Among the largest surviving collections of medieval court records are the documents from the inquisition into heretical depravity (inquisitio heretice pravitatis) conducted between May 1245 and August 1246 at Toulouse by the Dominicans Bernard de Caux and Johan de Saint-Pierre. During this ‘great inquisition’1 – the most expansive known to historians – over 5,500 lords, ladies, doctors, tanners, scribes, charcoal burners, and cow-herders were summoned from across the Lauragais (the verdant agricultural region between Toulouse and Carcassonne) to the cloisters at Saint-Sernin to depose about themselves and others, living or dead, regarding the crimes of heresy (requisitus de veritate dicendo de se et de aliis, vivis et mortuis, super crimine heresis et valdesie).

The evidence for this immense inquisition has come down to us in a registry contained in manuscript 609 of the Bibliothèque de Toulouse (MS 609).2 MS 609 entered professional historiography3 as part of the great wave of documentation of medieval records by nineteenth-century French ‘positivist’ historians (Carbonell, 1978: 173–85) when Charles Molinier (1880: 173) described it as the registry of a ‘large inquisition which spanned an entire region’;4 Molinier’s opinion has been confirmed by generations of historians of heresy and inquisition as a ‘general inquest of the state of religion in the whole diocese’ (Wakefield, 1974: 173–74), a judgement based on the human and geographic scope of the registry (as well as the great number of denials it contains, as one might expect from general survey).5 By way of demonstration, the 45 folios of depositions from the inhabitants of the first two village parishes in the manuscript alone (and the subjects of this essay: Mas-Saintes-Puelles and Saint-Martin-Lalande) contain 668 deponents testifying 685 times about 1,356 distinct persons and 236 places, in recollections going back as far as the 1180s.6 The registry contains depositions from 106 villages.

When Molinier issued the first report on MS 609, the power of the prevailing theory of the still-new theory of Catharism7 was already such that, despite the sheer scale of information contained in the registry, it was relegated to a supporting role in conveying the history of heresy in the Midi. Why this should be merits investigation beyond the scope of this essay (Pegg, 2015: 225–79; Pegg, 2016: 21–52), however there are several factors which bear directly on this study. It was Charles Schmidt who constructed Catharism in his two volume Histoire et doctrine de la secte des Cathares ou Albigeois (1849), and he noted with self-awareness his innovative ‘contribution’ of inquisition depositions to the historiography of medieval heresy. Schmidt did not know of or consult MS 609 nor any other documents in the south of France, but rather worked from the document extracts given to him by the curator of the Doat Collection at the Bibliothèque nationale de France in Paris (Schmidt, 1848: vii–viii, note 1). The Doat Collection – copies of selected inquisition documents made in the seventeenth century for the political operative Jean-Baptiste Colbert – has since remained central to defining the relationship of inquisitorial registries to Catharism due to a series of historiographic coincidences, including the easy paleography of the large script (transcribed from medieval originals into classical Latin) and early circulation from its central location of Paris.8 For Schmidt, the Doat documents reflected not only the heretical Cathar theology he believed was found in the Midi, but also the sociology undergirding the Cathar sect. The particular Doat depositions relied upon by Schmidt feature a striking focus on ‘heretical sympathies’ of nobility in the lands of the Count of Toulouse (including the Lauragais), one which he found adhered to the anti-noble medieval polemics and chronicles from which he argues his Cathar sect and theology.9 The central role of nobility in fostering Cathar heresy was for Schmidt the defining social feature of Catharism, and has been carried forward in historians’ use of early inquisitorial registries from the Midi.

Mark Gregory Pegg (2001b, 2015, 2016) has demonstrated the continuity of research methodology from Schmidt to the current generation of scholars who hold to Catharism and how it deforms our investigation of the content of inquisitional records.10 He has labelled (Pegg, 2001b: 191) one particular methodological problem as the ‘strip-mining’ of inquisitorial registries: a method based on raiding the inquisitorial registries for folio references to match the a priori expectations of Catharism, from theology to sociology, regardless of the consequences such fragmentation has on obscuring the reality of the inquisition, its subjects, and any heresis inquisitors found in the lands of the Count of Toulouse. In his monograph, the only sustained study of the contents of MS 609, Pegg inverted the logic of strip-mining, using a historio-anthropological method to understand the subjects’ beliefs (and more) from the stories contained in the depositions of MS 609.11

The fact there has only ever been one monograph published on the contents of MS 609 in the 120 years since Molinier’s catalogue of it – objectively an extraordinary document containing details of the lives of over 5,500 people who lived over 750 years ago – is surprising. There is of course the fact that it has never been edited to make it widely available for study, despite a variety of attempts.12 This may be due to the manuscript’s reputation as a massive, difficult volume brimming with formulaic depositions, judged by historians since as ‘dry, terse and repetitive’.13 It may also result from certain early uses of MS 609 which conveyed to historians its limited value as a source. The earliest inventory of MS 609 was a fifteen-page summary by the medievalist abbot and bishop Célestin Douais (1891), which he began by remarking that the manuscript ‘ne s’est encore trouvé personne pour le dépouiller’ (dépouiller: go over, study, dissect, strip, denude) before concerning himself with summarizing the contents according to categories oriented at theological matters.14 Douais’ article was followed by the work of his student (and editor of the Catholic journal La Croix) Jean Guiraud: his edition of the Cartulaire de Notre-Dame de Prouille (1907) and opus Histoire de l’Inquisition au Moyen âge (1935) were the first monographs to draw significantly from MS 609. The latter is famous as the first synthesis of Catharism since Schmidt (Dossat: 1967; Amargier and Fortanier, 1979: 217–18), and for providing the rubric for describing Catharism, followed by professional and public historians of Catharism since, including: compiling exhaustive lists of Cathar ‘perfects’, ‘deacons’, and ‘bishops’ mined chiefly from MS 609 which reify a Cathar Church into the depositions;15 citing a flurry of citations from MS 609 which exemplify and reflect a priori Catharist conceptions, while ignoring the multitudinous contradictions from depositions themselves;16 and a guiding hypothesis, still followed by many Cathar scholars today, that the fons and origo of Cathars in the Midi was Southern noble anti-clericalism, and therefore the focus on seigneurial depositions is of prime importance.17

Douais’ summary and Guiraud’s lists were the sum of knowledge of the contents of the registry that passed into the hands of arguably the most influential proponents of Catharism in the Midi: the cleric-historians of Toulouse who created and lead the Cahiers de Fanjeaux for decades (Biget, 2016), and the savants locaux (public historians) lead by Jean Duvernoy. Despite the immense productivity of this generation, the work reified Catharism as an historical object, and MS 609 as a source for citation, but not study in and of itself.18

It was only over eighty years after Douais published his inventory that the first articles proceeding primarily from MS 609 emerged, ostensibly reflecting the historiographic trends of quantitative analysis (Abels and Harrison, 1979; Semkov, 1984) and microhistory (Cabirou, 1995b; Roquebert, 1995; Wakefield, 1983). Three of these focused exclusively on the depositions from the village of Mas-Saintes-Puelles, the most fully represented village in MS 609, and, as it happens, the first in the manuscript. The most influential of these articles, Walter Wakefield’s ‘Inquisitors and Heretics: The Case of le Mas-Saintes-Puelles’ (1983), demonstrates the historiographic power of Catharism to sway the objective reporting of the contents of MS 609 through biased sampling of source material.19 The most pronounced example is Wakefield’s declaration that ‘social distinctions are not easy to detect’ in the manuscript, but he in fact passes over them as they do not adhere to his theme. With his keen eye for social history, he picks out a rare mention of the village consuls of Mas-Saintes-Puelles from a contemporary royal oath (juramentum), noting their existence with a single sentence but without either referring to or researching the six consuls within the depositions (a significant oversight, as we shall see). But, following the above-noted tendency of Cathar historiography to prioritise nobility, Wakefield then returns to the theme with longer speculation about possible co-seigneurs of Mas-Saintes-Puelles that he found in a ‘mimeographed booklet, locally produced’ in 1969 at the village, a claim for which there is no evidence in MS 609 nor any other source extant. Wakefield thus proceeds to render the story of the villagers who appeared before Bernard de Caux and Johan de Saint-Pierre as the intricate doings of the seigneurial del Manso clan and heretics, and then other villagers. Other studies present sample bias in different ways: some have focused entirely on village seignieurie deeply attached to Cathar heresy as a family affair over generations at le Mas (Roquebert, 1995) and Saint-Martin (Cabirou 1995b); still others purport a prosopography (Semkov, 1984) of heresy at le Mas,20 or a broader registry-wide demography of women and their participation in heresy (Abels and Harrison, 1979), but without going outside the depositions to corroborate it.

My own curiosity about the contents of the registry was motivated by Pegg’s arguments and lead eventually to preparing a digital edition of it. While this article does not directly explore Pegg’s thesis nor the nature of any heresis inquisitors searched for in the Midi, it does owe its findings to Pegg’s caution to treat the inquisition based on what was recorded by inquisitors themselves, and not on medieval polemics and tracts. But where Pegg took an anthropological route to understand the lives and beliefs of the subjects of the inquisition of 1245–56 and how the first generation of inquisitors shaped that into heresy, I have been absorbed by indications of whom Bernard de Caux and Johan de Saint-Pierre aimed their inquisition at. What I discovered during the act of creating a native digital edition21 – reading, transcribing, encoding, tagging, and connecting depositions – was that these Dominicans were emphatically not conducting a general survey of the faith; rather, the friars Bernard and Johan collected evidence to incriminate a particular social group in the villages of the Lauragais: the village consuls. These consuls were not united by any special relationship to whatever heresis inquisitors believed was rife in the Midi, nor were they greater or lesser practitioners of any such heresis. Moreover, the friars did this while consciously ignoring and bypassing numerous opportunities for prosecuting others for the crime of heresy.

This argument is presented below in two parts, through a study of the two most represented villages in both MS 609 and the manuscript’s own historiography, Mas-Saintes-Puelles and Saint-Martin-Lalande (from hereonin respectively referred to as ‘le Mas’ and ‘Saint-Martin’). A large-scale, comparative study of two villages will, I hope, allow us to escape the aforementioned problem of low sampling. The first part of the study presents a macroanalysis of the depositions, a technique facilitated by my edition of MS 609.22 This macroanalysis looks for patterns of accusations across all depositions using network analysis, a method that I illustrate through comparative anecdotes for scholars unfamiliar with digital methods (but does not argue from anecdotes). This part of the study produces a list of the inquisitors’ targets in each village, striking in their identical socio-political constitution as consular families.

Some commentators might be inclined to suggest that this socio-political group was attracted to, susceptible to, or otherwise friendly to, heresy, effectively displacing the historiographic blame for the Cathar heresy from nobility to consuls.23 Others might suggest that the heresy accusations arose from internecine conflicts within the village.24 The laconic depositions of MS 609 do not allow us much direct access to inner motivations of the deponents, but the second part of the study proposes several tests for these problems through a close reading (against inquisitorial formularies), tests which reveal the specific construction of incriminations in the depositions. These constructions turn out to be widespread patterns in the depositions, yet are never discussed in historiography. In these patterns we can clearly see the inquisitor’s hand selecting evidence and testimony to target the consuls and their families, while ignoring others who objectively should have been equally incriminated. In this we find the answer to Wakefield’s (1983: 221–22) open question about why Bernard de Caux and Johan de Saint-Pierre did not press on every lead they were given. It was not that they didn’t have time, as some have argued (Biller, 2010: 107; Bruschi, 2009: 16), it was that they recorded precisely what they needed.

The digital edition of MS 609 – a foundation for a new research methodology

The registry contained in MS 609 is not filled with the verbose records, oft-studied in undergraduate courses, which have come down to us from the inquisition of villagers of Montaillou in the early fourteenth century (Duvernoy, 1965; Ladurie, 1975) or Menocchio the miller two centuries later (Ginzburg, 1980). MS 609 is famously formulaic, reducing most testimony to a handful of details that the first generation of inquisitors in the lands of the Count of Toulouse thought important.25 Take for example the deposition of Guilhem Ayalric junior of Saint-Martin:26

Item. Year and day as above. The sworn witness Guilhem Ayalric junior said that he saw, in the home of Ysarn de Gibel, Bertrand Marti and his companion, heretics. He saw there with them Bernard Alzeu and Bernard Martin. The witness and everyone else adored the aforesaid heretics there, bending three times at the knee saying, ‘Bless us good men, pray to God for us’. This was about 10 years ago. And afterwards the aforesaid Bernard Alzeu and Bernard Martin asked the witness to accompany the aforesaid heretics outside the village, but he would not do so.

He did not believe the heretics to be good nor to have good faith – even though he once adored them at the insistence of the above-mentioned Bernard Alzeu and Bernard Martin, nor heard them speak errors of visible things – that God did not make them, nor the sacraments of the Church. He has not seen heretics anywhere else, nor believed [in them], nor gave or sent [them] anything, nor listened to their preaching.

Item. He said that he was with Bernard Alzeu and with many others at the Abbey of Saint-Papoul when they wanted to free two heretics held by the abbot of Saint-Papoul. This was about 4 years ago.

Whatever Guilhem Ayalric might have said to inquisitors at Saint-Sernin, it was transmuted into a rigid structure describing specific, discrete events in the third person. This first generation of inquisitors codified who, where, what, and when, each within a restricted set of categories (Arnold, 2001: 19–47; Pegg, 2003). The digital edition, beyond including the usual philological apparatus and historical notes, encodes this information as ‘meta-data’, linking it to a database of similar instances. In this way the digital edition provides technical methods to traverse, search, aggregate, and compare data according to the inquisitors’ own methods and categories. The same occurs with the summary of ‘beliefs’ found at the end of depositions: a checklist of the deponents’ attitudes towards the heretici,27 what they heard heretici say, and their belief about those ideas – always restricted to specific sacraments (de matrimonio, de baptismo, de resurrectione, de eucharistia) and whether God made material things (de visibilibus).28

The net effect for the researcher is that rather than selective sampling driven by historiographic biases, one can query across an entire corpus for large-scale patterns at a macro-level using a heuristic that mirrors the inquisitors’ own categories of guilt. From the 685 depositions from le Mas and Saint-Martin, featuring over 1,350 people and 1,281 ‘stories’, we can begin to draw out who was named by whom, and for what. The object is to test for any ‘directionality’ of incriminations, and if there is directionality, who its ‘targets’ were.

A problem of (non-)reciprocal incrimination

It is now axiomatic that the villages of the Lauragais, as with all villages across the Midi in the thirteenth century, were communities of people who lived in close proximity regardless of social position (Cheyette, 2001: 133–43; Pegg, 2001a: 63–73). This axiom has been derived in part from the above-mentioned studies of MS 609, wherein thousands of villagers of whatever ‘class’ in each village demonstrated that they knew and interacted with each other frequently. Yet when depositions of people who should have mentioned each other are scrutinized side by side, the expected display of reciprocal relationships are not evident. To demonstrate the problem, I want to return to the deposition of Guilhem Ayalric junior and compare it to the depositions of Ermengarde Boer of le Mas.

Guilhem Ayalric gave a typically short deposition consisting of two events. The first cites Bernard Alzeu and Bernard Martin, taking place at the domus of Ysarn de Gibel. We do not have depositions of any of these three incriminated by Guilhem, so we cannot test if they would have mentioned him in return in their hypothetical depositions. However, Guilhem did add a second event to his confession after the recitation of beliefs (recitations of beliefs are virtually always found at the end of a deposition, suggesting that this addition was provoked by an inquisitor’s direct question). This same event was mentioned by eight other deponents at Toulouse, and concerns an ‘armed expedition of village men’ from Saint-Martin (and Issel and Labécède) to ‘liberate’ two heretici held by the abbot at Saint-Papoul in 1241. Guilhem only mentions ‘Bernardo Alzeu with many others’, however when we compile all the depositions from Saint-Martin, we obtain a list of 23 people from the village named as having gone on the expedition (Table 1).

Table 1

People named on the expedition to Saint-Papoul c. 1241, the number of times they are mentioned in depositions about the expedition, and whether their own deposition at Toulouse is extant.

Name No. of Mentions Deposition Extant?

Arnald Fornier 4 no
Arnald la Vauza 2 no
Arnald Valh 4 no
Bernard Alzeu 9 no – deceased
Bernard Folc 1 yes
Bernard Martin senior 1 no
Esteve Faure 5 yes
Guilhem Arnald 2 yes
Guilhem Ayalric junior 1 yes
Jordan Hugole 4 yes
Peire Aio 1 no
Peire de Canast 1 no
Peire de Comas senior 2 no
Peire Faure de Collo 1 no
Peire Faure 4 yes
Peire Folc 1 no
Peire Galhard 6 yes
Pons Traver 2 no
P. Azals 1 no
Raimund de Verazilh 3 yes
Raimund Folquet 3 yes
Raimund Fornier 5 no
Ysarn de Gibel 5 no
‘pluries alios’ 5

Of those 23 men, nine of their depositions have come down to us. Guilhem was only implicated by himself, as was Bernard Folc. Several others are only implicated just once or twice, even where they themselves implicated many others. Evidently other deponents were able to remember who among their fellow villagers journeyed with them to Saint-Papoul: Bernard Folc, Peire Faure, Guilhem Arnald, and Jordan Hugole, each implicated over 10 other people by name; that Guilhem and others are recorded as mentioning only one or two people thus stands out in comparison. Even more provocative is the fact that inquisitors repeatedly settled for ‘with many others’ (cum pluribus aliis), and that none of the depositions from Saint-Martin mention that this was a joint expedition with men from Issel and Labécède. In short, Guilhem appears in his own deposition, implicates others, and disappears.

If the problems of Guilhem’s incrimination of others and his disappearance can perhaps be explained away – and historians do so by arguing that deponents commonly conspire and lie (Abels and Harrison, 1979: 221–22; Given, 2001: 125–27), or that inquisitors didn’t have enough time to pursue all leads (Biller, 2010: 107; Bruschi, 2009: 16) – the depositions of Ermengarde Boer of le Mas are resistant to such explanations. Ermengarde was deposed two times, the first on 19 May 1245.29 Inquisitors noted in the margin of the first deposition that she was ‘suspected and could say more’ (suspecta est ista et possent multa dicere); she had not yet returned to le Mas after the deposition, as she gave a second deposition three days later.30 Her first deposition recounts a long life involved in heresis at le Mas: over thirty years previously she was ‘hereticated’ by Ysarn de Castres (fuit hereticata) and then later reconciled to the Church by Saint Dominic. It obviously did not have the required effect, as she recounts a number of interactions with heretici and many denizens of le Mas in the thirty years thereafter. Table 2 illustrates the dramatic difference in incriminations between the first and second depositions. In her first deposition, Ermengarde mentions a variety of people from le Mas, although only one is known to historiography: Saramunda, wife of the eldest of the co-seigneurial del Mas brothers, Bernard senior. However, Ermengarde’s second deposition identifies thirteen members of the seigneurial family and their in-laws (the de Quiriès) as well as members of the ‘great families’ of le Mas, all of whom she evidently socialized with.

Table 2

People incriminated by Ermengarde Boer, grouped by each deposition.

Deposition May 19 Deposition May 22

Arnald Maiestre Aribert del Mas
Companh Gaufred Bernard del Mas senior
Guilhem de Canast-Bru Bernard de Quiriès
Guilhem Vidal Flor del Mas
Raimunda Gondaubou Galhard del Mas
Raimund Gasc Garnier Senior
Riqua German Guilhelma Meta née del Mas
Saramunda del Mas Guilhem del Mas junior
Segura Vidal Guilhem del Mas senior
Guilhem Moreta
Jordan del Mas
Maria de Quiriès
Finas Moreta
Peire Gauta junior
Peire Gauta senior
Roger Sartre
Saramunda del Mas
‘pluries alios’

Almost all those cited by Ermengarde gave their own depositions in turn at Toulouse, and often they are among the longest depositions in MS 609. Yet not one named Ermengarde herself. These same people named 117 other people from le Mas, chief among them the same seigneurial and ‘great families’ of the village.31 Despite her evident long history with heretici and hereticae of le Mas, despite having been made a ‘good little girl’ (Pegg, 2001: 92–103) with several other important women of le Mas, and despite being married to a man who was equally sociable with the ‘great families’ of the parish, Peire Boer,32 she was only briefly mentioned by three other people, none of whom she herself had named: Peire Roger,33 Peirona Riqua,34 and Raimunda Bernarda Gasc.35

The depositions of Guilhem Ayalric and Ermengarde Boer give us two slightly different cases with which to formulate a testable problem: there is an unexpected low rate of ‘reciprocal incrimination’ among certain deponents. At the same time, we can begin to see from the small numbers presented above that others may be implicated more often. This ‘directionality’ of accusations should not surprise us, as historiography has claimed that heretical activity was centered on, and supported by, the nobility. This directionality can be tested against large databases using simple methods for investigating large networks of people.

Macroanalysis – networks of accusations

Social scientists often investigate the structure of relations between people and their communities using network diagrams (graphs).36 Graphs are flexible analytical tools as they only require two sets of objects: nodes and edges. A node can be a person, an object, an institution, or an event. An edge (tie) describes the relationship between the nodes: acts of communication, exchanges, or transmission, or bonds of acquaintances, friends, and family. Edges are classed into two types which then give their names to the two network graphs: undirected and directed (Figure 1).

Figure 1 

Two different ways to connect nodes (people) in a network graph.

An undirected graph tells us which people are connected to others and in what concentration (degree). These graphs are concerned only with knowing which nodes are tied together, without determining if those ties are equal. For undirected graphs, the only goal is to say whether the nodes A and B are connected; it doesn’t matter if A contacts B or vice versa, or if there is an imbalance in the relationship. This approach is usually how sociologists create basic diagrams of ‘who knows who’, with the assumption that the given social relations are reciprocal (symmetrical relationships).37

A directed graph on the other hand expresses the nature of relationships between nodes which may or may not be equal (asymmetrical relationships). We can portray the quality of the relationship through direction of ties and possibly their frequency or intensity of ties. For example, A may call B more frequently than B calls A. In terms of networks of accusations in MS 609, Ermengarde Boer incriminates more people than incriminate her. A graph of Ermengarde with all edges and nodes connected to her would look like the simple Figure 2. As we begin to add more people who in turn incriminate others and are themselves incriminated, the graph grows more complicated. By applying visualization strategies carefully, we can begin to see larger-scale patterns emerge. This expansion requires setting rules for how to calculate the edges and the nodes from the digital edition.

Figure 2 

Directed graph of Ermengarde Boer of Mas-Saintes-Puelles – those she incriminated and those who incriminated her. Thicker edges (lines) indicate the weight of multiple incriminations.

In the digital edition, all people are tagged by their role and individual activity in an event as recorded by the inquisitors in the registry. The roles are categories (owner, infirmus, participant), explained here in turn:

  • participant is someone who interacts with heretici, including adoring them (adoravit), leading them (duxit), listening to their preaching (predicationem eorum audivit), believing them (credidit), etc.;
  • infirmus is someone unwell or dying and attended by heretici;
  • owner is someone who is named as the owner of a location in an event (for example in domo Guillelmi Vitalis);

Everyone at an event is a ‘participant’ of sorts, but inquisitors themselves distinguished these roles in MS 609. Moreover, they are not mutually exclusive tags in the edition. For example, an event that took place in domo Guillelmi Vitalis will state explicitly whether he was there or not as well.38 If he was named at that event, the edition tags him twice: once as owner, once a participant. If the home where the event took place was owned by the infirmus, they would be tagged as both owner and infirmus. The researcher can then decide how to sort and filter on the basis of these tags.39 A fourth role, heretici, is a distinct category, but is excluded from these graphs as no deposition was recorded from someone who is otherwise labelled as hereticus in MS 609, and frequently they were even left unnamed in other depositions (…et socium suum, hereticos).

Every time a deponent incriminates someone, which is recorded by inquisitors in these roles, a directed edge is created between the deponent’s node (source) and the node of the named person (target). An unweighted schema creates one edge per source-target, whereas weighted schema increases the value of the edge for each instance found in the registry. Guilhem Ayalric’s testimony can thus be represented by two different formulas, shown in Table 3. The unweighted edge allows us to count how many discrete individuals implicated a specific person; the weighted edge allows us to count what the inquisitors recorded as ‘infractions’. In this case, Guilhem claimed that Alzeu adored heretici, forced Guilhem to adore them, and Bernard attended the march on Saint-Papoul. Because this is a directed graph, we describe the directionality as follows: Guilhem Ayalric has a weighted out-degree of 6 (he makes six incriminations), Bernard Alzeu has a weighted in-degree of three (he is incriminated three times), Bernard Martin in-degree = two (he is incriminated twice) and Ysarn de Gibel in-degree = one (he is incriminated once).

Table 3

Calculating weighted and unweighted edges from the deposition of Guilhem Ayalric.

Source Target Unweighted Edge Weighted Edge

Guilhem Ayalric Bernard Alzeu 1 3
Guilhem Ayalric Bernard Martin 1 2
Guilhem Ayalric Ysarn de Gibel 1 1

In summary, the heuristic for identifying a pattern of incriminations in MS 609 is straightforward: collect each instance where a deponent incriminates someone else under the inquisitors’ own rubric as recorded in MS 609. We count incriminations by their degrees within each of the two villages, le Mas and Saint-Martin, to quantify a ‘directionality’ of accusations. This data is passed to a network analysis application to visualize patterns across thousands of records. The same application also provides a variety of common calculations and reports for quantifying the ‘directionality’ in our graphs of depositions.40 The dynamics of incriminations within the registries of both le Mas and Saint-Martin are demonstrated by the graphics in Figure 3 and Figure 4.41

Figures 3–4 

Graphs of networks of accusations from le Mas-Saintes-Puelles (left) and Saint-Martin-Lalande (right). Node colour represents weighted in-degree (green to red = less to more). Node size represents Eigenvector centrality (larger = more important). High resolution images and all data can be downloaded from http://medieval-inquisition.huma-num.fr/downloads.

Macroanalysis – strong centralisation of incriminations

The graphs of le Mas and Saint-Martin unequivocally demonstrate three clear patterns. The first is that the absolute number of people implicated in heresis by inquisitors in the two villages is far greater than historians have previously calculated.42 Within this larger-than-anticipated demographic, a small, single subgroup of people are strongly incriminated. Moreover, these incriminations come from across the entirety of each respective village, and not just within the subgroup itself. Finally, although some seigneurial family members do feature in the subgroup, their position is demonstrably less pronounced than historiography would lead us to expect: the subgroups are overwhelmingly populated with non-seigneurial villagers. These three patterns are analysed below in turn.

A truism of network analysis is that one should expect to find hubs, groups, and cliques in any diagram of a significantly-sized social group like the villages of le Mas or Saint-Martin.43 This is not an effect limited to large groups – schools and classes, work places, and neighbourhoods also demonstrate these effects consistently. The first remarkable feature of both village graphs is that they do not contain any detectable subgroups. Even under the pressure of a force vector algorithm which generated the diagrams – an algorithm often used to visually emphasize the dispersing tendencies of groups in a given graph – both village networks remain singularly centralized. This is the first, and most dramatic, indication to us that the inquisition depositions do not represent the natural social ties of the village.44

In Table 4 we find the count of total people deposed for each village and the segmentation of the depositions by those which are denials and those which are ‘sources’ of incrimination of others.

Table 4

Aggregate deposition figure. A ‘source’ is a deponent who has incriminated at least one other person, i.e. a ‘target’.

Total People Deposed Denials Sources Total People Targeted Targets Inside Village Targets Outside Village

Mas-Saintes-Puelles 412 311 101 297 238 59
Saint-Martin-Lalande 256 196 60 268 189 79

The denials represent the lengthy lists of deponents in MS 609 who swore to have never seen nor interacted with heretics (dixit quod nunquam vidit hereticos nisi captos nec credidit nec adoravit nec aliquid dedit nec misit nec eorum predicationem audivit). Their confessions do not contain any incriminations either of themselves or other people, and as such they cannot be graphed.45 Sources constitute 24% of all deponents (although the absolute number of depositions is slightly higher because several sources, like Ermengarde Boer, gave multiple depositions). 101 deponents from le Mas implicated 297 people in heresy (targets), and 60 deponents from Saint-Martin implicated 268 people.

Within these lists of targets are a subset of people who come from other villages. Identifying who is and who is not resident in the village is not always self-evident, nor is it clear how one categorizes those people who appear to have residences in several villages. If the target has a domus where an event took place in the village, then they are understood to be a member of that village. In other cases, the context makes it quite clear that the accused hails from outside the village and at the same time sheds light on a small but visible anomaly in the graphs which must be addressed. Pelegrina del Mas of le Mas and Ermessende Mir Arezat of Saint-Martin both stand quite distant from the center of their graphs, although not disconnected. Pelegrina was the daughter of the seigneurial clan of le Mas, and she married outside the village to a certain (late) Ysarn, a co-seigneur of Mont Server. Ermessende, on the other hand, came from the seigneurial family at Montferrand to marry Bernard Mir Arezat, one of the milites of Saint-Martin. Both women were clearly connected to both villages, and inquisitors took advantage of these relations to incriminate as many people outside as inside our two villages. The ‘outsiders’ mentioned by Ermessende and Pelegrina do not concern us in their particularity, as they are only mentioned once and forgotten in these two village inquiries.46

What these two depositions do is define for us the ‘network diameter’. This diameter is a technical expression of the extent of a network.47 These relatively few instances where the network of accusations extends outside the circle our two villages, complemented by the fact that both women gave long depositions incriminating many others,48 creates a visual impression of large network diameter which is not actually representative of the dynamics of the graphs. This same effect repeats itself a few more times in the graphs for reasons unrelated to seigneurial marriage – rather because they relay particular stories that inquisitors picked up and explored which implicated people from outside our villages. It illustrates that the graph reaches outwards only in exceptional cases. Despite the fact that ‘outsiders’ make up 25% of the graph nodes, they only account for 5–10% of incriminations. The incriminations overwhelmingly turn inwards at le Mas and Saint-Martin.

As described above, an incrimination represents one person implicating another as participant, owner, or infirmus. The weighted incrimination represents the content of the accusation – an accusation of both adoring and eating bread blessed by the heretici is weighted more than only adoring them (Table 5). Moreover, we need to remember that each time a deponent incriminated someone else, they also incriminated themselves. If someone testified that Bernard senior, lord of le Mas, ate blessed bread with heretici, it is because they themselves also ate blessed bread with Bernard and heretici.49 This ‘hidden half’ is glossed or ignored in a historiography obsessed with ‘Cathar nobility’, but it constitutes a rapid multiplication of interactions with heretici. The fact that so many people were therefore ‘involved in heresy’, yet either confessed so little or are not deposed at all, makes the concentration of incriminations more striking.

Table 5

Summary of incriminations.

Incriminations Weighted Incriminations

Mas-Saintes-Puelles 1,255 1,932
Saint-Martin-Lalande 688 1,047

The center of the graphs can be measured in two ways. The first is the simple question of numbers: the more one is implicated with heretici, the higher one ranks in lists of in-degrees (i.e. being incriminated). However, simple counting of degrees doesn’t account for the placement in the graph of those most implicated in heresis. Those most accused could be dispersed throughout the graph, forming the individual clusters of accusations, like Ermessende Mir Arezat or Pelegrina del Mas, but on a greater scale. In order to understand why the graphs for le Mas and Saint-Martin direct towards a single center requires us to explain the directionality of incriminations, and how they work together en masse.

All accusations are composed of dyads, a pair of people in a graph. A dyad captures the relationship of these pairs: whether Guilhem Ayalric incriminates Bernard Alzeu or vice-versa (asymmetric dyads), or they incriminate each other (symmetric dyads). Table 6 reports the dyads found in le Mas and Saint-Martin. In the registries we find 1,303 dyads for le Mas and 715 dyads for Saint-Martin, of which 40% (522) and 61% (439) respectively are asymmetric. What is the reason for this asymmetry? Like Bernard Alzeu, one side of the dyad may have died before giving a deposition, or they fled, or they just did not show up to depose. Or perhaps they did, but their deposition did not survive. Or perhaps, like Ermengarde Boer, those she named didn’t name her in their depositions.

Table 6

Rates of reciprocal incriminations in depositions.

Total Dyads Asymmetric Dyads Total Dyads – Both Deposed Symmetric Dyads Reciprocity Rate

Mas-Saintes-Puelles 1,303 522 781 202 25.86%
Saint-Martin-Lalande 715 439 276 74 26.81%

Across both le Mas and Saint-Martin depositions, only a quarter of dyads where both testified are reciprocal! The failure of deponents to reciprocate an incrimination nearly 75% of the time explains why the graphs of both villages are highly centralized: those in the periphery incriminated those in the center more often than the inverse. In turn, those in the center incriminated each other more often than they did the periphery. This directionality is visually evident in the graphs, but it is not just a function of brutishly adding up incriminations. The algorithm responsible for the graph is based on Eigenvector centrality, and it calculates not only degrees, but the importance of those degrees by recursively identifying important nodes. These nodes may not have the most incriminations, but they are incriminated with others who are important. They have high Eigenvector scores, and they explain why certain of the dead, and others for whom we lack depositions, appear in the center of our diagrams. Neither Arnald Godalh and the brothers Peire and Bernard Cap-de-Porc from le Mas, nor Ysarn and Andrea de Gibel, nor Bernard Alzeu from Saint-Martin would appear on a simple count of degrees – none of them have out-degrees (incriminating others), only in-degrees (incriminated by others). But they were deeply linked with other people who were both accused by and who implicated others in the centre, and as a result they all have very high scores. On the other hand, Ermessende Mir Arezat has a very low Eigenvector score as her connections are to people who on average don’t count for much in the village depositions of Saint-Martin. Eigenvector score therefore represents the directionality of accusations, not just the volume, and indicates who was most central.

The center of incriminations: the consuls of le Mas and Saint-Martin

The brief analysis above describes in structural terms the forces at work in the depositions of le Mas and Saint-Martin which produce graphs with high centralization of incriminations. The measures of that centralization result in the same people and families at the top 10, 20, and 30 positions. Based on this analysis, the seigneurie of each village – the eponymous seigneurial clan of le Mas, and the Azerats and Campolongs of Saint-Martin – although undeniably the target of accusations, only rank half or a third in magnitude compared to other families (Rehr, 2018: Annex 5).

Some of the most targeted families have been called ‘great families’ or ‘influential families’ in historiography.50 This term is used alternately to describe seigneurial families, or those families who seem to consort with seigneurial ranks, or, even more ambiguously, those families whose members simply appear a lot in early inquisitorial records. ‘Great families’ is a historiographic invention which has no sociological or explanatory dimension, and is usually a generic descriptor floating somewhere in the middle of the ‘feudal pyramid’. We can dispense with the ambiguous ‘great families’ and substitute it for an identifiable trait shared by all the targeted people and families: they populated the consulates of le Mas and Saint-Martin. That these targeted families were consular families is not obvious from MS 609 (nor the few Doat depositions which contain references to our two villages). In fact, the existence of consulates in the Lauragais would not be known from MS 609 save for two sole references to the consuls Bernard Alzeu and Pons Martin of Saint-Martin – both of whose families appear at the center of the graphs. We must reach outside inquisitorial documents to identify others as consuls. But as Wakefield demonstrated, interest in researching beyond the long-standing historiographic bias towards village seigneurie has amounted to zero, even as he cited the very document (a juramentum) that can now tell us who Bernard de Caux and Johan de Saint-Pierre were targeting. Moreover, research on consuls in the lands of the Count of Toulouse has been mis-directed or inhibited by the dependence (Bourin, 1987a: 301–30; Bourin, 1987b: 145–80) on the work of Jean Ramière de Fortanier (1939: 44) in his Chartes de franchises du Lauragais where he counts only nine consuls in the Lauragais before 1249, but only because he only looked for consul charters and constitutions. He did not look outside narrow constitutional documents to identify the consulates active by the time of inquisition; I have identified 29 so far through other documents.

Numerous consulates are identified in a series of juramenta found in the Archives nationales at Paris dating to 1242–43: these are the oaths of village milites, consules, and probi homines to continue honouring the Treaty of Paris-Mieux (1229), sworn to Louis IX after Raimond VII’s attempted rebellion.51 There is a near perfect correlation between the names identified as targets of the inquisition and those consuls and important probi homines named at the beginning of the juramenta. This correlation in an oath, within 24 months of inquisition, cannot be an accident. Comparative analysis suggests that the correlation is the product of social relations in the villages similar to those found in the consulates of Toulouse, Montauban, and Albi.52 Year after year, consulates cycled through the same social groups and various families occupied seats in the consulate in rotation; the families intermarried and remarried.53 Table 7 represents lists of consuls and key probi homines derived from comparing the juramenta and Chartes de franchises with MS 609.

Table 7

Consular families from MS 609 at le Mas and Saint-Martin. * indicates a single extended family that intermarried or descended from shared mother or father.

Mas-Saintes-Puelles Saint-Martin-Lalande

Aleman* Alzeu
Amelh* André
Barrau de Gibel (milites?)
Canast-Bru de Gozens
Saint-André (Cap-de-Porc) de Saint-Nazaire
de Quiriès (milites) Faure-Verazilh
Donat* Hugole
Garnier* Martin
Gauta la Vauza
Godalh*
Picrelh*
Rosengue
Vidal

Incriminating the consuls of le Mas and Saint-Martin

How do the incriminations against the consular families manifest themselves in the depositions? And how can we know that these incriminating statements were orchestrated by inquisitors and are not some random coalescence of the spontaneous confession of memories, or the objective reflection of a given family’s ‘adhesion to heresy’, or the result of internecine village conflicts? Looking at any individual deposition in isolation, we could be convinced that these issued from some spontaneous recollection of an event, perhaps motivated by personal or political animus, even if it was reduced and contained in the inquisitor’s formulary. Many of these patterns have such a prosaic, innocuous quality that we can imagine producing them naturally ourselves were we pressed to recall memories of events that go back five, 10, 30, even 50 years. This goes some way to explain why they have not been identified before, particularly if one is already convinced that the inquisition at Toulouse was a general survey of the Lauragais. It is only by meticulously comparing hundreds of depositions in the details of their construction that patterns emerge which reveal the directing hand of the Bernard de Caux and Johan de Saint-Pierre as they obtained the testimony they wanted.

A full confession?

On 3 March 1246, inquisitors recorded one brief statement from P. Laurenc (vel Massota) of le Mas which in one blow incriminated Peire Gauta senior, Garnier senior, Guilhem Vidal, and the brothers Esteve, Arnald, and Guilhem Rosengue, for fraternizing with Peire Petit, hereticus, at the home of Peire’s sister, Rixen.5455 The recorded memory is vague, just a list of names dating back 20 years, although Laurenc was sure that he did not see anyone adore two decades before, nor himself adored. Nor, he claimed, did he see heretics thereafter. The memory, like so many in MS 609, stretches back through decades to incriminate just a handful of people – the people defined in the target list. And in a refrain now familiar from the macroanalysis above, all of those he incriminated were themselves deposed at Saint-Sernin – and yet none remembered to include Laurenc in their depositions. P. Laurenc’s story brings us to the problem which can be rapidly comprehended in Table 8, the apparent low incidence of reported events of heresy.

Table 8

Summary of incriminating depositions from le Mas and Saint-Martin.

Events per Deposition Incriminating Depositions Depositions Incriminating Consular Families

Mas-Saintes-Puelles Saint-Martin-Lalande Mas-Saintes-Puelles Saint-Martin-Lalande

1 29 24 20 69% 20 83%
2 19 16 15 79% 14 88%
3 17 8 11 65% 6 75%
4 8 5 8 100% 5 100%
5 10 2 10 100% 2 100%
6–17 34 14 34 100% 13 93%
Totals 117 69 98 84% 60 87%

Of the 117 people who journeyed for two to three days from le Mas and Saint-Martin to give incriminating depositions at Saint-Sernin, 30% only recited one event from their lives, while 50% recited one or two. Of those depositions, more than 80% were directed (i.e. 158 out of 186), like that of P. Laurenc (or Guilhem Ayalric), at the consular families. These numbers represent the intersection of two problems. The first is the low probability that any given deponent would have had only one or two incidences of heresis to report from their lives. The lack of reciprocal incriminations, as we saw with Ermengarde Boer, Guilhem Ayalric, and many others in the graphs, already tells us that events were under-reported. But from Laurenc’s testimony we can detect even more under-reporting: only one of the six people he cited, Garnier senior, reported ever being at Rixen’s domus, and that was 10 years later.56 This means that 80% of attendees at just one event, in just one deposition, are unreported. If we extrapolate from this, we are faced with an extraordinary number of unreported acts of interactions with heretici. The second problem is the low likelihood that so many deponents would have reported a small number of similar (or the same) events which implicated the same people. We will see in the next section how the impression of small groups was created by inquisitors.

The problem of under-reporting throws into relief the significance of how many depositions actually targeted consular families. We can scrutinize these numbers in greater detail, by event. Table 9 reports how many events feature accusations against consular families, how many feature seigneurial families, and how many feature either of them. Over 60% of events feature consular families, the majority of those exclusively. P. Laurenc implicated Guilhem de Rosengue and his brothers, one of the families made famous by Douais (1900, vol 2: 4) (recapitulated by Wakefield) as ‘une des principales familles du Lauraguais’ who ‘n’avaient cessé de montrer le plus grand attachement pour l’hérésie’. As mentioned, Guilhem himself was questioned, but never mentioned Laurenc or the Petits.57 Guilhem’s deposition reports four events, and despite others testifying that he appeared with dozens of people, including the seigneurial del Mas family, his deposition is solely focused on events which would indict other consular families: the Vidals, Gautas, Amelhs, and the Garniers. As the depositions become longer and recount more events – from one event to five and then 10 – we find the longer depositions come from the consular class, like those named by Laurenc: Peire Gauta, Guilhem Vidal, and Garnier senior.

Table 9

Summary of incriminating events from le Mas and Saint-Martin.

Incriminating Events Events Featuring Consular Family Events Featuring Seigneurial Family Events with Both/Either

Mas-Saintes-Puelles 510 328 64% 150 29% 365 72%
Saint-Martin-Lalande 230 135 59% 61 27% 157 68%

In a different way, but to the same effect, Bernard Hugo58 and Johan Cabrigas59 of Saint-Martin should have had many memories to recount as they were implicated by others frequently. But they managed to escape with deposing just one or two events about the consuls. In these various ways, the incriminations accumulate towards the target lists for le Mas and Saint-Martin, as consular families were forced to confess against each other and their own families, while those outside rarely felt the reciprocated heat of accusations.

‘…et plures alios’

There were many credentes who appeared with heretici whose names are unknown, and can never be known, as Bernard de Caux and Johan de Saint-Pierre were not interested in recording their details. We saw one example of this above with the expedition to Saint-Papoul of certain people of Saint-Martin (and Issel and Labécède) to free two heretici held by the abbot of the monastery there, by payment or by force.60 The list of those named on the expedition, and the number of times they are named, proved the same pattern just discussed: a repeated focus on a handful of people, the consuls of Saint-Martin. But these depositions about the march on Saint-Papoul demonstrate in another way the obviousness of Bernard de Caux and Johan de Saint-Pierre’s focus on the consuls. Esteve Faure, chief among those named on the campaign along with his brothers and uncles, gave the names of six others who marched on the abbey, and he added that there were up to 18 others from Saint-Martin.61 Friar Bernard, the inquisitor that day, ignored any names beyond consuls, settling for ‘and many others, up to 18’ (et pluribus aliis usque ad X et VIII). Others similarly questioned gave between one and 15 other names, followed by ‘with many others ‘(cum multis aliis de Sancto Martino, cum pluribus aliis, et plures alios, etc). Why would the inquisitors accept ‘and many others’ rather than the names of every possible credens and fautor who were prepared not only to pay to release two heretici, but prepared to release them from the abbey by force (in universo cum armis)? More generally, why did Friars Bernard and Johan not insist on obtaining or recording longer lists of the credentes and fautores which their own manual (Pegg, 2003: 111–25; Arnold, 2001: 19–47) – and council mandates from which they derived their manual – made the special object of the inquisitio heretice pravitatis?

This may seem an uncommon type of event by which to judge the habits of the inquisition at Toulouse. But ‘plures alios’ was accepted by inquisitors 122 times in lieu of the names of more credentes, about one out of 5 deposed events. Sometimes, like in the depositions of the brothers Galhard and Raimund Amelh of Le Mas,62 they reported attending events with between 30 and 60 credentes dating to 12 years beforehand, but they were stopped by inquisitors before even listing10 of them (or, the inquisitors told the notaries to only record specific names). The names that overlap are familiar from our lists: Godalh, Amelh, Gauta, Barrau, de Quiriès. Even the senior of the del Mas brothers, Bernard, remembered an event with 40 people, but the inquisitors stopped recording after only naming two of his brothers, and then Peire Gauta senior and Guilhem Vidal.63 At other times the numbers of attendees are smaller, but refer to the same target list. Raimund Causitz remembered adoring and listening to heretici and accepting the peace with Peire Gauta, Raimund Amelh, Arnald Godalh, Guilhem Vidal, Esteve de Rosengue, and Bernard and Bertrand de Quiriès on several different occasions across a decade, each time with ‘many others’.64

The pattern is no different in the depositions of Saint-Martin. Ermessende, wife of the miles and village seigneur Bernard Mir Arezat, remembered vividly a joke made at her husband’s expense fourteen or fifteen years beforehand. A number of the seigneurial family were gathered at Guilhem de Saint-Nazare’s home with consuls Bernard Alzeu, Ysarn de Gibel, several of the Faure-Verazilhs and ‘many others’.65 The joke was recorded by the notary across several precious lines, yet no space was spared for an untold number of other credentes.

Unsurprisingly, mentions of consular families correlates with ‘many others’ – 75% of mentions of ‘plures alios’ involved the consular families. The only way for these families to become so central to the inquisition at Toulouse is not their supposed greater affiliation with heresy, but that inquisitors made them central.

It might seem as though some of these anonymous ‘other people’ are just failure of memory while trying to recall events of the distant past (et plures alios de quibus non recolat). Or in some cases they were people they did not know (plures alios quos non novit). These responses interpret the problem incorrectly. Every one of the deponents who used the term ‘plures alios’, in whatever form, can be proven to have under-reported their appearances with heretici. The memories were thus chosen because of the content they conveyed, even if incomplete, because they suited the inquisitors’ hunt for incriminating evidence against consular families. Others memories, perhaps more recent, more complete, would have incriminated those who were of far less concern to inquisitors.

The first inquisitor formularies of the 1240s were not simply devices for structuring confession. They also controlled the content. A credens, fautor, and indeed other categories had specific definitions which categorized heretical actions and in turn controlled what the inquisitors looked for (Arnold, 2001: 19–47; Pegg, 2003: 111–25). The depositions of MS 609 are so consistently and rigorously structured around the same formula that the variations stand out. Some variations are formal: direct and undecorated with the usual dates, places, and co-actors, as when Na Flors, wife of Galhard del Mas, simply stated that she ‘sent bread and wine to Bernard de Maireville, hereticus, on behalf of Arnald Godalh’, we don’t know when or where.66 Less common are the strings of similarly direct but ‘off topic’ accusations that Guilhem P. Barba junior made against Peire junior, the son of consul and scriptor Peire Gauta senior. Barba said that Peire junior stole ponies from the abbey at le Mas and sacerdotal garments from the church, and moreover he had an affair with a close blood relative.67 What is rare about these accusations is that they have nothing to do with any usual heresis. They are more notable because of who these uncommon accusations are directed at. Whether accusations of theft or tattle-tales, they implicate the same people.

Peire Raimund Prosat, an acolitus of the Church,68 recalled gambling with some men of the village in Peire Gauta senior’s workshop 25 years before. According to Peire, at some point during the game, Bernard de Quiriès stood up, climbed on a chest, and proceeded to urinate on Peire Prosat’s tonsure while cursing the Catholic Church. One of the other gamblers, B. Lapassa, had testified just before Prosat, but he was recalled to Saint-Sernin to be deposed again for just this one event. He wearily confirmed a vague memory of it, but couldn’t remember all the details except that it was indeed at Peire Gauta’s workshop. Bernard de Quiriès himself was then recalled to Saint-Sernin seven weeks after his first deposition solely to account for this story. He recalled that he was provoked by the other gamblers and consequently urinated on the table in complaint, but he said that if any urine struck Prosat it was entirely by accident.69

These stories are in themselves interesting for how they take us out of the suffocating sameness of the formulaic depositions. But it is important to remember that historians have insisted that the formulaic repetition resulted from the early inquisition’s focus on a small number of heretical acts, thoughts, and relations, as well as brothers Bernard and Johan having to process so many people during a short period. Yet this results in a contradiction: Bernard and Johan left aside untold numbers of unnamed ‘others’ who were involved directly and verifiably with the heretici, yet focused specific effort in gathering whatever information they could which might build a case of any anti-Church inclination on the part of consuls.

At times, certain deponents stated that they adored the heretici at the request or entreaty of someone else (adoravit ad preces, ad instanciam, or rarely, compulit), or they were asked by someone to guide or accompany heretici from one domus to another, or from one village to another (duxit, associavit). Although numerous people guided, accompanied, and adored heretici (65 different people from le Mas and Saint-Martin at one time or another acted as guide), only a handful of people are recorded by inquisitors as having asked others to do so: Arnald Godalh, Bernard Alzeu, Ysarn de Gibel and his wife Andrea, Bernard Martin, and other consuls (those consuls themselves are also cited as guiding and accompanying heretici from place to place, three to four times more than average). The questions stayed within the usual diagnostic criteria for identifying credentes and fautores, but the application of the formulary was intensified when investigating certain people. There was a sliding scale of responsibility for the persistence of heresy, but that judgement did not come from the deponents. It was imposed on deponents and the accused by the inquisitors through specific questions about the consuls and probi homines. Those questions drew out these extra pieces of incriminating information that could demonstrate a given person was more inclined or attached to heresy than others. Inquisitors could then use this evidence to prove that someone deserved greater investigation. To these three patterns we can add others (Rehr, 2018: 55–71):

  • Hereticos non novit’: In the deposition of Bernard Hugo mentioned above,70 we find the inquisitors completely uninterested in the names of the heretici, settling for duos hereticos quas non novit. This same pattern is seen across the depositions: heretics are unknown (non novit) or not recalled (non recolit), but what is recalled is who attended the event, and they are always the consuls at the top of the target list.
  • Inquisitors’ marginal notes: The inquisitors’ marginal remarks, unstudied in aggregate until now with many left out of extant transcriptions, note occasional punishments or location of current residence. But the greater part can be linked together to see inquisitors investigating certain ‘lies’ as they sought to obtain indicting confessions from and against consuls, with the greater number of those labelled suspectus est, or hic dicitur catholicus, being consuls subject to investigation not applying to the thousands of others named in the manuscript.
  • Inquisitors’ supplementary questions: The term requisitus/requisita (‘asked about…’), frequently found in later inquisitorial registers, indicates a specific question asked by the inquisitor. It only appears three times in the 685 depositions from le Mas and Saint-Martin. But the inquisitor’s questions can be seen in other ways which can seem like a spontaneous recollection or addition by the deponent at the end of a recollection, or at the very end of the deposition, even after the abjuration. These are often framed as dixit etiam or postea dixit, or even simply added with et. The seeming randomness and spontaneity disappears once one correlates the people and acts mentioned, like the seemingly insubstantial addition that breaks the formula in very first entry of MS609, ‘Et quando ipse testis exivit domum invenit Willelmus de Rozergue intrantem ad dictos hereticos’.71
  • The same stories deposed by different witnesses: Wakefield remarked that it was difficult to match stories from person to person. But his concern was with recreating specific exact moments in time, rather than comprehend the motives behind evidence-gathering by the friars from obscure, half-remembered stories 5, ten, fifteen years old. The overlaps are only explained when one looks at the attendees who are mentioned in them, and not the heretici nor the deponents themselves. Inquisitors were only interested in the people found in the overlap of those stories, often containing plures alios and hereticos non novit.72

Results

From combined macroanalysis and close reading, a revised picture emerges from MS 609 which makes sense of both the structure and tone of the registry. The central position of consuls in the inquisitions of Le Mas and Saint Martin were neither an accident of spontaneous confessions of credentes nor a reflection of consuls having been more inclined to heresis. Their centrality was solely the result of the investigative goals of Friars Bernard de Caux and Johan de Saint-Pierre at Saint-Sernin. This alone explains why the two inquisitors could possibly have conceived of an inquisition of the entire Lauragais in one short year. This also answers Wakefield’s puzzle: Why did the inquisitors so frequently overlook leads pointing very clearly to many other credentes and fautores at Saint-Sernin (Wakefield, 1983: 221–22)? The inquisitors had very particular goals, but it was not a ‘general survey of the faith’. Depositions were not short because inquisitors didn’t have the time to explore every lead and challenge every lie. The depositions at Saint-Sernin were short because they recorded exactly what was needed by inquisitors to incriminate a very particular group of people in brief period, a group of people they had decided upon before the first deponents even journeyed to Toulouse. This is a falsifiable claim: the depositions make clear that when inquisitors wanted to spend time with a credens they could and would do so.

The patterns dominating MS 609 become even more striking when we realize that depositions were frequently given without the presence of inquisitors Bernard de Caux or Johan de Saint-Pierre, a practice noted by both Dossat (1959: 240) and Wakefield (1993: 65). The investigations focused on specific people, and once inquisitors turned up a story that could indict those they targeted, it could be easily followed up by the small staff of notaries who provided the day to day continuity at Saint-Sernin.73 The inquisitors made their notes so the key events and relations were later interrogated (and surely the few extant margin notes in MS 609 pale in comparison to the provisional notes and lists which must have been produced by the thousands of depositions at Saint-Sernin). Thus many credentes were deposed to confirm only one or two stories, despite the evidence of decades practicing whatever heresis the inquisitors believed was occurring in the lands of the Count of Toulouse. It didn’t matter that scores of credentes and fautores were ignored and sidelined in each village, nor did it apparently matter who the particular heretici were. Sometimes it didn’t matter if the evidence was even stricto sensu the heresis inquisitors had set as their criteria, because they were assailing the character of those consular families who inquisitors wanted to portray as allied against the Church.

Future research

I would conclude with a few considerations on the implications for future investigations into this study’s findings and historical research methodologies applied to MS609. The first is the obvious question, ‘Why the consuls?’. The current state of research makes this question impossible to answer. Notwithstanding the utter absence of rural and village consulates from the historiography of Catharism, we lack even published sources or studies which could shed light on this social group, their composition, their roles, and their influence, prior to Alphonse de Poitiers, brother of King Louis IX, assuming the title of Count of Toulouse upon Raimund VII’s death in 1249.74 I am disinclined to any generalized conclusions drawn from existing studies of inquisition and consuls or ‘urban elites’ in the lands of the Count of Toulouse for several reasons. The first is that little unites these studies outside of resting under the banner ‘inquisition of urban elites’: some feature Catharism and inquisition as a backdrop to, or means to conduct, principally social and institutional studies of Toulouse (Mundy, 1985; Mundy: 1990; Mundy 1997; Mundy 2006);75 others are early papal inquisitions at Montauban (Feuchter, 2007) where the author pursues a traditional relationship of inquisition to heresy (Catharism, Waldensian); still others are episcopal inquisitions which the authors have uncovered were rooted in local power struggles between bishop and secular ‘elites’ of Albi (Biget, 1971; Théry, 2002).

The second argument against any generalized answer is the context of this inquisition itself, which I outline here as a series of provocative observations and questions to be answered only by returning to regional, royal, and papal archives. The first begins with a precise historical question to reframe the inquisition: ‘Why prosecute the consuls over milites, across the Lauragais, and to the near exclusion and ignorance of everyone else who were implicated in heresis by the inquisitors?’.76 It is clear from the juramentum from Mas-Saintes-Puelles (and Lavaur, Puylaurens, Fanjeaux, inter alia) itself that King Louis IX, the milites, and the probi homines of the Lauragais thought their consuls in the 1240s were important enough that they should guarantee the oaths with their seals (nos supradicti consules de mandato speciali predictorum militum et aliorum hominum universorum de Manso presentem paginam sigilli nostri munimine duximus roborandam).77 These juramenta were enacted at the end of Raimund VII’s rebellion against the Louis IX (a rebellion broader than Raimund VII, allied with England, and extending up to within catapult range of Poitiers (Le Goff, 2009: 102–08)). The signees swore to council Raimund should he become ‘belligerent’ again, and that, should battle come, they will adhere to the King in his matters, and Church in its matters: against heretics, their supporters, or anyone allied against it. Two years later, these very same signees – milites, consuls, probi homines – then travelled by the thousands from this vast region to Toulouse, in a short period of time and to give depositions against those same consuls. I note as well here that the memories set down by inquisitors end in 1243, the year of the oaths, two years before this inquisition. The effect is that this inquisition seems less and less a ‘survey’ and more and more ceremonial, processional, and ritualistic.

Going further, it is not clear which power(s) organized, enforced, or compelled such coordinated action on such a massive scale. If the consuls were chosen, then by whom and under what criteria? Some historians have observed (Chiffoleau, 2006: 167–69, 188–96, 207–14) that inquisition is not conducted within the borders of princes, but rather controlled by those borders. This research direction, which shifts the study of this inquisition from religious considerations (i.e., where there was heresy, there was inquisition)78 to earthly considerations, is suggested by the temporal context of the inquisition itself. To whom were the consuls and milites of the Lauragais subject by the time of friar Bernard and Johan’s inquisition – Louis IX, or his ostensible direct vassal, Raimund VII? Why did Raimund VII, after fiercely resisting inquisition for 15 years, suddenly seem to throw his support behind an inquisition which targeted the very consuls and lords of the Lauragais who swore to the juramenta? Did this change in attitude have any relationship to Raimund VII’s concurrent quest to receive papal support for annulation of one marriage and proposed new marriage, in order to solidify his claims in Provence? Is there significance to the Lauragais’ geographic and political position at the nexus of ambitions of the kings of France and Aragon in Languedoc, Frederic II in Provence, and Raimund VII’s defense of his own patrimony in both?

Finally, returning to the manuscript that provoked these questions, historians are offered an unusually large and contiguous source with which to test and falsify the conclusions of this study (and other arguments about its contents). This marks a rare opportunity in medieval historical research, let alone for the ‘Cathar debate’. The famed difficulty and sheer scale of the manuscript, its historic lack of accessibility, and the deeply problematic transcriptions available until now, have allowed historians to judge its contents (and to select extracts from it) by expectations of what it should contain. The digital edition of MS 609 (and documents related to the social context of the inquisition) has been constructed in such a way as to maximize the opportunities for historians to investigate its contents, including by making the edition and its data available under Creative Commons. The methodology presented in this study, combining network analysis and close reading, is but one possible approach – in this case suited to the question of who accused who. Other techniques, manual or digital, can be applied to numerous other investigations of the contents of MS 609, from the apparatus of the inquisitorial record itself to the beliefs and social worlds of the subjects. But there is a double-edge sword to digital availability: our work must be rigorous and testable, knowing that it can be submitted to rapid, intense scrutiny against large samples. With this challenge, our methods must become more explicit and more fine-grained in argument. The real challenge this poses to old Cathar historiography is the question of what is to be found in the ‘negative space’ in the manuscript: that the immense amount of information which exists outside the cherry-picked, strip-mined samples (the ‘positive space’) that have served to demonstrate the value of MS 609 until now.