The 'Thesaurus of Old English' database: a research tool for historians of language and culture

Abstract
The “Thesaurus of Old English” is the not inconsiderable offshoot of the Historical Thesaurus of
English, a project based at Glasgow University.
Although it is a research tool of considerable
interest in its own right, it is also intended that the
TOE will serve as a “pilot”, in which the classification structures and working practices designed
for the Historical Thesaurus may be tested. The
printed “Thesaurus of Old English” (TOE), recently published by King’s College London Medieval Series, is drawn from an INGRES database,
and other published resources, paper and electronic, are planned. The proposed presentation will
explore some facets of the corpus of interest to
historians of language and culture, and display the
database in practical use. Some technical aspects
of the database design and management will be
addressed, and some lessons learned by us in the
process of managing it, and in generating the
printed work with its index, will be discussed.
These lessons will prove invaluable in feeding into
the work of the Historical Thesaurus, and in preparing the later TOE publications.
The “Thesaurus of Old English” has been created
as a database, using the INGRES relational database software. Although the relational capabilities
of the software have not so far been exploited, in
the future we may want to use them as we move
away from the original tasks of preparing both the
Old English constituent of the Historical Thesaurus and the independent volumes that comprise the
published, paper, edition of the “Thesaurus of Old
English”. The database contains (Modern English) subject headings, and for each subject heading one or more Old English words. In total there
are at present 72,612 database records, of which
50,502 represent Old English words and 22,110
Modern English headings. The subject headings
are grouped by category and sub-category. Each
group and sub-group has a classification number
associated with it for identification purposes. Each
database record can be uniquely identified by its
category and sub-category references and a serial
number. Other data stored for the Old English
words include part of speech, a primary alphabetical sort key, Roget categorization (dating from
the time before the new classification was devised
by Michael Samuels and Christian Kay, but still
of use in placing slips as a clue to where errant
ones should belong) and cross-references; also we
record a modern or earlier English etymological
tie-up. A further field affords a space for comments: for example, a location reference for each
rare word, meaning or usage; this field also serves
as a convenient space for recording observations
of the sort that would normally go in a footnote.
Finally, we have a field in which we put a code if
the lexical item is a hapax legomenon – that is to
say, it occurs once only in the corpus of Old
English texts – (o), is recorded only in poetry (p),
is recorded in glosses (g) or in some way has a
query attached to it (q).
Text is recorded in the database in ASCII, so the
Anglo-Saxon characters thorn and ash have to be
entered as symbols (we cheat slightly by having
thorn stand for both eth and thorn). Thus, in the
database thorn is represented by “}” and ash by
“{”, and these conventions, together with an underscore to denote the macron of a long vowel,
serve to present the language adequately. At present, using the database requires an effort of transliteration, one which is acceptable to its primogenitors, but not, probably, to the researchers we
hope will want to use the database materials for
ongoing research. We are therefore determined
that any electronic version should carry with it an
acceptable Anglo-Saxon font (in particular, it
must offer macrons to mark long vowels).
The database classification starts within each new
category with the most general concepts or ideas,
and descends to more and more refined headings.
Here, for example, are the categories at the beginning of the classification of words to do with the
activity of the mind:
06 Spirit, soul, heart
06.01 The head (as seat of thought)
06.01.01 Thought, the faculty of thinking, mind
06.01.01.01 Thinking about, minding, heeding
06.01.01.01.01 Thought, cogitation, meditation
06.01.01.01.01.01 Consideration, rumination
06.01.01.01.01.02 Forethought, consideration
It is often said that modern human beings have lost
the faculty their ancestors had of being able to
memorize very long stories or histories. Feats of
memory (such as the recitation of a saga many
hours long) have certainly been recorded in the
twentieth century, but in general this facility of the
mind is not something of which we boast. For the
Anglo-Saxons the mind, the active seat of thought,
is conceived of simply as within the body, which
itself may be viewed as a dwelling place (“sawolhus”) or a container, a coffer (“hordcofa”) for the
inner life. The mind, the heart, the spirit and the
soul are all manifestations of that inner life, and
the thesaurus illustrates the range of ideas recorded for the way in which the mind works. That the
mind may be regarded as a treasury kept within the
body is illustrated by the concepts of “sawolhord”
or “lichord”, the metaphysical hoard corresponding in value to a hoard of gold and jewels. It is
here that the intelligence is locked away for security (“gewitloca”). More than fifty nouns attempt
to give some concrete account of the inner life of
the body. Over forty verbs express the ideas of
thinking, considering, reflecting, pondering, illustrating a way of looking at the mind that recognizes the hugely wide-ranging capacity of this most
mysterious of human functions.
Searching the database at present is by one of two
methods: the Query-by-forms procedure provided
within INGRES, or SQL. However, making effective and fruitful searches is at present very much
a matter of competent use of SQL. In the future,
the database will have a more user-friendly frontend which will allow ready access to the data
(possibly with an option of adding comments to
the discussion of a particular word). The results of
a sequence of searches may perhaps be placed in
adjacent windows for comparison, or further refinements of a search already made may be used to
bring the researcher ever closer to the fields required. Progress with these developments will be
described, and demonstrated.
The paper will describe briefly the technical problems encountered in producing the paper publication of the TOE, and in particular those related
to the generation of the word-subject index, including issues in the secondary and tertiary sort sequences which were needed to accommodate both
words and phrases. The automated procedures
developed for producing camera-ready copy from
the database using LaTeX will also be briefly
described.
However, the main emphasis in the technical part
of the paper will be on the issues related to the
electronic publication of the thesaurus material.
The project is addressing questions of medium
(CD-ROM and/or on-line database access and/or
Internet access), format (“mainframe” database,
personal computer database(s), SGML mark-up)
and user interfaces and manipulation tools. The
project is not only taking note of work done elsewhere, but is also carrying out small-scale comparative studies with sample materials drawn from
the database. The key objective is to find an appropriate balance between effectiveness of use and
technical complexity. The paper will describe the
progress of these developments, and the conclusions drawn.
With the ongoing publication of fascicles of the
Dictionary of Old English (by the University of
Toronto) there is a great deal of interest in the
vocabulary of Old English at the moment. The
thesaurus offers a unique way of exploring what
words were available to the Anglo-Saxons when
they wanted to call a spade a spade or when they
wanted to consider the most remote elements of
the nature of God (although naturally this record
must be recognized as incomplete because of the
haphazard survival of Anglo-Saxon texts). The
thesaurus documents, using all the sources available to us, the way the Anglo-Saxons denoted their
world.