Working on Vissernet: an impression

Take a monumental work called An Historical Syntax of the English Language (by the late F.Th. Visser, published in parts between 1963 and 1973), take the even more monumental Oxford English Dictionary, and then extract the bits and pieces in the first work that should also be in the second. That is, in short, what Vissernet is all about: a project which started about a year and a half ago, involving a small multinational group of individual cooperators under the supervision of Edmund Weiner, Deputy Chief Editor of the OED.

Visser's Historical Syntax is a traditional grammatical work, containing thousands of sections, well-organized but without an index. Very much like the OED, it includes dated quotations, many of which are not in the OED and thus represent potential antedatings, 'gap-fillers', or postdatings. Additionally, some of Visser's grammatical observations differ from, or expand on, information provided by the OED.

Practically, most work on Visser boils down to reading through the grammar section by section, finding the relevant corresponding entry (or entries) in the OED and sorting out which quotations (if any) should be extracted. Relatively little time is devoted to comparing and extracting actual grammatical information. The selected extracts are keyed in templated form (North American Reading Programme SGML format) and sent off by e-mail.

From my personal experience, I find the main problem is locating the relevant OED entries. Visser regularly organizes his senses/forms differently from the OED, which means that somehow the often large number of quotations from a section in one work, presented as a block, must be classified according to the structure of the other work. In this process, I find the OED-on-CD-ROM's text search mechanism a great, if not crucial, help. Still, much time is spent on unsuccessfully reading through complete OED entries, only to find, say, that Visser presents a form with a prefix which the OED leaves off or puts under its own entry.

As usual, matters are rarely straightforward and in a number of cases I find myself wondering whether a spotted 'gap' in the OED is not actually a result of my failure to find the right entry. Yet the number of searching tricks I have got up my sleeve increases with experience, and yes, tracking down an obscure, irregularly spelt OE form scattered over three different OED (sub)entries ('Gotcha!'), and then coming across a nice juicy antedating ('Kazam!'), provides considerable satisfaction.

In the meantime, besides giving me obvious general experience, working on Vissernet has taught me a great deal about how close the historical bonds beween (Old) English and Dutch (my native language) really are. Sometimes I even notice that a OE word is closely related to a dialect form lost in standard Dutch. (I still speak my local 'Brabant' dialect, which is related to Flemish.) But perhaps most importantly, Vissernet has given me the opportunity to experience at close quarters the fascinating tension between the two main pillars of language description: grammar and lexicon.