Introduction to Scaife Viewer

N.B. More than a generation of learners have grown up with accessing and manipulating texts online with Perseus or the TLG. Now there is something that provides us with new tools and the contents of both: the Scaife viewer. I am happy to have a short guest post from Leonard Muellner, Emeritus Professor at Brandeis University, my first Greek teacher, and the one who introduced me to digital classics way back before Y2k

The Scaife Viewer, https://scaife.perseus.org, is an interface for the next version of the Perseus Digital Library. Here are some distinctive aspects of this new tool for reading and research:

1) The majority of the texts visible through Scaife are in Ancient Greek and Latin, but there are also texts in Persian, Chinese, Hebrew, and, as time goes on, other classical languages. All of the primary texts in the corpus are open and freely available in a variety of formats for the general public. There is a list of the several sources with links for downloading here: https://scaife.perseus.org/about/. Among the links is the ongoing First1KGreek Project, https://opengreekandlatin.github.io/First1KGreek/, which is intended to complete and supplement the Greek texts available from the current version of Perseus for the first thousand years of Greek from Homer to the Third Century CE, though it also includes later texts that are standard research tools for classics (like the Suda or Stobaeus). The plan is to complete this particular corpus by June, 2021.

2) The project aims to provide multiple editions of primary texts, multiple translations of primary texts into the same or different languages, and searchable apparatus critici of texts when copyright law allows. All of the texts in Greek and Latin have been tagged as to their parts of speech and forms, and several have also been treebanked, in other words, have embedded in them the results of morpho-syntactic analysis. As a result of this data, it will be possible to align translations, word-for-word, with the texts, so that anyone can survey what are the various ways of translating a specific word in a primary source, or what any given word in a translation goes back to in the original. All of these features are in various stages of development — some are, others are not yet available but will start to become so.

3) The Scaife viewer has two parts, a reading environment (Browse Library, on the home page, screen shot above), and a search environment (Text Search, on the home page. In the reading environment, users can call up translations alongside primary sources (“add parallel version” in screen shot, top of middle pane), and the software automatically generates word lists with vocabulary for the primary source on display in Greek as well as morphological and lexical information for any word in Greek or Latin (in Highlight mode, just click on the word). For Homeric texts, there is access to the New Alexandria commentaries (lower right pane in screen shot)— more is forthcoming in this space. Readers can also search within a given text, with lemmatized search — in other words, search for all the forms of a given word given its base form — available at the moment only for Ancient Greek. Any passage being read can be exported as a text file or with its XML markup (whole texts can be downloaded from the list of repositories given above under #1).

4) The search environment (screen shot above) of the Scaife Viewer is sophisticated: users can search for a group of words (by putting double quotes around them), combinations of words (“and” or “or” searches), partial word searches whose initial letters are known (with the rest indicated by *), and so forth. For Greek, lemmatized searches, for example, for phrases or combinations of words, can return helpful results. The interface allows for elasticity in the search terms as well, on a scale of 1-10; they can turn up thematic as well as dictional associations that you might not anticipate.

5) The Scaife viewer is an interface to a corpus that is in ongoing development, but also, the viewer itself is in ongoing development. In other words, neither of these is complete, and there are bugs in the software. The teams sponsoring the development of both projects, a consortium of institutions in the USA and Europe, is also developing tools and manuals for participation in the development of the corpus of texts by people everywhere. Another consequence of the incompleteness of the corpus and the software is that there are significant gaps in coverage and functionality, but many common texts and some exceptionally helpful functions are already for the public to use. Please give it a try.