AIO (epigraphers love abbreviations) has a simple premise: translations are essential to make the inscriptions of ancient Athens and Attica more accessible; and translations are rarely available. CIL and IG continue to use Latin as their lingua franca for all supporting information, although IG has begun to include online translations for new editions, primarily in German, but also reproducing those of AIO for Attica. As someone reared on sourcebooks without translations such as R. Meiggs and D. Lewis, A Selection of Greek Historical Inscriptions (revised 1989), I only wish AIO had existed when I was a student. AIO was founded by the Greek historian and epigrapher Lambert and launched in 2012. It currently hosts 2,727 inscriptions, with new batches of material uploaded several times a year. Lambert has now stepped back from the immediate direction of the project, but the baton has been ably picked up by de Lisle, Low and Liddel, supported by an Advisory Board; and the list of contributors is almost a who's who of Attic epigraphy.
AIO does what it sets out to do remarkably well and to a very high level. The site is clean and simple in its presentation, and set up simply to get you to the inscription(s) you are looking for with the minimum of fuss, allowing you to search by a variety of basic metadata categories (date, findspot, original or current location, inscription type, monument type, date of publication in AIO) or by bibliographical references. Having found your inscription, you are presented with the translation, which clearly follows the arrangement of the original text. In some cases the translation can be toggled with the specific edition of the Greek text that was used, in others you are provided with a link to the source text, available online in another resource, such as PHI or IG. Translations are annotated with transliterations of key terms and placenames; both features can be turned on or off, the latter linking to the Pleiades gazetteer of ancient places. Translations are accompanied by a detailed commentary, rich in hyperlinks to cross-referenced texts and places. Given that most inscriptions have been repeatedly edited, the different editions are all recorded, and the translations can be reached via any of those references (which are in turn detailed on the page); the source edition (usually the most recent) is made explicit, and the translator(s) noted.
There is much more to AIO than translations alone: the site hosts three additional resources: ‘AIO Papers’, ‘AIUK Papers’ and ‘Attic Inscriptions: Education’, a set of supporting resources for schools. The first of these offers a forum for discussion of Attic epigraphy, especially work arising out of the translation project. Papers are published as online PDFs. Among the twelve available, one might single out no. 4, by Lambert, which is a fine demonstration of what can be done by one person systematically working through a fascicle of IG for the purposes of translation, illuminating brilliantly the extent to which close translation requires detailed scholarly engagement with the material. AIO Paper 12 should also be highlighted, a thorough overview of the Athenian ephebate and its epigraphy in the Roman period by de Lisle, a small monograph in its own right. The second resource, AIUK, edits the Attic inscriptions held in UK collections, work undertaken by Lambert, Liddel, Low, de Lisle and R. Pitt. Fifteen papers range across country houses and museums, producing first-rate, fully illustrated editions with commentary, including a previously unpublished ephebe list of the Roman period in the National Museums of Scotland in Edinburgh.
In the face of such remarkable work, all available in open access under a CC-BY licence (except for AIO Papers, which are copyrighted and reserve all rights), and such a genuinely valuable resource, it seems deeply unfair to quibble, but in a spirit of constructive criticism, a few observations might be made. Firstly, an epigraphic point: if I read the data correctly, there are currently 106 texts for which AIO presents a ‘new’ text (see ‘AIO’ under ‘source texts’), the editors/translators having determined that the existing editions are in some way unsatisfactory. One could hardly take issue with that. However, there is no indication of responsibility (the translator?), and neither is any sort of apparatus provided to indicate where or how this text diverges from other published texts. In the random example I interrogated, AIO 803, I could not identify any supporting discussion in the AIO Papers. Some old-fashioned spade work in the library confirmed that the online text diverges from P.J. Rhodes and R. Osborne, Greek Historical Inscriptions, 404–323 bc in three places, but follows that of IG II2.42, except at line 20, where it adopts the reading of Matthaiou 2013 (which is listed among the previous editions). Given that the difference matters, it is slightly disconcerting that this is nowhere highlighted or discussed, and it is oddly out of line with the quality and attention to detail characterising most aspects of this project. To add a positive note, comparison of the translation with Rhodes and Osborne makes clear the consistent attempt to offer a translation that is closer to the structure and grammar of the original Greek.
If AIO has a weakness, it lies in its documentation and its data accessibility, in other words, its performance as a digital resource. The onsite instructions and explanations are minimal, and for a non-specialist it will still be a challenge to find what they are looking for. The only documentation lurks in AIO Paper 10, last updated in 2020, which offers some basic FAQs, but does not, for instance, explain the principles adopted by the translators, which it would be interesting to know (including the principles underlying the decision to provide transliterations for some terms). Slightly disconcertingly, while you can reach any inscription via reference to any of the published editions, and although the individual inscription page spells out which edition is used for the translation and so is primary, the URL at the top of that page will be constructed on the basis of the edition by which you searched, so that exactly the same content page is represented by as many as five or six different URLs. This cannot be considered good practice and risks being very misleading: unique AIO numbers are assigned to every text, but are relegated to tiny text at the foot of the page and do not seem to serve any public function. This matters, since stable URLs and unique identifiers are an essential part of meaningful citation, human or digital: and as a real piece of scholarship, citation should be enabled and supported. Documentation on and access to the data is also rather obscure. By chance, I stumbled upon the seemingly undocumented existence of an API (https://www.atticinscriptions.com/api/), which is how I believe there to be currently 2,727 unique AIO records (not information otherwise published). The API will return both Greek and English texts in EpiDoc, although they are encoded in Base 64, adding an extra hurdle. The EpiDoc encoding is minimal, lacks any of the metadata and does not include line numbers; so its purpose remains unclear. Similarly, documentation for the searches is lacking, and it is not immediately obvious why each of the searches returns a different total number of inscriptions, ranging from 591 to 9,639 (although explanations could certainly be imagined); only ‘publication date’ returned a total of 2,727. The current text-searching facility also still needs work, as exact string searching with quotation marks does not in fact return exact string matches only, contrary to what is suggested. These points matter, not simply as quibbles, but because the whole point of generating digital data is that this enables one to do more with and ask more of the material than might have been intended by a simple translation project: currently the opportunity is missed.
Quibbles aside, this is a fantastic resource. Furthermore, it is already showing quite remarkable endurance and longevity as a digital resource. AIO-inspired resources such as Greek Inscriptions Online (https://www.greekinscriptions.com/) have not fared so well (first and last update 2018), and related resources such as the Epigraphic Landscape of Athens (http://www.epigraphiclandscape.unito.it/index.php) have similarly fallen dormant (last update 2019?). Doubtless much of this is due to Lambert's dedication. Having been brought under the aegis of the British School at Athens in 2021, and with a refreshed editorial board, one can only hope that AIO will eventually achieve his aim of including all 20,000 Attic inscriptions.
