Over fifteen years, the concept of virtual unwrapping as only a possibility has moved to a demonstrated process that has produced text from a scroll that cannot be physically opened (Seales et al. 2016). The intact Herculaneum scrolls, of which there are almost 300, however, continue to elude successful analysis and, as of this writing, remain unread. There are three primary challenges which, as the presentation will explain, will imminently be overcome, clearing the way for complete texts to be extracted from intact Herculaneum scrolls. When that happens, the argument for further excavation to find more scrolls will escalate considerably.
The first major challenge is geometric. The shape of every layer as ii appears internally in an intact, carbonized papyrus scroll had never been seen in situ prior to 2009 (Seales and Delattre, 2010). That groundbreaking volumetric scanning and subsequent analysis revealed the complexity of the interior form of each section of rolled, carbonized papyrus. The layers of papyrus are thin and brittle and often broken and frayed within the wraps. The shape of layers is unpredictable. It is a major challenge to find automated computer algorithms that can transform the damaged and unpredictable papyrus shape within a scan volume into a computer representation that is well-behaved and supports subsequent virtual unwrapping. Over the last five years, however, tremendous progress has been made to address this difficulty. Advances in tools based on computer vision and machine learning algorithms that help solve this geometric segmentation challenge makes converting the tangled inner Herculaneum geometry into regularized, virtually-flattened surfaces finally within reach.
The second major challenge is elemental. The first characterization of the trace elements in Herculaneum ink from fragment studies (Seales et al., 2008) showed the presence of calcium, strontium, and lead – but only in very small amounts. The minute quantities of elements found only in the ink suggested that a sensitive imaging method might be developed to leverage that fact. It also revealed that the quantities were so small that they would not easily be detected with conventional volumetric methods such as micro-CT. The prediction that phase contrast tomography would solve this ink-detection problem (Seales, Delattre 2013) was absolutely a step in the right direction, but the technology proved to be inadequate to the challenge. In fact, recent claims that ink was clearly identified in Herculaneum material through the use of phase contrast tomography are sketchy and deeply unsatisfying, primarily because the results did not reveal a complete text and are also impossible to confirm (Mocella et al., 2015). So while the problem of seeing contrast at ink that is carbon-based is generally acknowledged to be a major challenge, this presentation will reveal the major and systematic progress being made through new methods for revealing contrast in x-ray-based imagery of pure carbon ink. The results are remarkable and have impact for Herculaneum and beyond.
The third major challenge is diplomatic. Somehow the scholarly community, which includes conservators, has not been able to come together around the work of virtual unwrapping and provide technical teams who are positioned to meet these challenges by capturing data safely and reliably, consistent access to materials. Institutions have not made durable commitments, and researchers argue with each other rather than working together. This slows down progress and makes work that should take months or years to end up taking decades. This presentation will call for a unified, diplomatic approach in the service of scholarly progress. The call is premised on the strong likelihood that all Herculaneum material can and will be substantially revealed using the framework of virtual unwrapping; and as stewards of this process and its profoundly important impact, the scholarly community should embrace and encourage it.
Herculaneum: Works in Progress