Wednesday, March 13, 2013

Benevolent Magic and "The Survey of Scottish Witchcraft"

[This post discusses the extraordinarily informative website The Survey of Scottish Witchcraft. Here is how they like to be cited: Julian Goodare, Lauren Martin, Joyce Miller and Louise Yeoman, 'The Survey of Scottish Witchcraft', (archived January 2003, accessed March 2013).]

Over at the "The Survey of Scottish Witchcraft" website, one can read the following, on the page "Introduction to Scottish Witchcraft":

"Q. Were the witches midwives or healers?
A. Not usually. We have recorded 9 individuals whose occupation was recorded as being a midwife, and for 10 people midwifery practices were included as part of the accusations of witchcraft levelled against them. This is a tiny percentage of the overall total. Folk healing was more common and featured in the witchcraft accusations of 141 people—about 4%. Even so, it was not something that the typical witch seems to have engaged in—though the beliefs that underpinned folk healing were closely related to witchcraft beliefs. If magic could be used to heal, it could also be used to harm."
On the face of it this seems very straightforward. A simple matter of arithmetic. But if we drill down a little into the database itself, the picture quickly become more interesting.

"Folk healing" and "midwifery" are two of the sixteen different categories that the researchers at the Survey of Scottish Witchcraft call "characterizations". Here are all sixteen in descending order (according to how many cases were found to possess the particular "characterization" in question): 
  1. Implicated by another Witch 948
  2. Demonic 528
  3. Maleficium 414
  4. Folk Healing 141
  5. Neighborhood dispute 120
  6. Fairies 113
  7. Unorthodox religious practice 85
  8. Demonic Possession 79
  9. White magic 70
  10. Property Motive 62
  11. Political motive 52
  12. Consulting 46
  13. Other/Unknown 23 
  14. Treason  19
  15. Refused charity to economic dependent 19
  16. Midwifery 10
Right off the bat one can see a major problem, for there is no way that these numbers add up to 3,413, which (apparently) is the total number of cases that was used to arrive at the "about 4%" figure, which in turn was used to justify the claim that magical healing was not something that the "typical" Scottish Witch was involved with. The disparity is even greater when we realize that there is a great deal of overlap due to the fact that many trials involved two or more of the sixteen "characterizations".

Let me put that another way, before getting into the gruesome details: the percentage given ("about 4%") is artificially small because the total number of cases used to calculate the percentage is greatly exaggerated. The total is exaggerated because it is mostly (see below) made up of cases for which we have no idea what the accused Witches in question were actually accused of. It is highly misleading to include those cases in which we have no information of any kind about what the accused person actually did, as if these cases represented positive evidence of what accused Witches did not do.

In fact, it turns out that less than half (1,511) of all the surviving trial records in the database have any information that allows for assigning any of the sixteen categories to them. This is easily verified by going to the "Search the database" page, then selecting "Search for cases of witchcraft by date and characterisation" link. Once there you go to the "characterization" table and select everything except for "any" (that particular selection is fairly misleading, for it does not mean "any characterization" but, rather "any case, regardless of whether or not it has any characterization assigned to it"), and then over to the right be sure to select the "Any selected option (OR)" button. This means you are selecting every case for which at least one of the sixteen different "characterizations" has been assigned. Then you click the 'Search Cases" button and, voila, a list of all cases matching that search appears, with the total number of cases helpfully appearing at the top of the list.

But it gets even worse, because almost half of these 1,511 cases, in turn, have "implicated by another Witch" as the only "characterization" that has been applied to the case, and this obviously has no bearing on what the accused Witch was actually accused of doing (or wasn't accused of doing). And so if we now eliminate those trials for which the only characterization was that it involved "implication by another Witch", which, to repeat, tells us nothing about what the accused Witch was actually accused of, there are only 876 cases left. Here is a brief summary:

3,413 cases total (the number of cases returned with all filters set to "any")
1,511 cases with enough information for one or more "characterization"
876 cases with any "characterization" other than "implicated by another Witch"

This suddenly increases by fourfold (up to 16%) the percentage of Witches involved in folk healing, if we limit ourselves, as we should, to only those cases where we have some factual, documented basis for saying whether or not the accused were involved in such practices.

But there is also another category that demands our attention: "white magic," which is category #9 in the full list of all sixteen "characterizations" listed above. For some reason, the term "white magic" isn't included in the (otherwise) very helpful glossary at the site, but it is discussed in the paper "Survey of Scottish Witchcraft Database Documentation and Description," where one reads that "white magic" encompasses a variety of activities including astrology, love magic, and finding lost goods. The grand total of cases involving either "folk healing", "midwifery", or "white magic" brings us up to 181 cases, which is 21% of all cases for which such information is available.

In other words, the answer to the question "were Witches midwives or healers?" is a simple "yes". This becomes even more the case if we further include other documented cases of beneficial magic ("white magic") along with "folk healing" and "midwifery". And on top of all that we must add the fact that Witches who practiced healing and other forms of beneficial magic tend to be systematically underrepresented in the trial records, because they were less likely to have charges brought against them .

The Survey of Scottish Witchcraft is an extraordinarily informative website. It is a shame that it is to some extent marred by such an egregious misrepresentation of the very data that it provides. The relationship between beneficial magic and Witchcraft is now a very "hot topic" in the scholarly study of historical Witchcraft. Unfortunately, a few scholars involved in this field appear to have an agenda that impels them to exaggerate the malefic characterizations of Witches and Witchcraft, while either denying outright, or systematically trying to diminish the clearly documented historical relationship between Witches and beneficial magic. Because certain widely respected scholars are now very aggressively promoting such an agenda, it is possible even for unbiased researchers, let alone the "lay" public at large, to unintentionally become convinced of (and even complicit in the spread of) crude misrepresentations of historical Witches.