & height.

Forthly, formal context sheets provided an important continuity in the investigation, but would have to be modified for future research. Even so, some of the formal prompts, such as “texture” and “inclusions” and “execution” provided a welcome decentering in our excavation of the hard drive. What is the texture of a file folder?

1620 - Image of MAD-P artifacts in bags
1620 – Image of MAD-P artifacts in bags

Finally, MAD-P revealed a certain ambivalence in archaeological definitions of artifacts, contexts, sites, and sequences. During the investigation of both of the phases–the hard drive and the user interface–we moved back and forth between our understanding of how to evaluate an artifact and how to record an archaeological site. The destabilization of these definitions was an unexpected resistance to archaeological investigation from these media, and this resonates through our subsequent archaeological practice. This ambivalence can also count as one of the benefits of the investigation.

Additional benefits reinforce the utility of the archaeological method. The best example is the continued usefulness of drawing in archaeological recording. During MAD-P, we sketched each context on the back of our context sheets, and created a formal scaled drawing on permatrace. Interestingly, the sketches were very useful during the investigation of the user interface phase, while the drawings on permatrace were more useful during the hard drive phase. Sketching user interface icons was jarring, and felt silly, but became immediately compelling. Drawing the object of your research encourages a depth of involvement, forcing your attention on its complete visualization and how it interacts with the surrounding context. The formalized drawings of the hard drive on permatrace, a semi-transparent tracing paper, allowed us to overlay the drawings to understand the stratigraphy of the hard drive and the relationships of the components to each other.

Another affordance of the archaeological investigation was the formalized separation of the constitutive components of the hard drive into finds bags. This provided an interesting contrast to the relative ephemerality of the “finds” of the user interface investigation, various folders and music files, though they were contained on the platter of the hard drive. These user interface artifacts, though not as apparently present and sorted into bags, are actually more omnipresent–the best “find” during the investigation was the David Byrne song hidden under a generic label in an unremarkable folder structure. The song, Like Humans Do, was included in Windows XP to demonstrate the Windows Media Player, leading us to wonder–was it the most ubiquitous song in the world? Perhaps now eclipsed by the U2 album, Songs of Innocence recently embedded in iTunes? [As it happens, following Colleen’s presentation at Bradford’s media archaeology conference, @pbenzon has pointed out that the Nokia Tune may, in fact, be the frontrunner–a subject that’s been explored by Jeff Thompson.]

Most importantly, MAD-P was conceived as a critical, creative exploration of the intersections between media archaeology and archaeology, but it was also an incredibly fun project. Applying archaeological methods to a computer screen was the best kind of mischief–it encouraged critical play to reconfigure our approach to research. This mode of critical play is being more fully investigated in our Heritage & Play working group at the University of York, and is part of a larger series of questions that we are exploring around the relationship between doing, making, knowing, learning and the crafting of expertise. We would have liked to engage with some of these questions in more depth here on Savage Minds, but this month has flown by for us, with MAD-P rolling out alongside various other related projects, including Colleen’s recent Archaeology and Minecraft event at York. We will continue our work, then, on our own web profiles, so please stay tuned via @clmorgan and @ArchaeologistSP.


Sara Perry is Director of Studies of Digital Heritage and Lecturer in Cultural Heritage Management in the Department of Archaeology at the University of York (York, UK). Sara blogs about her academic life at The Archaeological Eye (saraperry.wordpress.com) and tweets at @archaeologistsp.

4 thoughts on “What archaeologists do: The site report & what it means to excavate a hard drive

  1. Brilliant. I wonder what would have been the result if a media archeologist had been included in the team? I imagine her poring over the music on the hard drive, concerned about what it says about the collector’s taste and the cultural background or institutional context that shaped it.

  2. Thanks for this work: I have an archaeology background and I teach digital archiving in the School of Information at the University of Texas-Austin. I’ll share this with 1) a PhD student in archivy who is actually working on archaeological recordkeeping (which she is treating as including artifact-keeping) and 2) my spring class in digital archiving, as an interesting and unexpected angle on parts of what we do. BTW I was living in the same dig house with Ed Harris in Winchester when he was putting the final touches on the Matrix and you might want to contact him because I think he would be interested in how far it has gone, here, as a diagnostic: he is now Executive Director of the Bermuda Maritime Museum.

  3. Fascinating perspective! I am an archaeology PhD student at the University of Western Ontario who researches archaeological representations, particularly digital representations, and user engagement. BUT, I am also a co-founder of a non-profit, community driven program called the MakerBus (www.makerbus.ca) hosted in London, ON. As part of the maker-movement ethos, we literally have a bus that transports technology and other maker tools like 3D printers, 3D scanners, screwdrivers, saws, and whatnot to various parts of the city and engage people in various projects related to deconstruction, reconstruction, and making/innovation as a methodology in order to understand various subjects. We use a cell phone hack that involves combining dollar-store items to make a high powered macro-lens subsequently used to explore archaeological artefacts (as an example). We have had a program in the past where we had young students (most around age 12) take apart a DVD player to retrieve the lens from inside as an extension of the macro-lens hack. I love your approach to critical play and look forward to see what comes out of further explorations. See you both on Twitter!

  4. This August I visited the (currently) world biggest e-waste dump in Accra, Ghana
    this August and bought around 24 harddrives.

    See some photos and a workshop proposal here:

    I am now in the process to organize workshops to excavate them and would
    like to document the outcome in an artistic media installation. Your
    Archive report might come in handy for that.

Comments are closed.