Clarifying terminology

Before outlining the approach proposed in this chapter in greater detail, it may be sensible to introduce some technical terminology that will facilitate its understanding. Namely, I will discuss the distinction between first-order, second-order and third-order scholarly concepts; the so-called insider/outsider problem in the study of religion (and/or magic); and, finally, the concomitant differentiation between emic and etic scholarly analyses.

The most important theoretical distinction on which the present approach relies is the distinction between first-order, second-order and third-order concepts. The terms first-order and second-order have originally been applied in philosophical logic, where second-order logic has usually referred to an enhanced set of predicate symbols. ¹ Over the course of the past decades, these terms have been taken up by scholars of religion and often employed – yet, with a somewhat translocated meaning – in the ongoing debate about the concept of religion. The sociologist of religion James A. Beckford, for example, has claimed that religion is a “second-order concept. It is an observer’s construction that is supposedly based on the first-order beliefs, practices and experiences of human actors”. ² His main argument is that whereas social actors may or may not subsume their actions and beliefs under the term “religion” (which would then represent a first-order use of the term), scholars may nonetheless adopt a – supposedly more clearly defined – concept of religion while analysing these actors (this would then represent its second-order use). Accordingly, Beckford insists “on a clear conceptual distinction between first-order and second-order notions of religion”. ³ The idea that social actors (that is, the objects of scholarly research) use first-order language, whereas scholars use second-order language to analyse these actors (such as comparative categories or taxonomies), has become an established pattern of argumentation in the ongoing debate on the concept of religion. ⁴ Some scholars have used the distinction with a different meaning which shall not interest us here ⁵ but Russel T. McCutcheon’s addition of a third level of reflection which he calls “third-order category of redescription” ⁶ is nonetheless noteworthy (as it could be applied to the tendency of scholarly debates in the humanities to become increasingly abstract, up to the degree of being completely detached from any real-life “data” – this has often happened in the debate on magic).

In this chapter, I suggest that the distinction between first-order and second-order language should be implemented more systematically in the study of medieval learned magic. I will thus argue that the use of the term magic within medieval sources should be referred to as first-order, whereas its second-order use refers to the application of the concept of magic by modern scholars who investigate these (or other) sources. A third-order use of the concept of magic may refer to its eventual generalization and universalization, for example when scholars (such as medievalists) move beyond their particular corpus of sources and engage in interdisciplinary debates on whether magic is a human universal or not. As will be argued below in greater detail, both second-order and third-order talk on magic entail, in my view, a range of basic methodological difficulties and should be avoided on principle.

Related to the differentiation of first-order and second-order language is the so-called “insider/outsider-problem”, which has been a major point of discussion in the study of religion over the past years. ⁷ The main argument here is that believers and/or practitioners of a particular religious tradition (who may be considered “insiders”) tend to speak very differently about their tradition than adherents of other (for example “competing”) religious traditions or further “outsiders” to the tradition in question. ⁸ Seen from this perspective, scholars of religion are necessarily “outsiders” to their objects of study, even when they have acquired substantial insider knowledge (for example by reading primary sources or conducting interviews). This may also be due to their use of second-order language which is often considerably different from the first-order language used by the “insiders” of a given tradition.

With regard to medieval learned magic, this distinction is useful as it points to the necessity of distinguishing two very different medieval discourses about magic which could be called insider and outsider discourse. The first may refer to medieval ritual texts that have been written, copied or used by practitioners (or theoreticians) of the art who have, in fact, often applied the first-order concept of magic to refer to themselves or the rituals described or theorized in these texts. ⁹ The second may refer to medieval sources that have spread and advocated polemics against such insider texts or against magic in general (elsewhere, I have used the analytical terms “discourse of inclusion” for insider sources, and “discourse of exclusion” for polemical outsider literature). ¹⁰ This distinction is, by now, fairly established in medieval studies (and it also underlies the rationale of the present volume) and is applied more and more frequently in other historical contexts and disciplines, too. ¹¹ Note that both medieval insider and outsider sources are on the level of first-order language, as both make frequent use of the Latin term “magia” and/or established medieval synonyms of that term (for example “necromantia”, “nigromantia”).

The distinction between emic and etic approaches is directly related to the differentiation of first-order and second-order categories and the “insider/outsider-problem”. In an important article published in 1967, the American linguist Kenneth L. Pike derived both terms – emic and etic – from linguistics (where they originally referred to different, namely phonemic and phonetic, conceptualizations of sounds) and suggested applying these to the study of human behaviour. ¹² His main argument is that it makes a great difference whether a scholar tries to analyse human behaviour as “from inside the system” ¹³ – that is, from the perspective of the actors’ own experiences, understanding and terminology (this would then represent an emic approach) – or whether a scholar reframes such first-order data by making use of more systematic and/or comparative second-order concepts and categories (which would then correspond to the etic approach). These second-order categories may be completely alien and incomprehensible to the social actor(s) in question but may nonetheless be useful for systematic and/or comparative purposes. The differentiation between emic and etic approaches has been particularly important in anthropology, but has also been applied more and more frequently in the study of religion and various historical sciences over the past decades. Note that, with regard to medieval learned magic, an emic approach is not to be equated with the insider perspective of medieval practitioners, nor the outsider perspective of medieval polemicists: both emic and etic standpoints are an exclusive preserve of scholarly analysis. ¹⁴

Magic as a discursive concept

On the basis of these conceptual clarifications, I shall now move on to elucidating what I mean by a discourse historical approach towards medieval learned magic. To cut a long story short: the approach proposed here strives for an emic analysis and reconstruction of the medieval insider discourse of learned magic. The latter is here understood as a collection or “group of statements” ¹⁵ (to quote Foucault) that employ an etymological derivate, linguistic equivalent or culturally established synonym of magic as an identificatory first-order term of self-reference. The addendum “learned” is used to point out that the analytical focus lies on such medieval insider sources (which are here referred to as learned magic for various reasons – see below), while contemporaneous polemical (outsider) sources are mostly neglected. The approach is called “discourse historical” as it is essentially an application of the method of discourse analysis in a historical setting. ¹⁶ Accordingly, it is argued that magic here functions as a discursive concept, as the analytic focus lies on its first-order use within historical sources and discourses, and not as a second-order scholarly category.

Applying a discourse historical approach towards medieval leaned magic entails some important methodological implications that may or may not differ from the analytical use of magic in other medievalist works on the topic (including some chapters within this volume). Let me sketch out some of these implications in greater detail.

1. The approach proposed here refrains from any second-order or third-order notions of magic: that is magic is neither defined, nor theorized, nor generalized nor universalized in any essential manner. In line with other discursive approaches in the study of religion, the approach can be considered anti-essentialist, ¹⁷ as magic has no intrinsic meaning in itself but is “constructed and informed by the particular discourses that surround it in particular historical, social, and cultural contexts”. ¹⁸ Scholars have sometimes applied the term “empty signifier” in this sense, ¹⁹ but I find it more plausible to perceive magic as a floating signifier: its semantics are obviously not empty but rather floating, in the sense of being dependent on the context of its use. ²⁰ In contrast, second- or third-order (scholarly) concepts of magic are usually essentialist in the sense that they refer to an “essence” of magic that may be derived from substantial definitions, disciplinary habits or simply everyday language. In my view, such essentialist second-order notions of magic are neither necessary nor helpful for understanding the insider perspective(s) of medieval authors and practitioners of learned magic. ²¹
2. As already mentioned, the approach proposed here is oriented towards sources that include an etymological derivate, linguistic equivalent or culturally established synonym of magic as an identificatory first-order term of self-reference; this is in fact its main criterion for setting up a corpus of medieval learned magic. ²² I would like to stress that this criterion is neither arbitrary nor trivial: over large parts of Western history, it has obviously involved serious personal risks to consciously denote one’s own ritual practices and texts as magic, when at the same time powerful elitist discourses have devalued and condemned magic or even initiated legislative action. This is even more evident with regard to late medieval author-magicians. ²³ What is more, the criterion is empirically well-documented. Not all medieval texts that seem to belong to the medieval insider discourse include the Latin term “magia” or one of its synonyms (partly because their authors may have avoided the term on purpose), but many of them do, so we are indeed able to reconstruct a fairly coherent discourse of inclusion in the European Middle Ages. ²⁴ Apart from this purely philological criterion, there appear to be numerous family resemblances and further stereotypic narrative and/or ritual patterns across the texts that justify allotting them within the same category (consider the use of common pseud-epigraphs such as Solomon, Hermes or Apollonius; the recurrent narratives of a holy art or angelic transmission; the adoption of “voces magicae”, “charaktêres”, talismans, sigils, ritual circles and other sophisticated ritual techniques; and so on). It may be useful to continue exemplifying and fine-tuning these family resemblances in future research as they justify the scholarly configuration and investigation of a fairly clear-cut textual corpus of medieval learned magic without the need to define magic as a second-order category. ²⁵
3. The analytical focus on medieval learned magic calls for a pronounced differentiation of medieval insider and outsider discourse. Of course, there has always been a dialectical interplay between these two discourses in (and beyond) the European Middle Ages. ²⁶ We come across sceptical but curious encounters with learned magic, enthusiastic discussions – albeit without practical experience – or nuanced, ambivalent judgements that seem to represent rather intermediary positions. Medieval examples of such positions may be the discussion of the concept of “magia naturalis” by medieval theologians ²⁷ or surveys of relevant literature by authors such as (Ps.?) Albertus Magnus (Speculum Astronomiae), ²⁸ Michael Scotus (Liber Introductorius), ²⁹ Berengarius Ganellus (Summa Sacre Magice) ³⁰ or Johannes Trithemius (Antipalus Maleficiorum). ³¹ However, as the conceptual history of magic was – particularly in premodern times – extremely controversial, morally and religiously value-laden, full of social stereotypes and often a matter of legislation, the insider/outsider poles are fairly clearly marked, so that most medieval actors and sources will be assignable to one of either of these two poles (this does not preclude that some authors may have partaken in or adopted arguments from both discourses). From a discourse analytical perspective, this differentiation is crucial because the first-order concept of magic is employed not only in very different ways (for example with a valorizing or polemical function) but also while referring to very different things in medieval insider and outsider sources ³² (according to Foucault, discourses are “practices that systematically form the objects of which they speak”). ³³ These differences call for analysing medieval insider sources as independently as possible from the misleading claims and distortive narratives that pervade medieval polemics against magic. Certainly, outsider texts can be useful, for example when they provide lists of insider texts that were circulating at a given time (for example Speculum Astronomiae). However, as soon as one wishes to know more about the contents of such texts, outsider sources should be read with great caution as they tend to simplify, distort or – consciously or not – misunderstand the insider tradition of medieval learned magic. A telling example is Thomas Aquinas’s argument that the angels invoked during the Ars notoria ritual (I am now referring to the first-order terminology used within the text, that is to the apparent insiders’ perspective) may really be demons (Summa Theologiae 2, 2, 96, 1). ³⁴ From the viewpoint of the modern study of religion, the “demon pact” narrative – which has informed Christian polemics against magic ever since Augustine ³⁵ – is obviously nothing more than a tool of religious “othering” ³⁶ : it neither leads to, nor is it interested in, a proper understanding of the “other” (in this case the medieval learned magician), but rather distances the latter by projecting a stereotypic, distortive pattern of interpretation onto him. Given this basic tendency, one might generalize that outsider accounts are usually not trustworthy for reconstructing the insider perspectives of authors and practitioners of learned magic (this is the case also in other epochs).
4. The approach proposed here also adopts an anti-essentialist stance towards medieval insider sources. In fact, a comparative reading of these sources quickly reveals that medieval learned magic is not a homogenous category: there is no conceptual “core” or ritual “essence” that can be deduced from the sources apart from the fact that the first-order concept of magic usually refers to a “ritual art” (as most insider texts are ritual texts or theoretical reflections on such texts; the corresponding first-order formulation is thus “ars magica”). ³⁷ Even though medieval insider narratives tend to suggest otherwise, this art is strikingly heterogeneous, hybrid and ever-changing from an analytical perspective. It is heterogeneous, as the ritual procedures described in most medieval insider sources consist of a vast variety of different ritual (micro-) techniques and varying concepts of ritual efficacy. It is hybrid, as these techniques are usually combined in the manner of “building-blocks” in a multiplicity of ways, depending on the preferences of the respective author or copyist (this is what I refer to as “ritual hybridity” elsewhere) ³⁸ ; what is more, these techniques have often been derived from different cultural or religious contexts (this is what I refer to as “religious hybridity” elsewhere). ³⁹ It is ever-changing, because learned magic is – within and, the more so, beyond the European Middles Ages – in permanent motion: it continuously adopts ritual patterns and techniques from older sources, it discards unnecessary or unwanted elements, it adapts to novel cultural and religious environments or practitioner milieus and it continuously invents novel modes of ritual performance or efficacy. All three features call for non-essentialist analyses of medieval insider sources and nuanced modes of analytical description. In contrast to adopting essentialist second-order definitions of magic that may obscure or completely bypass the heterogeneity, hybridity and changeability of medieval learned magic, I suggest adopting a typological perspective, maybe inspired by the concept of “family resemblances” (as suggested above) or the recently proposed concept of “patterns of magicity”. ⁴⁰ The idea would be to develop open and flexible taxonomies of ritual techniques, ritual goals and concepts of ritual efficacy (and/or other features) that may be consecutively derived from medieval insider sources, but also applied to these for comparative purposes and for the reconstruction of intertextual dependencies or ritual dynamics.
5. I speak of medieval learned magic – and not of magic in general – in this chapter for two reasons. First, the addendum “learned” refers to two fairly distinctive characteristics of medieval insider sources: (i) they stem from people who were not only able to read and write (already a tiny elite in the European Middle Ages), but often quite sophisticated authors with apparent expertise in several languages and religious traditions; and (ii) the “ritual art” described in these sources tends towards complex, time- and resource-consuming ritual performances. The addendum “learned” thus operates as a marker of specificity of this particular corpus of sources and thereby helps to demarcate it from other (allegedly magical) ritual traditions that may have been transmitted only orally and whose ritual performances may have remained rather short and simplistic (consider so-called medieval folk magic traditions). ⁴¹ Second, the addendum “learned” points to the fact that medieval insider sources are not unique but the result of a complex interplay between intercultural transmission ⁴² and inner-cultural appropriation. ⁴³ Seen from this entangled perspective, medieval learned magic is obviously part and parcel of a much larger textual–ritual tradition that is considerably older (as it goes back at least to late antiquity) and continues up to this day: “Western learned magic”. ⁴⁴ The historical embeddedness of medieval insider sources within the overall history of Western learned magic calls for interpreting these sources not (only) by reference to medieval polemics (that is, contemporaneous outsider sources), but (also) by reference to other – that is, preceding and/or subsequent – insider sources.
6. The last argument ties in with a recently proposed research programme on “Historicising Western learned magic” which consists of eight theoretical issues that should, in my view, be considered in the course of its historicization: continuity, changeability, hybridity, deviance, morality, complexity, efficacy and multiplicity. ⁴⁵ It is, of course, impossible to go through all these issues in the final section of this chapter. Apart from word-count restrictions, such a discussion would require a coherent, diachronic and cross-cultural narrative on the history of Western learned magic at hand in order to systematically relate medieval insider sources to preceding (for example late ancient, medieval Jewish, Islamic or Byzantine) and subsequent (for example early modern, modern or even contemporary) insider sources. For the time being, such a work is still a scholarly desideratum. ⁴⁶ However, even on the current state of research, the analysis of medieval insider sources from the viewpoint of the overall history of Western learned magic may reveal interesting insights and thus serve as an important complement to their (so far, prevailing) interpretation and contextualization within medieval studies.

Note that such an analysis poses different questions to medieval insider sources than medievalist in-depth studies or critical text editions. For example, it puts greater emphasis on the changeability of learned magic texts and techniques from a diachronic and cross-cultural perspective (thereby tying on novel approaches towards ritual dynamics). ⁴⁷ If one combines all eight theoretical notions mentioned above – continuity, changeability, hybridity, deviance, morality, complexity, efficacy and multiplicity – and projects these onto medieval insider sources, one might even come up with a bold research hypothesis: the European Middle Ages may have operated as some sort of “bottleneck” within the overall history of Western learned magic. In fact, medieval insider sources seem to display a more or less significant decrease on a range of domains, compared to antecedent and subsequent insider sources: regarding (i) the quantity of circulating insider texts, (ii) their conceptual complexity (this refers to the existence of elaborate insider definitions and systematizations), (iii) their ritual complexity (this refers to the length and complexity of the “ritual art” outlined in insider sources), (iv) their social evaluation (this refers to the quantity and quality of liberal milieus where learned magic may have thrived for the time being) and (v) their position towards morality (greater restriction in this matter may lead to rejecting malevolent ritual goals, for example). ⁴⁸

For the time being, this is nothing but an ambitious hypothesis that may provide food for thought and eventually point to future avenues of research. Regarding the latter, it might call for enhanced and more systematic modes of cooperation among those historical disciplines that may be considered relevant to historicizing Western learned magic – such as classical studies, medieval studies, Arabic, Jewish and Byzantine studies, early modern history or the study of Western and contemporary esotericism. Like other contemporary approaches in historiography (consider entangled history, transcultural history or global/world history), ⁴⁹ historicizing Western learned magic challenges the plausibility of investigating a limited set of insider sources within the boundaries of single historical disciplines. Instead, it calls for engaging in interdisciplinary, diachronic and cross-cultural analyses – and, eventually, large, communal research projects – that do justice to the complexity of this novel and fascinating field of research. The present volume is a laudable step in this direction.