Building External Coherence

Abstract

This chapter explains the principle of external coherence within an idea system. It presents how witch accusations in early modern Europe acquired a strong coherence beyond the local contexts—venturing into trans-local and regional contexts. The principle of oracular reasoning—consisting of core propositions and secondary elaborations—is explained through an idea-orbit visualization.

External coherence encompasses the bridging of patterns or ideas pertaining to local cases with patterns or ideas pertaining to a broader set of cases. The bridging can be solid or tenuous, depending on the amount and strength of the bridging mechanisms.

As a starting point, activating the same set of preexisting knowledge, built on the same codes and coded things, can establish the basic precondition for external coherence. But more intricate bridging works need to be performed as it is necessary to discuss and work on the details of similarities and differences between things.

Sophisticated theories, treatises, and manuals are intellectual resources that played a central role in forging external coherence. They allowed individuals, particularly judges, to quickly establish similarities across cases, often borrowing the general ideas from texts and authorities to draw “parallels” between data and between ideas. They also helped judges to quickly process apparent contradictions, ambiguities, and differences. In effect, doing so helped judges to effectively mitigate theoretical contradictions that would threaten the specific case. In turn, these processes contributed to the overall integrity of the larger idea system, when the local interpretations do not contradict the larger body of ideas. An ideal scenario of maximal external coherence is a global-local unity of sorts, in which the ideas, voices, and data of the “global” level cohere and harmonize with those of the local level.

Relational Bridges: Thin Versus Thick Parallels

The words, ideas, and textual resources that bridge different bodies of data or ideas can be called relational bridges (or relational bridging points). Bridges can be established when people could cognitively connect two or more objects or ideas that are otherwise perceived to be unrelated.

In the context of establishing external coherence, seeing relational bridges between a local and an extra-local text or case does not necessarily mean seeing the text or case to be identical. Rather, a point of relation exists, and that relation is built on a point of parallel of sort. The parallel is based on the assumption that that the two bodies of data or ideas are not completely independent, but due to some principles of parallel operation they have particular relational connectedness. They may indeed be connected by their analogous similarities, or they can be connected by other kinds of relations.

Imagine a scenario where one person tells another that they are “long-time friends” in a context—say when they need the other’s assistance, allowing all kinds of communication, feeling, interaction to flow. Upon a change in context, that person degrades the analogous relationship to an “acquaintanceship,” asserting much more mutual independence and seclusion. These two statements evoke different parallels based on similarities with two classes of idealized mental objects within a culture.

In another scenario, if the Greek god Zeus indeed casts a lightning bolt onto Earth, the parallel relation there is not to say that humans are “the same” as gods, or that the two worlds’ conditions have collided into one. Instead, the parallel relation is that what goes on in Earth—at least for one single event—is causally connected, that the outcome in one world could be causally explained by what goes on in another world running “parallel” to it. That relation is not treated as analogous but actual.

It is the latter kind of connectedness that was most relevant to the ubiquity of witch trials. Due to the perceived actual parallels between the supernatural and the natural, the data and ideas of local cases, to a certain extent, became conceivably applicable to other cases, and vice versa, if there is a common set of causes. Many efforts were conducted to establish and theorize specific points of parallels, bridging the local cases to a broader context.

Parallels can be built on incredibly thin connections—or thin relational bridges. The requirement is low: for example, one could simply make two things “look alike” and then invite the audience to attend to their similar properties, prompting the audience to feel that the two things indeed “operate alike.” Saying “money is just like cocaine” is an example. It accounts for the similarity in addictiveness but not their broad range of functions. People could use numerous analogies, proverbs, maxims, metaphors, or similes to create flash- or momentarily intersecting points of parallel, and in turn a momentary coherence.¹

These thin relational bridges, however, could be supplemented with or bolstered by further works—with thick relational bridges. Thick relational bridges are not merely many thin bridges that add up in numbers; they are built with more enduring, complex, reinforced structures. When there are thick relational bridges, people could transpose the knowledge between different domains much more easily. In other words, rather than quickly looking at two things that “seem alike” and presuming they must “operate alike”—thus walking a tightrope cognitively—people could actually get into details of explaining why the two things are alike or are connected. They would, metaphorically speaking, be walking on a wide and solid bridge with ease and grace.

French studies scholar Virginia Krause discusses the ways individuals relate particular cases to the broader demonological corpus, which serve as bridges to various biblical, literary, and legal eruditions. Demonological writings set up numerous “parallel structures” (and methods to create such structures) that helped individual judges and actors to relate what happens in one case to other similar classes of cases, the missing gap or information from one party or one case could be “filled in” by another party or by general understanding derived from other cases at the other side of the bridge.²

Furthermore, numerous existing quotes, tales, and facts in the larger corpus could also be resources for problem-solving. Suppose one finds that some inconsistencies exist between an understanding of what witches could do, or that some unresolved mysteries or contradictions have arisen. A wealth of past cases could then be consulted and referenced to determine how similar mysteries and contradictions were resolved elsewhere. This process evokes new, plausible explanatory hypotheses to resolve the existing case. Even if the individual actors in a location come up with this conclusion on their own, their findings are doubly reinforced by other cases.³

Consider one problem that had surfaced. If a convicted witch has children who were implicated in a trial, should the children simply be assumed to be free of diabolical qualities? By looking into past cases where witchcraft played out as a hereditary crime, Nicholas Remy’s stated: “experience has shown that they [the children of witches] who have fallen into the power of the Demon can rarely be rescued except by death.”⁴ After reviewing a few cases to support his idea that the Demon seeks to corrupt the offspring of children, making them commit sin or execute witchcraft at an early age, Remy quoted this text:⁵

The breed shows its descent;

Degenerate blood reverts to its first type.

(Seneca, Hippolytus)

Here, the universe of the Devil—as described in demonology—is made parallel to the universe of local cases pertaining to children raised by witches. Beyond simple quotations, this relational bridge was thickened by multiple precedents cited by Remy.⁶

Theorizing variations is especially interesting for the purpose of building relational bridges. The basis of the parallel relation could be one of correlation, of resemblance, of variance, or of derivative. Establishing the relation of a variance is to assert that the common cause that has a unifying essence may show up in different forms. Taking sexual experience with the Devil as an example, as discussed by Remy:

But all they who have spoken to us of their copulations with Demons agree in saying that nothing colder or more unpleasant could be imagined or described. At Dalheim, Pétrone of Armentières declared that, as soon as he embraced his Abrahel, all his limbs at once grew stiff. Hennezel at Vergaville, July 1586, said that it was as if he had entered an ice-bound cavity, and that he left his Schwartzburg with the matter unaccomplished. (These were the names of their Succubas.) And all female witches maintain that the so-called genital organs of their Demons are so huge and so excessively rigid that they cannot be admitted without the greatest pain. Alexée Drigie (at Haraucourt, 10th Nov., 1586) reported that her Demon’s penis, even when only half in erection, was as long as some kitchen utensils which she pointed to as she spoke; and that there were neither testicles nor scrotum* attached to it.

To offer more cohering evidence, Remy goes on to say:

Claude Fellet (at Mazières, 2nd Nov., 1584) said that she had often felt it like a spindle swollen to an immense size so that it could not be contained by even the most capacious woman without great pain. This agrees with the complaint of Nicole Morèle (at Serre, 19th Jan., 1587) that, after such miserable copulation, she always had to go straight to bed as if she had been tired out by some long and violent agitation. Didatia of Miremont (at Preny, 31st July, 1588) also said that, although she had many years’ experience of men, she was always so stretched by the huge, swollen member† of her Demon that the sheets were drenched with blood. And nearly all witches protest that it is wholly against their will that they are embraced by Demons, but that it is useless for them to resist.⁷

* Remy added a footnote cross-referencing De Lancre.

† Remy added a footnote cross-referencing Bouguet’s and De Lancre’s experiences.

There is clearly a cohesiveness in the vivid details across cases, even though the actual details differ. Each differing description adds some partial understanding and empirical data to the general topic. Coldness, coercion, and discomfort are the common motifs. Varying in exact details, the more general idea is that sexual intercourse with the Devil is overall unpleasant, abnormal, and negative. The larger body of the idea posits that the Devil is duplicitous and full of false promises. The negative sexual experience is a derivative of the Devil’s dealings with the seduced.

In summary, individual authorities were the immediate actors who made case interpretations and legal decisions, and they had considerable room to do so creatively. But an organized body of thought—that of “demonology” treatises—offered pre-organized, ready-to-use methods for them. With these tools, they could efficiently and elegantly synchronize different information, ideas, and authoritative voices with one another, rendering a harmonious account.

Multiple Voices and Multivocality

Another process of building external coherence is the organizing of “voices.” In choral music, different voices of males and females are organized into soprano, tenor, countertenor, and so on. In empirical inferencing, “voices” that embody emotions, social positions, interpersonal perspectives, among other things, are often borrowed.

Theorist Mikhail Bakhtin, in exploring the concepts of “voices,” articulated that “multivocality” can embody “the interaction of several consciousnesses, none of which entirely becomes an object for the other”; the voices can be aesthetically and sensationally distinct. ⁸ Borrowing from music, theorist Mikhail Bakhtin uses the term polyphony to describe how different voices could, without merging into one or strictly following one common scheme or rule (e.g., the rule of octave), complement one another in a state of plurality. Several distinct musical segments—made up of the diverse forms of monophony, homophony, and heterophony—could be creatively integrated to form a polyphony, without these forms ever needing to resemble one another.

Relating to the building of relational bridges in witch trials, each party on different sides of the bridge need not sing at the same frequency or harmonize themselves according to a common rule. Multivocal polyphonic formats (as an overarching coding format) could tolerate many epistemic styles, even minor contradictions that arise, even accommodating minor contradictions. Everyone from different social positions or institutional roles could play a part in the polyphony, exercising agency with autonomy. While disorganization is a risk—as perhaps arranged and coordinated by an invisible composer or author—a more complex and accommodating form of coherence is created.

A complete trial story, especially that of a mass trial, often involves an intricate web of voices. To help us visualize the diversity, we can borrow from Willumsen’s characterization of actors representing difference voices (see Table 4.1).⁹

Table 4.1 Polyphonic coherence between parallel voices, adapted from Willumsen’s Witches of the North^a

Each of these voices offers something unique—such as a different perspective, a different articulation, a different kind of detail—that another would not. The voice of the scientist, for example, would add a different contribution to the polyphony than the voice of a scribe. In some scenarios, these simultaneous voices may work like the voices of each individual blind person touching an elephant, thus offering an incomplete observation of a broader subject, but with far fewer contradictions. In a successfully unified account, these blind persons act more like a de facto division of labor, each offering a piece of information that complements another. Even if their articulations are different, the differences arguably enhance coherence if they are well synchronized. In the presence of noticeable epistemic gaps, playing a coherent polyphony from diverse voices helps to strengthen confidence.

More importantly, an effective polyphony is often performed by people who are representatives of different epistemic communities, with different degrees of epistemic authorities. In addition to “convergence”—or synchronization—of empirical details, then, we also have a synchronization of authorities. In fact, since their authorities build onto each other, their synchronization gives the cohere idea accumulated authority. The unity in authority also creates such a mental force that it would be difficult for unorganized individuals daring enough to question or depart from their shared assumptions.

Preemptive Design and Secondary Elaboration

Aside from parallel structuring and multivocality, two other mechanisms essential to building external coherence are preemptive design and secondary elaboration. Both of them help to minimize and contain contradictions.

By preemptive design, I mean a design that can preempt contradictions to surface within an idea system. This can be accomplished by the very design of individual propositions or a group of interwoven propositions.

Consider the Devil’s characteristics as an example. At the outbreak of witch panic in Germany, the Devil had been popularly associated with tricking women into sexual intercourse. But men too started admitting having diabolic intercourse in a German town—the Devil’s sexual properties (that is to say, the propositions) expanded further to the domains of homosexuality and bestiality gradually.¹⁰ The original design of Satan’s properties as supernatural, odd, negative, and mysterious afforded such seeming departure from the original image. By being somewhat generic and vague, rather than resolutely specific, the advantage is resilience: such designs would be “built to last,” so to speak.

Other alleged characteristics of the Devil’s power include becoming invisible, changing shape, going through walls, transforming into animals, simultaneously existing in two locations (bi-location), and casting spells that would cause long-term harm (with imprecise duration). These ideas preempt most of the normal ways by which humans empirically refute and disprove propositions using principles of natural laws. Two witnesses who testify that the accused appeared in far distant locations did not necessarily indicate probable lying, as in the case of their testifying to an ordinary case of theft or murder. Witchcraft consists of non-natural constructs—that is, it consists of supernatural constructs—for which the laws of physical sciences do not straightforwardly apply.

Overall, the many years of refinement of the concept of witchcraft helped to create a fine balance between elegance, vagueness, flexibility, and specificity in the ideas’ designs, helping them weather stormy and turbulent conditions.

Secondary elaborations (or secondary explanations) are a set of propositions or pathways of thinking developed to handle the rise of forthcoming contradictions that—almost by anticipation—would come to the surface following the initial set of propositions. ¹¹

A simple illustration is this. If a person believed to be punctual in habit (initial, core proposition) has been observed to be late to work regularly (potential counterevidence), one may seek to explain the phenomenon by heavy traffic, sickness, exceptional emergencies, after-hour work tasks, and so forth. Even if the secondary explanations are not completely satisfactory answers, they often orbit around the initial propositions like a defense mechanism, serving to deflect, diffuse, or even resolve counterevidence or contradictions that emerge—making the initial propositions become almost incorrigible.

Parameters of Acceptable Contradictions

When applied to witchcraft cases, often a puzzling situation (e.g., caused by counterevidence) emerged, and a set of explanatory ideas formed in response. These ideas were somewhat contradictory to one another, and none was sufficiently satisfying in its own right.

To resolve the amount of contradictions, these contradictory ideas sometimes formed an orbit of potential explanations revolving around certain central, core theses. The secondary ideas might have ties to the core ideas, but they were never an indispensable part of them. They could not fully “resolve” or “answer” a problem, but they have enough evidential and intellectual merit to warrant consideration. Perhaps they have enough merit to work as a hypothesis; a more developed one may even warrant the status of a “theory” or a “school of thought.” If one of these pathways of ideas is temporarily questioned or rejected, an alternative pathway of explanations could be put into work.

Within this orbit, secondary explanations operate within certain parameters of acceptable contradictions.¹² Certain internal disagreements were allowed, even quite significant ones. But these explanations must affirm the original ideas—the core propositions. Explanations that challenge the core proposition would present an unacceptable amount of contradiction. The core propositions are “defended,” in a way, by those pathways of explanations within the parameter. Considerable work and evidence may be required to undo each of them resolutely, eliminating each of them as possibilities.

Figure 4.1 depicts a scenario involving several propositions pertaining to the witches’ Sabbath.¹³ The “incorrigible” and essential propositions are like the central nucleus around which the three pathways evolve, each with its set of derivative interpretations and accumulated forms of evidence. Believers of the idea system could choose one or another to explain an issue, or simply to leave open multiple interpretations.

Fig. 4.1

An orbit with three potential explanations operating within parameters of acceptable contradictions

Explanations within the orbit share a certain symbiotic relationship with one another as well as with the nucleus. Each pathway of thought was created due to the imperfection of others (and the imperfection of the nucleus propositions). Their co-presence increases the complexity of the idea system enormously as well as the difficulties in unraveling it. To thoroughly disable all defenses, the attackers may have to review and undo different lines of arguments, concepts, and pre-coded information.

The three pathways in Fig. 4.1 represent, hypothetically, three interpretations of the witches’ Sabbaths. Suppose a court had problems actually finding the physical locations of the Sabbaths—where hundreds or even thousands of witches allegedly gathered. This problem could be explained by the theory that the Sabbaths were held in secret places, and they were so clandestine that they could not be easily found.¹⁴ And if the Sabbaths were so far away, there must be a means for so many witches to attend. But the belief that witches—partly associated with the image of “night witches” in classical literature—could transport themselves in the air through a fork or a broom, or that they could turn themselves into animals to evade villagers’ notice, explained how attending the Sabbath could happen.¹⁵ A husband of an accused could testify that he knew his wife lay beside him every night instead of attending the Sabbath. But this contradictory evidence could be countered by other elaborated claims—for example, that the accused had used magic to deceive the husband, tricking him by placing an object on the bed.¹⁶ Other theories were also used as candidates for explanation, such as the Devil’s ability to move body or spirit (or both), or that nocturnal adventures of many women were mental events caused by trickeries of the Devil.

On the surface, night flight by physical means may be “more absurd” than night flight in the mind or in spirit. But these two equally outlandish ideas provide a range of plausible possibility. Given that multiple ideas remained plausible, their collective presence allowed people to process contradictions within acceptable paradigms via different pathways; it tolerated unresolved contradictions—even transforming them into expanded possibilities. The believers might not need to resolve any specific point in particular within that zone/space. And in any given approach or theory they decide to momentarily opt for, there they could find a rich set of compact, elaborate symbols and accumulated evidence. If one disliked the relative conservative doctrines of Delrio or the Canon Episcopi, they could adopt the more aggressive stance of Remy—or the local stock of knowledge that coheres with these theorists in many aspects. The expansion of variant and possibility thus occurs within that parameter, a parameter that characterizes the epistemic community. They might even, in selectively rejecting the savage or backward approach of Kramer and Sprenger or village beliefs (e.g., the belief in the witches’ bi-location ability), signify their relative moral or epistemic superiority.

The title of Pierre de Lancre’s 1612 treatise, called On the Inconstancy of Witches (Tableau de l’inconstance des mauvais anges et demons), illustrates such an outcome of concept expansion. Written at a time when many major demonological writings were already published, Lancre often juxtaposed the differing opinions by demonologists in non-antagonistic ways, suggesting more than one possibility for the properties of the Devil. By stressing “inconstancy” as a theme—a synthetic idea—it deferred identifying the absolute, definitive manifestation of the Devil, using many possible pieces of evidence to elaborate on the ambiguity and possibility of what the Devil did or could do. The Devil had a penchant for branding witches with a Devil’s mark “just as God marks the saints” but did not have to do so; sometimes, he erased or omitted making such marks, “depending on whether or not he thinks this would be more advantageous.”¹⁷ The demons might have the power to keep witches—or anyone—afloat when they are thrown into water, but they did not always have to do it: they did it “in order to deceive the world, and in particular to abuse the judges whom the Devil amuses with this madness [of a cold water test].”¹⁸ The Devil might have considerable power to the extent permitted by God—even the power and strength to move mountains¹⁹ or to “move and transport a specific measured weight with as much speed and with the agility of the angel”,²⁰ but such power is enabled by God, is inferior to God’s, and could be constrained by God.²¹

Should the core propositions as to why God may permit the Devil to inflict harm upon the innocents be questioned, Lancre considered several pathways of explanation.²² For example, God may allow the power that witches use to kill children “in order to increase their sin and deepen their eternal punishment” (Explanation 1). God may foresee that unbaptized children will commit “enormous sins… were they to live, and because He does not want them to fall into even graver damnation” (Explanation 2). For baptized children, “Sometimes God permits such things to punish or test the parents” (Explanation 3), or “by their premature deaths they are protected from several more serious crimes already foreseen by God that the occasion and their age would have made them commit” (Explanation 4). Each of the above ideas is backed by particular stories or textual excerpts. Challenges of the core propositions could be deflected by debates regarding these explanations.

External Coherence with Competing Knowledge Systems and Authorities

At the height of the witch panics, a variety of epistemic authorities—the viewpoints of scholars, jurists, political figures, physicians, and religious authorities—developed cohering ideas on the subject of witchcraft.

As time developed, judicial caution generally increased. The paradigms of natural science helped to propel the trend by contributing to legal re-classification. Poisoning had often been linked to supernatural causes, but later poisons were increasingly recognized as physical agents, and the acts thereby became simple acts of murder. Psychological interpretations—such as melancholy and feeblemindedness—were also used to interpret cases in legal settings.²³

The mindset of the law faculty had undergone a transformation in the late seventeenth century; eventually, a majority of them openly advocated against capital sentences or torture for witch cases.²⁴ The skeptics had gradually developed their own compact symbolic structures, accumulated evidence, and gathered various other epistemic furnaces that made it harder for witch hunts to achieve dominance. As Goodare notes, new demonological treatises after 1620—shortly after Lancre’s The Inconstancy of Witches—were few and far between, even if old treatises were reprinted. Newer demonologists engaged in the topic of witch prosecution at a more philosophical rather than a practical level; their views on witchcraft merged more with medical problems, and they demanded far less punitive action against witches compared to earlier demonologists.²⁵

Among the new wave of works published by skeptics, particularly noteworthy was the rise of “anticlericalism” in the seventeenth century.²⁶ These radical writers were important because they operated outside of the normal parameters of criticism. They were willing to challenge religious authority at large. As much as Spee’s Cautio Criminalis (1631) was eloquent and critical, many critics of witchcraft, including Spee, never challenged the fundamental assumption of the existence of God and the Devil, nor did they openly challenge religious authorities. These new cadre of radical writers expanded the parameters of discourse—they did not rely on the same core propositions as their moderate counterparts. If they did not achieve a fully-fledged “paradigm shift,” they at least succeeded in making Spee and others look more tempered and moderate.

Historians had cited the rise of other ideational movements and intellectual figures that might indirectly foster the decline of witch prosecutions. This new intellectual zeitgeist included individuals such as Isaac Newton, Thomas Hobbes, René Descartes, and other significant luminaries of the Enlightenment.

Although the new intellectual “revolution” did not specifically concern witchcraft, it nonetheless affected the idea system’s efficacy indirectly. The same information encoded as “evidence” for the witchcraft idea system could alternatively be encoded by other idea systems. While the Devil and his witches might be interpreted by demonologists as being able to commit maleficent acts through natural causes, a supernatural framework was no longer the primary or domineering lens to understand or solve a problem. Witch trials would be constrained by a variety of other ideas—the ideas of scientific rigor, “mechanical philosophy,” judicial caution, medical understanding, and anticlericalism, built by the voices of polyphonic authorities—that were in direct or indirect conflict with them. This kind of challenge in external coherence inevitably affects internal coherence as well. As more and more people had the opportunity to use other methods to make sense of the same data and observed patterns, more competing thoughts were likely to have surfaced in the public domain, if not also within the minds of individuals.