Once the available examples of a particular word have been gleaned from existing slips and from digital databases, they are read through with a view to determining the precise meaning(s) of the word from the given context. Sometimes it is necessary to read more of the context than has been written out on a slip; at other times the context seems to give little help, as for example when all we have is the word in a list of unrelated commodities in household accounts or a ship’s cargo and the immediate context is simply the price of the item. Sometimes we are dependent on what our medieval predecessors thought a word meant, if our only evidence is the definition of the word given in a gloss in Latin, Old or Middle English, or Anglo-Norman French.
Sometimes we may only have one slip for a word, while for other words there may be several hundred slips to assess, from which we select those where the quotation best seems to illustrate the meaning of the word. If there are many slips for one word we will sort them into groups according to possible sense or subsense and by date, sometimes distributing them in piles on the desk or using a purpose-built, compartmentalized box. Working through these slips methodically by sense and/or date, we prepare definitions and also integrate the evidence we have found in digital resources, taking care that our definitions properly cover the evidence we have found and refining them as necessary during the process.
We always have to try and remain open to what the text is trying to tell us, rather than coming to each quotation with a preconceived idea of what the word means, for it is possible to have 99 examples with one meaning of the word and then find a new meaning in the 100th. Many words may have a classical sense, still used in medieval texts but supplemented by new senses covering, for example, new social, legal, ecclesiastical, mathematical, or agricultural developments. The most difficult words to define are perhaps the botanical names because neither we nor the medieval writers know what the hell is going on, most of the time: often all we have is a vague description of a plant (e.g. it has long hairy leaves) or a gloss where we may not be certain what the English plant name denotes either. Other common challenges include assessing ambiguous examples, where a writer has accidentally or even intentionally used a word in such a way as to be unclear which of two or more possible meanings is intended.
Historically, the use of slips written in longhand was specifically designed to facilitate this open-minded approach: slips could be quickly and easily reordered and regrouped as the process went on.
Similarly, in our modern XML-based digital drafting system, quotations are checked and entered so as to build up groups for which definitions can then be added and refined. These groups of quotations can be readily restructured and moved (or deleted) as necessary, whether as a group or individually. For instance, we may find at first a single example of a phrase, which we cannot therefore yet recognize as a phrase and so we enter it within the particular relevant subsense, but later when we find many more examples, we may divide these off as a separate adjacent subsense and include some of the further examples to document what we can then see as a common phrasal idiom. We may also find that we wish to regroup the subsenses differently into senses, and the drag-and-drop system is set up specifically to support editors in this kind of flexible drafting process.
Division and ordering of senses
We divide the evidence into senses and subsenses in such a way as best reflects the evidence we have.
For words with earlier Latin usage (in Classical or Late Latin) we are mindful of the ways in which they were used in the earlier periods, since many of those earlier senses continue into Medieval Latin. We thus consult relevant dictionaries, such as the Oxford Latin Dictionary and Souter's Glossary of Later Latin, when we create our structure of senses, fitting any new senses in and around the existing ones.
For new words, whether coined in Medieval Latin or borrowed from other languages, we start out from what we take to be the word's basic meaning, taking account of its source, its form and its use. Where it makes sense to do so, we proceed from more concrete to more abstract meanings, from more to less animate ones, and from literal to transferred or figurative senses.
The entries of related words are generally structured in a similar way to each other, where this makes sense.