Article body

While scholars almost immediately recognized the 2005 launch of Google Books as a watershed for archival research, the response was presumably more muted among educational publishers like Gale and Adam Matthew (AM), which by then had invested heavily in subscription databases featuring many of the books and periodicals now freely available online. With survival suddenly depending on staking a claim to some corner of the literary and historical archive that Silficon Valley’s scanners had missed, Gale has redoubled its focus on digitizing early newspapers, the print genre most glaringly omitted from Google Books.[1] Meanwhile, AM has looked beyond the world of type to handwritten texts, striking deals with several leading British and American archives to digitize (and monetize) their manuscript collections. As a result, those fortunate enough to have institutional subscriptions, can now, from the comfort of their home or office, explore the Huntington Library’s collection of all plays licensed in Britain between 1737 and 1824, the Stationers’ Company’s ledgers of texts printed in England since 1554, and, as of 2020, the hoard of book-industry records in the National Library of Scotland’s John Murray archive.

One of AM’s first major coups in this area came in the early 2010s, when it secured the rights to digitize the Wordsworth Trust’s (WT) entire manuscript collection and many of its artworks. The resulting database, which AM somewhat hyperbolically markets as Romanticism: Life, Literature, and Landscape, is less a smorgasbord of all things Romantic than an archive concentrating specifically on a single set of writers (the so-called “Wordsworth Circle”) and a single geographic region (the English Lakes). Building upon foundational bequests by the Wordsworths’ descendants, several generations of curators at the WT and its forerunner, the Dove Cottage Trust, have gathered 90% of William Wordsworth’s verse manuscripts, all of Dorothy Wordsworth’s extant journals and notebooks, and a wealth of correspondence by the siblings’ closest friends and relations. By digitizing these and hundreds of other major literary manuscripts owned by the WT, AM has largely obviated the need for researchers at subscribing institutions to make pilgrimages to Grasmere to pore over drafts of The Prelude, peruse Dorothy Wordsworth’s notebook poems, or explore manuscripts by such writers as Samuel Taylor Coleridge, Robert Southey, John Wilson, Thomas De Quincey, Maria Jane Jewsbury, Felicia Hemans, and Matthew Arnold. Further, the Romanticism database digitizes some 2,500 of the 9,000 works in the WT’s art collection, concentrating on portraits and landscape paintings with direct relevance to William Wordsworth’s life and works.

Compared to the most robust single-archive databases—including some of AM’s own recent releases—Romanticism underwhelms in the supplementary resources it offers for appreciating and navigating this remarkable collection. While valuable in its own right, the “User Guide” is less an instruction manual for this site than a descriptive catalogue of the WT’s holdings. Other contextual materials, including a set of concise “Literary Lives” and five short introductory essays on the Wordsworths and Romantic art, only obliquely reference the collection’s strengths, rarely mentioning and never linking to specific manuscripts or paintings found therein. Worse, two of the four features grouped under the “Further Resources” tab—a set of “External Links” and an “Archive Explorer”—are so riddled with broken or outdated hyperlinks as to be essentially unusable. While “link rot” eventually infects almost any website, decay on this scale is usually associated with abandoned personal websites, not pricy educational databases.

That the site’s supplementary features have fallen into such disrepair may suggest how rarely they have been used since the product’s 2012 release. Despite the myriad ways the Romanticism database might be used in classrooms, its heaviest users have presumably been specialist scholars looking for specific manuscripts rather than a general introduction to the period. For such users, a greater source of aggravation than the site’s broken links, then, will likely have been its maddeningly unreliable search tools. Browsing the WT collection is significantly more cumbersome using AM’s interface than it would be on site in Grasmere. Of course, this may be more a liability of the medium than this product, for, as Jon Klancher recently observed, the digital archive “vastly increases our access” but “seriously compromises our ability to browse or explore” large collections.[2] Still, AM has made no effort to mitigate the browsability problem by offering, say, curated sub-collections focusing on commonly researched genres or themes. And clicking on the navigation menu’s “Documents” tab yields not a road map for the entire archive but a generic “search results” page listing every item in the collection (see Fig. 1 below).

Figure 1

The site’s standard header and navigation menu and, beneath it, the generic “search results” list that appears under the “Documents” tab.

-> See the list of figures

Since Romanticism’s search functions pull largely from metadata in the WT’s catalogue, the most dependable results generally come from inputting proper names or formal titles. A basic keyword search for “Bishop of Llandaff,” for instance, yields not only William Wordsworth’s incendiary Letter to the Bishop of Llandaff from 1793 but digital copies of several unpublished letters by and to this eminent cleric and politician. The internal search engine strains, however, to account for alternate spellings. Hence, typing in “Bishop of Landaff”—a spelling the prelate himself generally used—produces just a single result. More vexingly, as seen in Figure 2 below, searching for “Grasmere journals” (omitting the surrounding quotation marks) yields ten hits, but not a single one of the four manuscript notebooks (DCMS 19, 20, 25, and 31) comprising Dorothy Wordsworth’s famous diary. Searching instead by manuscript number is nearly as unfruitful, as keying in “DCMS 20” (sans quotes) does bring up the correct notebook but buries it midway through a list of 39 hits, many of which have no discernible connection to this specific manuscript.

Figure 2

Search results list for “Grasmere journals” (sans quotation marks).

-> See the list of figures

Having located the desired manuscript, though, the user experience improves considerably. All items I have accessed, whether letters, images, or diaries, have been expertly digitized in high-resolution formats. AM offers three ways to study any particular document: perusing it on the main landing page using directional and zoom controls, saving it to a separate “Lightbox” and exploring it there, or downloading it as PDF. Especially when working with multi-page manuscripts and sloppy or cross-written letters, I find it considerably easier to download texts and read them in a PDF reader like Acrobat than to study them within the database itself. But when working with relatively legible, single-page manuscripts, the built-in reading tools are perfectly adequate.

All told, then, AM’s Romanticism: Life, Literature, and Landscape is a decidedly mixed bag. Although the site’s interfaces and supplementary features leave much to be desired, the remote access it provides to the WT’s extraordinary collections is a godsend to scholars researching the Wordsworths and their circle. According to AM representatives, the company plans to significantly revamp this database’s interface in the near future. Hopefully any such reboot will not only improve basic functionality but also capitalize on major technological advances from the decade since AM first designed this product. Topping the wish-list of most heavy users would presumably be integrating the Handwritten Text Recognition (HTR) technology featured in AM’s new database of the Murray archive, Nineteenth Century Literary Society. According to the company’s marketing materials, HTR “is a form of artificial intelligence that uses probability to assess handwriting and deliver full-text search results in manuscript documents.” In other words, this emerging technology aims to enable researchers to find incidences of a certain word or phrase in a corpus of manuscripts as readily as they have become accustomed to doing with print since the advent of Optical Character Recognition (OCR) software. Adding HTR to the Romanticism database would therefore dramatically enhance scholars’ ability to find mentions of specific people, places, and texts that don’t appear in the catalogue’s metadata. In short, it would be the type of upgrade that would give this world-class literary archive the sophisticated and reliable interface it deserves.