Someone had to go first. Someone needed the courage to press their hard won intellectual content into service and develop a new apparatus for publication in the twenty-first century. A Mid-Republican House from Gabii represents a crucial, if unsteady step into that future. Based on nearly a decade of fieldwork by the University of Michigan's Gabii Project, this resource — as it is clearly neither fully a book nor merely a website — marries a multi-layered history of Building B/Tincu House with an immersive 3D model linked to a research database.
The resource is divided into two main narrative sections: (1) an extended, open-access introduction to the project and (2) a three-level discussion of the history of Building B, including specialists’ reports, behind the pay-wall. The introduction, though long, gives a brief account of the site, the Gabii project's history, and the methods employed. Two additional sections, on ‘activities’ and the theory of 3D models in (at least this) publication, leave the reader intrigued, but insufficiently informed on concepts behind the design of this resource.
Once inside the ‘contents’ section, the narrative is divided into three overlapping sections intended to offer different levels of engagement with the site's history. The first level, Story, is a 1,500-word synopsis written in an accessible style for a non-academic audience that is hyperlinked to the More level, where a fuller discussion of the relevant interpretations can be found. Within the More level, key sections of text are linked either to activate the related content in the 3D model or to drill down further into the evidence within the third level, Details. The Details level is organised by phase into a series of stand-alone, data-driven descriptions that can be ‘dropped into’ from the More level. These are followed by the excellent specialists’ reports. Almost every stratigraphic unit in the Details level is linked to the 3D model, which is out-linked to the project's research database. Unlike the narrative, the database has no ‘on-ramp’ for the uninitiated, a problem the editors have attempted to tackle with full descriptions of what data each field is supposed to contain.
The centrepiece of this resource is the innovative 3D model that serves as an interactive illustration of the authors’ arguments, an interface to the raw research data and a sandbox for the curious, permitting an unstructured perusal of the site and its evidence. The model opens with a help-screen to explain its features and navigation. There are two viewing modes: (1) a limited 3D orbit with zoom and (2) a first-person ‘walk’ mode. An expandable menu at left loads 3D models, stratigraphic units and reconstructions from each phase. A one metre grid can also be toggled on and off. Trowel icons mark the location of key stratigraphic evidence and clicking opens an information pane with a short description, image, indicators of the artifacts and activities present, and a ‘Go to Ark’ button that links to the database record. For those who desire it, an accessibility mode removes the model. For others like me who find the immersive experience within the data so compelling, the absence of a full-screen mode is regrettable.
When open, the model pane occupies 75 per cent of the screen, leaving only seven lines of text visible below it (using default display settings). An ‘>’ icon at top left minimises the model. Many readers will make good use of this when trying to engage with the narrative. That narrative properly begins in the More section, where the editors lay out four phases of development, from the Iron Age (B-0) through the house's abandonment in the first century c.e. (B-3). Nearly all of the activity at Building B, however, occurs in the last three centuries b.c.e., including the foundation of the house c. 280–260 (B-1a), its southern extension c. 200 (B-1b) and its reinvention as an annex to the monumental construction across Road 4 a century later (B-2). A useful, but too brief, synthetic section on urbanism and domestic architecture comes near the end of the More level, followed by synopses of the artifactual reports.
Unsurprisingly, one finds the narrative based upon excellent archaeological practice (though I was alarmed by ‘hand collection’ and the need for metal-detectors in lieu of sieving), but not always excellent interpretation and argumentation. One example is the discussion of Phase B-3, which awkwardly conjoins the abandonment of Building B with a re-occupation of the area. This leads to confusing statements about a long, well-built wall reflecting the importance of traffic on a street that was buried and cut by this same wall. Why such construction was not dealt with as its own phase or sub-phase is a mystery. Another poor argument is the idea that a floor antedates the mint date of an ‘embedded’ coin. Even if a coin could become ‘embedded’ in a crushed tufo floor (and its findspot was not recorded), it could have done so any time after the mint date, allowing the floor to be built anytime thereafter as well. Unfortunately, there is also a significant interpretive error: perhaps the most important deposit (SUs 1306, 1385 and 1279) is phased as B-2 in the stratigraphic section, but phased in B-1b in the ceramic and faunal analyses. What is more, the depositional processes are contradictory in these sections; ‘accumulating over time’ in the former and ‘a single action’ in the latter. The impact of this error is either to undermine the connection between the closing of the house's exterior drains and the inversion of its Phase B-1 roofing system to slope inward in Phase B-2, or to place both those events further back in time, separating the origin of Phase B-2 from its other activities by as much as a century. The editors seem to average out the error, directly claiming the new entrance in B-2 was paved c. 150 b.c.e., but ignoring (or implicitly conceding to) both the post-quem dates for the external drain fill (decades earlier than 150 b.c.e.) and the post-quem dates for the building's interior renovations (decades later than 150 b.c.e.).
The impetus for this date perhaps derives from a strong desire to see the Tincu House converted into an ‘annex’ of the large public complex (Building F) across Road 4 in Phase B-2. The claim of the transformation is repeated but never supported by physical (connecting walls, shared drainage, etc.) or analogical (construction styles, assemblages, etc.) connections between the buildings. More than annoying the reader, the certainty of this connection has the consequence of foreclosing other interpretations. To this admittedly biased reviewer, it is clear that in Phase B-2 the building has become something akin to an inn and stable: a paved ramp, including clear cart ruts in the threshold stone, leads into a courtyard with a water basin against one wall (not a central impluvium) that is connected to a long room appropriate for animal housing. Of course, these observations are not incompatible with Building B still being an annex of Building F, but that interconnection must be made.
There are also a host of problems with the resource that appear to stem, in one way or another, from the structure of the publication apparatus and its over-reliance on the 3D model. The simplest is the absence of the necessary illustrations. Nowhere are Road 4 and room numbers ever labelled. Users should refer instead to Mogetta and Becker, AJA 118 (2014), figs 5, 11. Similarly, the absence of a scale bar (despite the 1 metre grid) and elevation information (even in the database) makes dimensional comparisons impossible. Most readers will find the table of phases and associated activities especially useful, but be astonished to find it only in the ceramicist's report. Additionally, it is only here that the notion of ‘Activities’ returns to the text. Likewise, stratigraphic arguments, even of crucial sequences, are left to be made implicitly by the 3D model and by Harris matrixes in the database: collectively, words like abut, bond, cut or over(lie)/under(lie) are used in fewer than three dozen instances.
The structure of the narrative produces its own curiosities. For example, why is the non-academic Story behind a $150 pay-wall, while the methods section is in the open access introduction? Likewise, why is other important methodological information embargoed in the ‘apologia’ section, not least the notice that the database and ceramics report are out of sync? Why is the database behind the pay-wall, but then given a CC-BY copyright status? Typos and infelicities are too many to mention, though one is important. Table 10 of the zooarchaeological report is missing and its caption mislabels table 11, causing table 11’s caption to accompany the first table of Phase B-3, which now goes uncaptioned and is intrusive in the Phase B-2 discussion.
Having piled on necessary criticism, let me offer an equal heap of praise. Nearly all of the foregoing critique — especially my attempt to reinterpret Phase B-2 — was only made possible by the vast quantities of data made available by this resource and by the careful articulation of those data between different data structures of text, 3D model and database. More importantly, it is clear that such openness derives from a place of remarkable humility, courage and integrity. I am fond of saying that an archaeological phase plan hides as much information as it reveals. Coming as a fait accompli, the static image of a phase plan presents hundreds of interpretive decisions — and unconscious (or worse) deceptions — with the finality of a solid black line. By contrast, editors have exposed to critique the entire interpretive process, from the individual observation and object, to its position within a series of nested arguments for Building B's development, to the discussion of how that building fits into the history of Gabii, and further still, Roman urbanism. This novel apparatus has another unique intellectual benefit. Having spent hours within the 3D model and text, I simply feel like I know Building B at Gabii better from this manner of publication than I know any other site I have not visited. The editors therefore are to be commended for making the kind of embodied knowledge that one creates by being on site for a long time — the habitus, the communion of simple presence — more accessible to those who may never be there.
In sum, the editors have made a brave choice to prioritise an openness to their entire interpretive process, but with the consequence of compromising the quality of those interpretations. What is a ‘win’ for novel modes of scholarly communication is a ‘loss’ for Gabii. Each reader must decide the worth of this bargain, but because their apparatus easily allowed me to suggest new interpretations, I side with the editors. I hope they (and others) will build on this model and improve it.
