Abstract. Ancient undeciphered scripts present problems of different nature, not just tied to linguistic identification. The undeciphered Cypro-Minoan script from second millennium BCE Cyprus, for instance, currently does not have a standardized, definitive inventory of signs, and, in addition, stands divided into three separate subgroups (CM1, CM2, CM3), which have also been alleged to record different languages. However, this state of the art is not consensually accepted by the experts. In this article, we aim to apply a method that can aid to shed light on the tripartite division, to assess if it holds up against a multi-pronged, multi-disciplinary approach. This involves considerations linked to paleography (shapes of individual signs) and epigraphy (writing style tied to the support used), and crucially, deep learning-based strategies. These automatic methods, which are widely adopted in many fields such as computer vision and computational linguistics, allow us to look from an innovative perspective at the specific issues presented by ancient, poorly understood scripts in general, and Cypro-Minoan in particular. The usage of a state-of-the-art convolutional neural model that is unsupervised, and therefore does not use any prior knowledge of the script, is still underrepresented in the study of undeciphered writing systems, and helps to investigate the tripartite division from a fresh standpoint. The conclusions we reached show that: 1. the use of different media skews to a large extent the uniformity of the sign shapes; 2. the application of several neural techniques confirm this, since they highlight graphic proximity among signs inscribed on similar supports; 3. multi-stranded approaches prove to be a successful tool to investigate ancient scripts whose language is still unidentified. More crucially, these aspects, together, point in the same direction, namely the validation of a unitary, single Cypro-Minoan script, rather than the current division into three subgroups.
Website of publisher: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0269544
Roberta Ravanelli, Lorenzo Lastilla, Silvia Ferrara
Abstract. Any attempt of decipherment and language identification of the scripts from the Aegean dating to the second millennium BCE (namely Cretan Hieroglyphic, Linear A, and Cypro-Minoan) has relied, until today, on traditional catalogues of inscriptions, consisting of incomplete or subjective 2D representations, such as photographs and hand-drawn copies, which are not suitable for documenting such three-dimensional writing systems. In contrast, 3D models of the inscribed media allow for an accurate and objective “autopsy” of the entire surface of the inscriptions. In this context, this work presents an efficient, accurate, high-resolution, and high-quality texture photogrammetric workflow based on focus-stacked macro images, designed for the 3D modeling of small Aegean inscriptions, to properly reconstruct their geometry and to enhance the identification of their signs, making their transcription as unbiased as possible. The pipeline we propose also benefits from a pre-processing stage to remove any coloration difference from the images, and a reliable and simple 3D scaling procedure. We tested this workflow on six inscribed artifacts (two in Cretan Hieroglyphic, three in Linear A, one of uncertain affiliation), whose average size ranges approximately from 1 to 3 cm. Our results show that this workflow achieved an accuracy of a few hundredths of mm, comparable to the technical specifications of standard commercial 3D scanners. Moreover, the high 3D density we obtained (corresponding to the edge average length of the 3D model mesh), up to ≈ 30 µm, allowed us to reconstruct even the smallest details of the inscriptions, both in the mesh and in the texture layer of the 3D models.
Website of publisher: https://authors.elsevier.com/a/1eYUz4ckJEyOlX
Abstract. The focus of this paper is to present a new methodology that examines Cretan Hieroglyphic seals from both epigraphic and glyptic standpoints to be understood as parts of an integrated and multimodal system of communication. As our premise, we consider the newly published material from Myrtos-Pyrgos (Ferrara, Weingarten, Cadogan, 2016), and then compare and contrast local trends impacting the presence and use of inscribed seals from reasonably well provenanced Middle Minoan contexts in the East of the island. The goals are : 1) to throw light on the cultural significance of the administrative and symbolic functions played by Hieroglyphic seals and seal impressions ; and 2) to gain a synoptic appreciation of the emergence and use of this relatively short-lived writing system.
Website of publisher: http://www.libraweb.net/articoli.php?chiave=202233301&rivista=333
Abstract. This work has two specific objectives : 1. The reconstruction of a palaeographic model that describes the relationship between the Cretan Hieroglyphic script and Linear A ; 2. The structural analysis of the sign distribution of the two scripts and the distribution of their sign sequences (or sign groups). The two sets of analysis aim to establish via rigorous, solid parameters the degree of relation between the two scripts, by reconstructing through an evolutionary dimension the development of each sign and by establishing its positional distribution in each script corpus. The conclusions indicate that, despite sharing a nucleus of signs, the two scripts show notable divergences : a percentage of Linear A signs is not shared by the Cretan Hierolgyphic, which implies innovation strategies from a common template. In addition, the analysis of the two writing systems shows that no sign sequences are shared, in addition to different positional distributions. This evidence indicates that Linear A was the result of substantial adaptation strategies of the Cretan Hieroglyphic model.
Available in open access here: http://www.libraweb.net/articoli.php?chiave=202233301&rivista=333
Abstract. Ever since the publication of the first and unsurpassed corpus of Cretan Hieroglyphic inscriptions by Louis Godart and Jean-Pierre Olivier, Corpus Hiero-glyphicarum Inscriptionum Cretae, known as CHIC, there has been no systematic or comprehensive reassessment of several difficult readings of signs and sign groups geared towards a rationalization of the sign-list. In this article, we discuss several readings in depth, by engaging with issues that relate to script classifica-tion, sign frequencies, interpretative uncertainties, and new finds post CHIC. Spe-cifically, we highlight inconsistencies in the graphic behavior of signs, especially those of singular attestations (‘hapax’) or low frequency, and we also reassign uncertain cases that may be more likely read as Linear A rather than Cretan Hiero-glyphic. As a steppingstone in rationalizing the sign-list, we apply a systematic approach to the material by means of a thorough sign-by-sign paleographic and structural reassessment, also presenting the resulting inventory.
Abstract. What is the origin of the earliest script in Europe? Is it invented locally or borrowed from an external template? How can we go about addressing this problem? A common view is that the script in question, Cretan Hieroglyphic, is locally created but externally inspired, probably through an influence from Egypt. But this appreciation should be the result of a full examination of the evidence, rather than a superficial appraisal of the script signs. This article reframes this approach, starting with generic assessments on origin and stimulus, and so opens a new avenue that takes into account the following aspects: 1) the establishment of a methodology for cross-comparisons between the Egyptian and Cretan scripts; 2) the situated context of the Egyptian and Cretan scripts in the mid-3rd to late 3rd millennium B.C.E.; 3) the local Cretan seal imagery; and 4) case studies of sign shapes, representing physical and immaterial referents across the Egyptian Hieroglyphic and Cretan repertoires. Only from this broad, multicentric framework, which has input from archaeology, epigraphy, iconography, and paleography, can we establish a solid method to address the origin of Cretan Hieroglyphic.
Article we be available soon at our website (open access embargoed till November 2022).
Website of publisher: https://www.journals.uchicago.edu/doi/full/10.1086/716098
Abstract. The Rongorongo is a system of writing, still undeciphered, from Easter Island in the Pacific. It consists of a corpus of twenty-six inscriptions, scattered around the world. This article presents the state-of-the art in the study of one of these inscriptions, Text D or the ‘Échancrée’ tablet housed in a museum in Rome, Italy. Through an integrated methodology based on photogrammetry and high-precision structured light scanning, a 3D model of the inscriptions is made available through a public 3D Viewer for the first time. The technique made use of the benefits of both methods of image acquisition: a very accurate, precise, high resolution, and metric reconstruction of the tablet geometry gained through the scanning process, and a high-quality texture achieved through photogrammetry. In addition, we present a new analysis of the text, through a close palaeographic examination of its signs, and corrections of previous hand drawings and transcriptions. The ultimate aim is to reach unbiased ‘readings’ of the signs through an integrated synergy of traditional palaeographic analysis and an advanced 3D model. These, applied to all the inscriptions, constitute the necessary stepping-stones for any decipherment attempt.
Available for download at: https://academic.oup.com/dsh/advance-article/doi/10.1093/llc/fqab045/6387816?guestAccessKey=c7841dea-2e5e-4e48-b978-01c0e27272d6
Abstract. This article examines the earliest attestations of writing on Crete at the beginning of the second millennium BCE, the so-called ‘Archanes formula’. The aim is to reassess its origin, purpose, significance and ‘reading’ through a multi-step analysis taking in details of palaeography, correlations with iconographic seal motifs, and material culture. Key issues are considered, namely the extent to which is it comparable with the Linear A ‘libation formula’ a-sa-sa-ra-me, or, conversely, whether it should be singled out as a separate writing tradition. To address these questions, the ‘Archanes formula’ is brought under close scrutiny, vis-à-vis the graphic repertoires of Cretan Hieroglyphic and, in parallel, Linear A. Our conclusions point towards a strong connection with the Cretan Hieroglyphic milieu, in terms of sign shapes and direct links to seal imagery. In this light, the earliest writing in the Aegean is revisited not so much as a script in itself, nor as a prequel to Linear A religious sequences, but as a direct manifestation of the iconic glyptic practices of the Hieroglyphic tradition.
DOI: https://bmcr.brynmawr.edu/2020/2020.12.28/
Keywords: Numeracy; Abstract; Concrete; Writing; Tokens; Numerical notations
Abstract. Numeracy and writing constitute different phenomena, whose paths of formation often appear intertwined. Here we reassess the theory that numeracy evolved universally from a concrete to an abstract concept of number, and that that shift is correlated with the invention of writing. First, we gather contemporary linguistic data and early Mesopotamian epigraphic evidence that indicates that the ‘concrete’ vs. ‘abstract’ dichotomy is not useful to understand the emergence of numbers. Then, we discuss evidence from other regions where writing was probably invented independently, in order to investigate the conceptualization and formation of early numerical notations.
Download "Ahead of Print" PDF here: https://doi.org/10.1016/j.hm.2020.08.002
Keywords: Linear A script: Minoan Crete; Fraction signs; Constraint programming; Numerical notations
Abstract. Minoan Linear A is still an undeciphered script mainly used for administrative purposes on Bronze Age Crete. One of its most enigmatic features is the precise mathematical values of its system of numerical fractions. The aim of this article is to address this issue through a multi-stranded methodology that comprises palaeographical examination and statistical, computational, and typological approaches. Taking on from previous analyses, which suggested hypothetical values for some fractions, we extended our probe into assessing values for some problematic ones. The results achieved, based, on the one hand, on a close palaeographical analysis and, on the other, on computational, statistical and typological strategies, show a remarkable convergence and point towards a systematic assignment of mathematical values for the Linear A fraction signs. This contribution sets the agenda for a combinatorial, novel, and unbiased approach that may help advance our comprehension of some standing issues related to ancient undeciphered writing systems.
DOI: https://doi.org/10.1016/j.jas.2020.105214
Download the full article here.
Read the press release here.
Keywords: Rebus, acrophony, phonetisation, script formation
Abstract. Rebus and acrophony are crucial in the development of ancient invented scripts from Mesopotamia (cuneiform), China, Mesoamerica (Maya), Egypt, and scripts which may have been created through exposure to literate cultures (Anatolian Hieroglyphic and Nahuatl). Yet, these two linguistic mechanisms have been understudied from a terminological, contextual and comparative perspective. This article aims to address issues regarding their definition, development and application in script formation. The scope of our study is all attested writing systems that are largely iconic in their sign repertoire, and whose phonetic values were generated anew based on an underlying language (hence ‘invented’). This allows us to chart how writing systems are created ex novo and what trajectories of development are put into practice when phonetisation takes place. We show some reliable patterns of universal mechanisms, observable from a comparative perspective. We also demonstrate that these patterns attest to a verifiable degree of phonological awareness that ties the process of phonetisation to the path to script formation. We further highlight that the tendencies discerned from deciphered writing systems provide ways to test hypotheses in the study of iconic writing systems which are undeciphered, such as the Indus Valley script and the Rongorongo of Easter Island.
Reference: Miguel Valério & Silvia Ferrara (2020) Rebus and acrophony in invented writing, Writing Systems Research, DOI: 10.1080/17586801.2020.1724239
Accepted manuscript available for download here.
Keywords: 3D modelling, Laser scanning, Structured light, Precise 3D digitization, Rongorongo inscriptions
Abstract. Rongorongo is an undeciphered script inscribed on wooden objects from Easter Island (Rapa Nui) in the Pacific Ocean. The existing editions of the inscriptions, and their widespread locations in museums and archives all over the world today constitute a serious obstacle to any objective paleographical assessment. Thus, with a view to a potential decipherment, creating 3D models of the available corpus is of crucial importance, and one of the objectives of the ERC INSCRIBE project, based at the University of Bologna with Professor S. Ferrara as Principal Investigator. In this preliminary work, we present the results of the 3D digitization of the Mamari tablet, one of the longest inscriptions in Rongorongo, housed in the Museum Archives of the Congregazione dei Sacri Cuori di Gesù e Maria in Rome. The tablet is made of wood, with a shiny reflecting surface, characterized by a mainly dark texture. The 3D modelling was carried out with the ScanRider 1.2 laser scanner manufactured by VGER, based on Structured Light technology, taking care to ensure the legibility of each sign while preserving the overall shape of the object as precisely as possible. To overcome the difficulties inherent in the object’s complex fabric, the Mamari tablet was acquired in separate sections (joined together during processing through specific markers), thus managing to optimize the optical parameters of the laser scanner, such as the exposure of the camera and the depth of field of the projector. Furthermore, an evaluation of the 3D reconstruction precision was also carried out, highlighting a precision of few hundredths of millimeters, in agreement with the claimed nominal standard deviation. In addition to the 3D model produced, one of the main results of this endeavor was the definition of a successful method to scan such complex objects, which will be replicated to finalize the complete 3D modelling of the whole Rongorongo corpus of inscriptions.
Full reference: Lastilla, L., Ravanelli, R., Valério, M., and Ferrara, S.: 3D MODELLING OF THE MAMARI TABLET FROM THE RONGORONGO CORPUS: ACQUISITION, PROCESSING ISSUES, AND OUTCOMES, Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W18, 85–89, https://doi.org/10.5194/isprs-archives-XLII-2-W18-85-2019, 2019.
Keywords: 3D Modeling, Close-range Photogrammetry, Laser Scanner, Focus Stacking, Small Inscribed Objects
Abstract. 3D modelling of inscribed archaeological finds (such as tablets or small objects) has to consider issues related to the correct acquisition and reading of ancient inscriptions, whose size and degree of conservation may vary greatly, in order to guarantee the needed requirements for visual inspection and analysis of the signs. In this work, photogrammetry and laser scanning were tested in order to find the optimal sensors and settings, useful to the complete 3D reconstruction of such inscribed archaeological finds, paying specific attention to the final geometric accuracy and operative feasibility in terms of required sensors and necessary time. Several 3D modelling tests were thus carried out on four replicas of inscribed objects, which are characterized by different size, material and epigraphic peculiarities. Specifically, in relation to photogrammetry, different cameras and lenses were used and a robust acquisition setup, able to guarantee a correct and automatic alignment of images during the photogrammetric process, was identified. The focus stacking technique was also investigated. The Canon EOS 1200D camera equipped with prime lenses and iPad camera showed respectively the best and the worst accuracy. From an overall geometric point of view, 50 mm and 100 mm lenses achieved very similar results, but the reconstruction of the smallest details with the 50 mm lens was not appropriate. On the other hand, the acquisition time for the 50 mm lens was considerably lower than the 100 mm one. In relation to laser scanning, the ScanRider 1.2 model was used. The 3D models produced (in less time than using photogrammetry) clearly highlight how this scanner is able to reconstruct even the high frequencies with high resolution. However, the models in this case are not provided with texture. For these reasons, a robust procedure for integrating the texture of photogrammetry models with the mesh of laser scanning models was also carried out.
Full reference: Lastilla, L., Ravanelli, R., and Ferrara, S.: 3D HIGH-QUALITY MODELING OF SMALL AND COMPLEX ARCHAEOLOGICAL INSCRIBED OBJECTS: RELEVANT ISSUES AND PROPOSED METHODOLOGY, Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W11, 699-706, https://doi.org/10.5194/isprs-archives-XLII-2-W11-699-2019, 2019.