Legends
Ancient coins from the Iberian Peninsula used various writing systems: Greek, Punic, Latin, Northern Paleo-Hispanic, Southern Paleo-Hispanic, and Southwestern, alongside unique glyphs and countermarks (Ripollès and Sinner 2019; Velaza 2019; Correa and Guerra 2019). During the 5th–2nd centuries BCE, only Emporion and Rhode used Greek writing, employing a fairly standardized alphabet, so reproducing their glyphs poses no difficulty. The Latin legends from the mints of the Ulterior province are in a similar situation to the Greek ones, as they adopted a very regular alphabet when introduced in the early 2nd century BCE. Therefore, Greek and Latin legends can be easily represented using Unicode fonts, which can be read and displayed by all browsers on any internet-connected device. Although the most common signs do not pose difficulties, it is true that occasionally some special forms are included, such as flipped or inverted Latin letters. The remaining writings, especially Punic and Iberian, present multiple issues since their numerous graphic variants do not fit the highly standardized Unicode model.
Punic issues from the 3rd–1st centuries BCE can be organized into various groups based on the variant of the Punic script used (Alfaro 1991, Alfaro 1998). The first group consists of mints that used a fairly standardized Punic script, such as Gadir, Ebusus, Seks, Olontigi, and Ituci. A second group includes mints that adopted Neopunic script with many licenses and peculiarities, such as Ebusus, Malaca, Abdera, and Seks. Finally, a third group contains mints that used an aberrant Neopunic script, also known as Libyphoenician, a group that includes workshops such as Asido, Lascuta, Bailo, Iptuci, Turri-Recina, Oba, Vesci, and Arsa, which used numerous glyphs that are hard to distinguish from one another and are difficult to read.
The coins of Iptuci are a good example of the diversity of aberrant Neopunic legends and the challenges they present in identifying their signs. Their inscriptions consist of vertical, angular, or rounded glyphs, whose notable variety of forms makes it difficult to identify the corresponding sign. Fortunately, many coins in this group also display Latin legends that mention the name of the issuing city, which is helpful for suggesting values for the signs and locating the city.
The legends written in the different local Iberian scripts are also characterized by numerous peculiarities (Ripollès and Sinner 2019). The Northern Paleo-Hispanic or Iberian script consists of 28 signs with well-known phonetic values but with many formal variants in their application as part of monetary legends. While many mints were very consistent in the choice of signs, others like Arsaos or Sesars created very inconsistent series, in which different alographs were alternated and mixed. The Southern Paleo-Hispanic or Iberian script is less well known than the Northern one because there are greater uncertainties about the phonetic values of some of its signs. The number of mints using the Southern Iberian script is small and located in the southeast of the Iberian Peninsula. Their series also show notable variability in forms.
Other Iberian legends are indecipherable or used non-existent forms attempting to emulate signs, as happens with the Iberian imitation drachmas of Emporitan style (Villaronga 1998). Only a small proportion of the legends on these Iberian drachmas are intelligible (De Hoz 1995). In some cases, engravers simply copied the Greek legend EMΠORITΩN, but in many others, they created signs devoid of meaning. These types copied the design and legend as a whole, so the meaning of the legends is merely ornamental, without any intention of imparting a particular meaning to the whole.
Given these circumstances, it is undeniable that creating a rigorous catalog of coin types requires a broad graphic repertoire of glyphs that allows for the faithful reproduction of the various scripts, their abundant variants, and even all those designs without meaning. Unicode fonts can be used to reproduce Greek and Latin legends, although a few singular forms like ligatures or orientation changes understandably cannot be part of their standard. Problems primarily arise when trying to represent Punic and Iberian legends, as Unicode fonts do not cover the broad range of graphemes present in these scripts. It is understandable that the numerous alographs of a single sign, forms with occasional use, or those whose phonetic value is unknown are not suitable to be part of such a standard. Therefore, the international Unicode system does not offer a complete solution to the needs of the project due to its highly normalized nature, unsuitable for incorporating rare signs or those of unknown meaning, which are of high interest within the MIB project. The need to represent such a wide range of signs and variants, sometimes requiring the combination of Unicode characters with others that are not standardized to reproduce monetary legends, was a challenge from the start in the development of the Numisdata management system.
Digital Representation of the Scripts
In 2000, Pere Pau Ripollès began drawing graphemes and organizing them into font characters to accurately represent the legends of ancient Iberian Peninsula coins in books and academic papers. By 2017, when the digital planning of the MIB project began, several fonts were already available that included almost all the graphemes and countermarks used on coins from the Iberian Peninsula between the 5th century BCE and the 1st century CE. This repertoire was distributed across six Paleo-Hispanic fonts, three Punic fonts, one with special characters, and three countermark fonts, all fully functional for use with any common program. This system of working based on the use of typographic fonts provided the reference graphic repertoire but was not suitable for developing the online work environment or the planned web catalog. The MIB project set up a new work and publication context that required designing a model for generating legends compatible with any browser.
The idea of creating a tool to construct and manage the legends of the coins emerged when the MIB digital project began working with Dédalo. The goal was to generate an environment in Numisdata where users could organize and select the abundant glyphs from these ancient scripts, which were outside any standardized system, and combine them with standardized characters like Greek or Latin ones. The possibility of starting a process to standardize ancient Paleo-Hispanic scripts within the Unicode consortium was also considered, but it did not fully solve the problem and was therefore discarded. The process of regularizing new scripts is slow and complex, and it does not account for multiple alographs or characters with unknown or absent phonetic values. It is logical that these monetary alographs do not fit into a coding system whose purpose is not to document variants. However, it is also understandable that within the MIB project, it is useful to meticulously record them, inevitably subject to some degree of subjectivity in choosing the forms. Another drawback of this standardization model, which becomes evident in the medium term, is the difficulty of incorporating updates as new signs are documented. Although the representation of some glyphs in the Unicode consortium is necessary, many others are rare or uncommon, and there is no reason to incorporate them into the standard, even though their representation in the MIB catalog is relevant. However, it is worth noting that, fortunately, a recent initiative has begun the process of standardizing the signs of Northern Iberian script (Ferrer et al. 2015).
Within Numisdata, the graphemes could have greater capabilities if managed from a standard section of the database, which would initially allow attaching metadata to them. Their inclusion as a graphic resource in a section would resolve most of the limitations of the Unicode standard, eliminating potential conflicts and enabling the association of additional information to the glyphs for setting search parameters. Any update would work internally and externally without the need for additional file installations. However, the glyphs would be stored in a basic table that would not allow grouping or customized sorting, operations that were essential for organizing this complex information. The definitive solution quickly became their integration into the system, but not through a standard section, rather through the thesauri, a section with hierarchical relationships within Dédalo that provides complete freedom for performing complex sorting and grouping of content.
Organization of the Glyphs in Thesauri
From the beginning, the project had a vast graphic repertoire consisting of hundreds of signs that could represent almost all the legends of ancient Peninsula coins. Despite this initial advantage, it was necessary to make an important effort to make them operational and expand their capabilities within the new content management system. After being incorporated into Numisdata’s thesauri, it would be possible to: a) sort and group the glyphs, b) associate information with each sign, c) search within different alphabets, d) create and edit legends using a user-friendly tool, and e) ensure the resulting inscriptions are compatible with any browser.
The thesauri in Dédalo adopt and share a single record model, where the term field constitutes the main concept of the record. These records are displayed in a tree interface, where descriptor terms can be freely moved and grouped hierarchically. In the thesauri of signs, the term field would include the differentiated descriptions of the various graphemes (e.g., inverted ki, r with extended peduncle, regular a 1, regular a 2…). Their vector drawings would be included in an image field in SVG format. Thesaurus records would also allow associating additional information with each glyph in different fields, such as definition, description, chronology, geolocation, and relationships with other records in the thesaurus.
The final organization of the glyph sets in Numisdata should be organized into six thesauri: Greek, Latin, Punic, and Northern, Southern, and Southwestern Paleo-Hispanic. Two more thesauri were added to gather symbols and countermarks. Since glyphs can be organized into groups and subgroups, they can be presented in order within hierarchies based on their phonetic value and form. Sorting is easy when the scripts are well known, but it becomes more complicated when grouping glyphs with uncertain meanings or singular forms.
To transfer all the information stored in TTF font files, it was first necessary to export all the graphemes contained in these fonts as individual SVG files. The open-source SVG vector graphic format was chosen for graphical representation within Dédalo, as it is a W3C standard with broad implementation that retains vector editing capabilities. The glyphs were organized into folders based on their phonetic value (vowels, consonants, syllables, or ligatures) and, within them, in subfolders based on their form (for example; a_1_Triangular inclined, a_2_Triangular inclined with peduncle, a_3_Triangular right angle, a_4_Triangular right angle with peduncle, a_5_Rounded, a_6_Rounded with peduncle). In this second level of formal description, it was inevitable to work using the subjective criteria of the research team. Once the forms were selected and organized, the next step was to create the corresponding epigraphic thesauri within Dédalo, reproducing the folder and subfolder hierarchies used to organize the SVG files. After importing them into Numisdata, it would be possible to attach linguistic, temporal, spatial, or any other type of information, allowing searches under various criteria.
After converting and incorporating this epigraphic information, it was necessary to develop a tool to create and edit legends within Dédalo. The new thesauri for ancient scripts, symbols, and countermarks included a total of 2,177 glyphs as of May 2019. The Numisdata system had to provide a tool that allowed researchers to easily create and edit legends. A tool was designed where, alongside each type's image, users could search for glyphs by typing their descriptions in a text field. The field for the legends also allows combining Unicode characters with any SVG glyph from the Paleo-Hispanic or Punic scripts, providing the necessary versatility to represent all types of inscriptions. The final result of this configuration is that any browser on any device can view the legends without problems, which can be considered a decisive achievement for presenting the coin types in the web catalog monedaiberica.org.