Working some lexicographical magic: the revision of witch

As we move from summer to autumn and Halloween appears on the horizon, it is apt that a number of the words for the latest OED Online quarterly release have spooky or spectral associations: ghost, ghoul, hag, voodoo, witch, and zombie, to name but a few. As luck (or fate) would have it, I was involved with the revision of a number of these entries, but witch in particular stood out both in terms of the size of the entry and its continuous use from the Old English period onwards. How does it look now that we’ve worked our magic on it?

In Old English the word appeared in two forms: wicca (pronounced something like ‘witch-ah’), denoting a man who practises witchcraft or magic, and wicce (pronounced something like ‘witch-eh’) denoting a woman who practises witchcraft or magic. Because of this—and because for hundreds of years witch has been used much more often to denote a woman than a man, meaning that examples of witch referring to a man seemed like a different sense—the first edition of the OED opted to present the ‘male’ and ‘female’ senses as two distinct headwords. However, after some discussion, we decided to merge the two: the grammatical distinction between wicca and wicce in Old English is a separate issue from the way in which witch has been used from at least the Middle English period to denote women more often than men, and we decided that it would be more helpful to tell the word’s story in one place rather than two.

In fact, it was the story of the word, rather than its simple meaning, that I kept returning to. The new definition of sense 1a of witch, the oldest and most common sense, is relatively short and straightforward:

A person (in later use typically a woman; see note) who practises witchcraft or magic, esp. of a malevolent or harmful nature.

However, the ‘note’ in question is a miniature essay, covering inter alia the history of belief about magic and magical power, the way in which witch came to be associated particularly with women, and popular depictions of witches, all of which is important in understanding the way in which the word has been, and continues to be, used, and provides the context in which the core definition operates. The same also applies (to a lesser extent) to sense 1b(b), which deals with a specific type of witch; namely, the ones associated with the period of increased fear of witchcraft in the late middle ages and the early modern period that led to witch trials such as those at Pendle in Lancashire in 1612, and at Salem in Massachusetts in 1692-3. The first edition of the OED covered this as part of its ‘female practitioner of magic or witchcraft’ sense, but it seemed sufficiently distinct to be worth treating as a sense in its own right. Again, the note provides the historical context in which this sense is to be understood. You will also note, if you look at the section towards the end of the entry devoted to compound words, that a large number of these originate in this period of heightened fear of witches, including witch-cake (first recorded in 1616), witch-lock (1682), witch’s mark (1624), witch-meeting (1693), and witch-trial (1694). Others are coined later but refer to this period, perhaps most notably witch hunt n. (1853), which subsequently acquired a more general sense of ‘a period of persecution’ and so is treated by OED as a separate headword.

The fact that sense 1b(b) has been so productive in terms of compounds is a good argument for its treatment as a separate sense, and this shows how it is important not to think of individual senses and lemmas as isolated units of meaning but as parts of a whole, influencing and being influenced by each other. This was particularly the case when considering the question of how sense 1a came to be so closely associated with women. Although, as we have seen, the Old English word wicce denoted a woman who practised magic, we don’t have enough evidence to be able to say whether or not people spoke more often of a female wicce than a male wicca. However, the appearance of sense 3a (‘as a term of abuse or contempt for a woman’) in the fifteenth century strongly suggests that witches were thought of as women at that time if not earlier, and the appearance of compounds such as ‘he-witch’ and ‘man-witch’ round about 1600 indicates that it was necessary by that point to make it clear that a particular witch was male rather than female.

Similarly, one of the new senses added to witch turned out to fit amazingly well with the chronology of another part of the entry. Sense 1c denotes practitioners of magic within other, non-Christian or non-European cultural contexts whose position resembles that of the traditional European witch. When I began drafting this sense, I had a clear idea of how it is used in the present day, but wasn’t sure how far it would go back. Eventually I was able to take it back to 1731, and noticed that this corresponds very closely with the date of the final Witchcraft Act (1735), which made it illegal to accuse someone of being a witch, and so marked the effective end of the ‘witch trial’ period in the English-speaking world. It therefore makes sense that the word began to be used at this time in a more ‘anthropological’ sense. 

Many other changes, large and small, have been made to witch: we might note, for instance, the newly added sense 5, ‘a follower or practitioner of a modern pagan belief system that draws on pre-Christian religious history in its beliefs and rituals’, which unlike other senses can often be found as a self-designation rather than a label bestowed by others. But I hope this gives a glimpse into the process of revising a word so closely embedded in the history of the English language and the culture of its speakers.

The opinions and other information contained in the OED blog posts and comments do not necessarily reflect the opinions or positions of Oxford University Press.