Reconsidering Donizone’s Vita Mathildis (again): Boniface of Canossa and the Emperor Conrad II

Introduction

Boniface of Canossa was the most powerful secular figure in Northern Italy during his lifetime. He held titles and lands across Emilia, Lombardy and Tuscany based on a core around Mantua, Reggio, Modena and Canossa itself. [1] Like his father Tedald and grandfather Adalbert Atto, Boniface is almost universally portrayed as a firm supporter of the German Emperors. [2] This relationship is often portrayed almost as an alliance between equals rather than a hierarchical connection between lord and vassal and is seen to reach its zenith under Conrad II (1024-1039) when Boniface participated in a series of campaigns within and beyond Italy in support of his emperor, [3] before collapsing dramatically and entirely under Henry III (1039-1054) and Henry IV (1054-1105) leading to open conflict between Canossa and imperial forces. [4] The main consequence of this alliance is portrayed as the extension of Boniface’s rights and landholdings demonstrated most especially through his installation as duke of Tuscany. [5] However, the strength and distinctiveness of this Cannosan-Imperial connection is generally overstated. The supposed alliance between Boniface and Conrad is not as clearly defined as is usually suggested. There is little to suggest that Boniface was particularly favoured by Conrad and several factors that indicate that the emperor took active steps to counter Boniface’s power. Boniface was simply one of many several powerful figures within the imperial relationship network, kept in balance through the empowerment of other members of this network.

Boniface’s exceptional role within imperial politics described in the typical narrative stands at odds with typical structures of rule and authority deployed by the German emperors within Italy and the Empire more generally. The Ottonian and Salian systems of itinerant kingship demanded the existence of a carefully balanced relationship network in every region of the empire, whose members were reliant to a large extent on a connection with the Emperor for the legitimacy of their authority [Bernhardt 1993, 50-51]. The need to maintain such a system is a partial explanation for the empowerment of Italian bishops under these dynasties; doing so provided the emperor with another means to influence the structures of power in the region [Fumagalli 1979, 70-81]. This system of balanced relationships held in creative tension had been a mainstay of Frankish rule [Nelson 1992; Wood 1994; Rosenwein 1999], and was continued by the post-Carolingian rulers of Italy [Rosenwein 1996a; 1996b]. The apparent installation of Boniface as an imperial viceroy in Italy on the basis of a unique bond with the emperor is a substantial departure from this system and therefore demands more careful attention.

The reliance of Boniface and his family on the largesse of the emperors, bestowed as a result of their exceptional loyalty, has been overemphasised. Although the Canossans obtained some of their lands and rights through imperial concessions, this only formed a small part of the basis of their power. The family had progressively expanded their lands, wealth and authority through several methods which did not rely on imperial support and many of these alternative sources of power worked at odds with the goals of the emperors. Most blatantly, Boniface’s construction of a palace in Mantua from which he issued judgements represented a claim to public authority and jurisdictions and challenged the rights of the emperor [Fumagalli 1976, 47; Marani 1987, 215-216; Houghton 2015, 399-400]. Beyond this, the widespread and well documented usurpation of Church property by the Canossans [Fumagalli 1987, 162; Golinelli 2001, 513-514, 1991, 82-85] is a further example of an alternate source for their expansion and also undermined the imperial policy of empowering and protecting these institutions. [6] Boniface’s ability to expand his power at the expense of the Church should not be seen as an indication of Imperial acquiescence to his actions: in several cases, as will be demonstrated below, Conrad actively sought to empower bishops as a counter to Boniface’s growing influence. Boniface’s expansion in this manner was enabled by imperial absence which left these bishops exposed, a key example of this is his seizure of lands from the bishop of Mantua [Fumagalli 1976, 47; Gardoni 2006, 224-226]. Likewise, marriage into major Italian families greatly extended Canossan lands over several generations without recourse to the emperor [Houghton 2015, 398-400; Lazzari 2008, 107]. Boniface’s father, Tedald, expanded his influence and holdings in Tuscany through his marriage to the daughter of Hugh, Margrave of Tuscany [Lazzari 2008, 109]. Boniface himself acquired vast lands in Lombardy (Brescia, Cremona and Mantua), Veneto (Verona), and Emilia (Reggio and Ferrara) as a result of his marriage to Richilde of the Gisilbertini [Anton 1972, 533; Rinaldi 1997, 74-76]. This marriage did little to further imperial interests: it did nothing to secure the support of the Gisilbertini who remained closely associated with Arduin of Ivrea, the main rival of Henry II in Italy [Lazzari 2008, 109; Houghton 2015, 398-400]. Boniface and his family were not dependent on the emperor for their power, and in fact much of their expansion ran counter to imperial interests.

The complexities of the relationship between Boniface and the emperors have been highlighted by several authors. The chronology of the rising tensions between Boniface and Henry III have been discussed extensively. [7] More recently Boniface’s supposed alliance with Henry II has been challenged [Houghton 2015]. Of most relevance here, Boniface’s loyalty to Conrad II has been questioned, particularly in the early years of Conrad’s reign [Anton 1972, 556]. However, this more nuanced consideration has had little impact on the common presentation of Canossan-imperial relations.

A key reason for this disconnect is the selection of sources on which the typical narrative is based. There are three core narrative sources for Conrad’s relationship with Boniface. Donizone of Canossa in his Vita Mathildis provides details of several events connecting the two in elaborate detail. [8] Arnulf of Milan in his Liber gestorum recentium mentions both Conrad and Boniface briefly and gives some further indication of the connection between the two figures. [9] Wipo’s Gesta Chuonradi II imperatoris presents an imperial perspective of Conrad’s rule, including his expeditions into Italy [Wipo 1978]. However, as Anton (1972) notes, most modern work on Conrad and Boniface has focused on the account given by Donizone. This is problematic, as Donizone’s goals frequently led him to exaggerate or even fabricate events to better suit his carefully constructed narrative. This has led to a misrepresentation of the relationship between Conrad and Boniface in much of the modern literature.

In order to redress this over-reliance on Donizone it is necessary to reconsider not just his aims in writing, but also those of Arnulf and Wipo. All three authors were driven by specific goals which led them to present different versions of events, omitting, including and fabricating details in order to fit their narrative. This in turn has influenced modern accounts of the relationship between Conrad and Boniface. After analysing the background and aims of the authors, this piece will reconsider the events of Conrad’s reign through three key periods: 1) Conrad’s initial expedition into Italy and its aftermath (1026-1032). 2) Conrad’s campaign for the Burgundian crown (1032-1034). 3) The final years of Conrad’s life, including his second campaign into Italy (1035-1039). Through each of these periods, Boniface has been presented as a loyal ally of the imperial cause and a prominent figure within Conrad’s system of rule. However, these case studies will be used to construct an alternative narrative of the relationship between Boniface and Conrad and to demonstrate that for much of Conrad’s lifetime Boniface did not hold any particularly special connection with the emperor and instead Conrad consistently took steps to balance Boniface’s growing power.

The Sources

Riversi, through a painstaking study of the text, has correctly identified a deliberate merging of history and fiction, or a programma di veritá, within Donizone’s poem Vita Mathildis [Riversi 2013, 260-264]. Donizone had read numerous medieval, classical and sacred texts, [10] and used the methods of rhetoric employed within these works as the basis for an account which presented events in a manner which suited his grand narrative of the Canossan family, often altering details or fabricating events in the process. [11] The ubiquity and subtly of Donizone’s programma di veritá necessitates a firm understanding of the poet’s goals. The most obvious aim was to aggrandise Matilda and, by extension, her family, [12] but Donizone also set out to glorify his monastery at Canossa [Golinelli 2008, x, 1992, 201; Riversi 2013, 194-197]. Both of these overt goals were connected to and driven by a more subtle aim: to legitimise Matilda’s control of her lands and counter the ongoing rebellions against her in the 1110s when Donizone wrote [Fumagalli 1978, 32-33; Golinelli 2001, 515; Haverkamp 1988, 91]. Donizone’s efforts to legitimise Matilda’s landholding extended into the chapters dealing with her ancestors who were depicted as close allies of the emperors [Riversi 2013, 251-259], and the poet went out of his way to minimise the conflict between Henry III and Boniface. [13] The construction of the Canossans as an idealised, loyal family which had received its lands lawfully and rightfully from the German emperors as a result of their unyielding support was absolutely central to Donizone’s programma di veritá [Riversi 2013, 251-259]. Through this Donizone sought to demonstrate Matilda’s legitimacy to Henry V (1105-1125) and the rest of the Empire.

Arnulf of Milan was the grand-nephew of archbishop Arnulf of Milan (998-1018) [Fasola 1980, 1020; Fliche 1912, 599]. His Liber gestorum recentium, written in five books, charts the history of the Church of Milan and its archbishops from 925-1077. Arnulf’s central goal was to uphold the glory and power of the archbishops of Milan [Violante 1971, 282] and their capitanei vassals, including Arnulf’s family [Fasola 1980, 1020; Stock 1983, 163; Violante 1971, 281-282]. When Arnulf was writing, he completed his first three books in 1072 [Violante 1971, 282], the powers and privileges of the archbishop were under serious threat from the reforming popes but also from within in the form of the Paterene movement. [14] Prompted by these challenges to the status of Milan, and as a corollary to this core goal, Arnulf sought to secure allies for the archbishopric. This included the Canossans: Boniface of Canossa was presented as one of two “Lights of the Kingdom” (duo lumina regni) alongside Aribert of Milan. [15] This praise was designed to secure the support or at the neutrality against the reform papacy of Boniface’s widow, Beatrice of Canossa, and his daughter, Matilda, and resulted in Arnulf’s emphasis of the importance of Boniface within Italy.

Wipo’s Gesta Chuonradi II Imperatoris is often ignored in studies of the Canossans, not least because Wipo makes no mention of Boniface whatsoever. However, this work still provides an important parallel view of Conrad’s rule in Italy and the supposedly pivotal events in his relationship with Boniface from the perspective of the German court. Wipo served as a chaplain to Conrad II and was present at his coronation and several other key events [Mommsen, Morrison 1962, 42]. He also claimed a close connection to the young Henry III: Wipo remained at the imperial court III after his Conrad’s death in 1039 [Mommsen, Morrison 1962, 43; Wolfram 2006, 26]. The extant edition of the text is likely based on a revision in or after 1046 with changes made to emphasise the role of Henry III [Mommsen, Morrison 1962, 43-44; Wolfram 2006, 26]. This late date of completion has led Wolfram to argue that Wipo altered events to fit the needs of his narrative [Wolfram 2006, 26]. This is certainly the case, like Donizone and Arnulf, Wipo had several good reasons to distort his account. Wipo sought to glorify Conrad and, by extension, Henry. He also set out to justify Conrad’s rule and aimed to influence Henry’s actions [Weinfurter 1999, 47]. In particular, Wipo sought to instruct Henry in the proper treatment of his clergy, going so far as to openly criticise Conrad for his attempts to remove bishops, including Aribert of Milan, from their positions [Weinfurter 1999, 47]. These goals led Wipo to produce an idealised account of Conrad’s rule designed to elevate not just the emperor, but also his clergy. These aims also meant that, unlike Donizone and Arnulf, Wipo had no need to inflate the position of Boniface: to do so would not serve his narrative purpose. As such, while Boniface’s presence in the Gesta would certainly be a strong indicator of an augmented role, his absence suggests that Boniface’s connection to the Emperor was not remarkable in the eyes of Wipo, the German court or Conrad himself. The Gesta is therefore a valuable, if overlooked, source for the relationship between Canossa and the emperor.

Beyond these narrative sources Conrad’s charters are an important resource for the study of his connections with Boniface of Canossa and his manner of rule as a whole. These documents allowed Conrad to exert and express his authority in several ways. They were a statement of his right to intervene in an area [Keller 2011, 104-105]. They underlined a claim to legitimacy of rule by drawing on the traditions of previous kings and emperors through the use of established phrases and formats [Goldberg 2009, 213; Guyotjeannin 2001, 17-35; Keller 1998, 431-435]. They provided a political connection between the emperor, the recipient and any witnesses or petitioners, tying their interests together [Bartoli Langeli 2002, 205-206; Bresslau 1967, 193-204; Koziol 1992, 47-54]. At the most basic level, they supplied a means for the emperor to empower his vassal's through grants of lands and rights [Keller 2011, 78-79]. These charters are important as they can provide an outline of political networks within Italy and clarify Boniface’s position within these networks in greater detail and with less rhetoric than the narrative sources.

Conrad’s First Italian Campaign and Aftermath (1026-1032)

In 1026, having secured his rule in Germany, Conrad entered Italy in pursuit of the Lombard and Imperial crowns. [16] We are told that he enjoyed the support of the Italian bishops, most notably Aribert the Archbishop of Milan, and Boniface of Canossa but was opposed by the rest of the nobility led by Ulrich Manfred of Turin, who offered the crown of Italy to Robert II of France and then William V of Aquitaine. [17] Conrad also came into conflict with the city of Pavia, which destroyed its imperial palace on the death of Henry II [Nasali Rocca 1955, 288; Weinfurter 1999, 47]. Most of Conrad’s opponents provided no resistance when he arrived in Italy and were welcomed back into the imperial fold. The exceptions were the city of Pavia with its allies, Adalbert of the Otbertenghi and William of the Aledramids [Wolfram 2006, 99], which maintained resistance to Conrad until early 1027 [Schumann 1973, 33-34; Wolfram 2006, 102], and Rainer duke of Tuscany who held out in Lucca in February and March of 1027 before submitting [Anton 1972, 535; Bertolini 1971, 99]. While Pavia was able to negotiate a reconciliation with Conrad [Wolfram 2006, 102], we are told that Rainer was stripped of his duchy which was awarded to Boniface of Canossa in return for his support of the emperor [Fumagalli 1978; Golinelli 2001, 511; Rinaldi 1997, 64].

It is important to place these events in the context of Canossa’s relationship with previous German emperors. The Canossans had enjoyed several moments of clear alliance with some of Conrad’s predecessors. Adalbert Atto supported Otto I in opposition to Berengar II. [18] Tedald fought Arduin of Ivrea in favour of Henry II. [19] However, this relationship was under strain in the early years of Boniface’s rule as demonstrated by his absence from the charters of Henry II combined with the installation and empowerment by the emperor of numerous groups and individuals at odds with Boniface in and around his lands [Houghton 2015]. At the death of Henry II there is little reason to believe that Boniface was a strong supporter of the German emperors.

Boniface’s alleged position as a supporter of Conrad in 1026 must therefore be questioned. The main evidence for this loyalty is Boniface’s installation as duke of Tuscany. However, Boniface’s supposed acquisition of the duchy in 1027 is problematic. A document of Jacob, bishop of Fiesole, produced on 16 March 1032 is the earliest surviving reference to Boniface as duke and margrave of Tuscany (dux et marchio Tusciae). [20] However, this new title could not have been acquired by Boniface in 1027. The dispositio of a document of the bishop of Florence, Lambert, produced in July 1028 prays for the salvation and redemption of the souls and the health and safety of the dukes or margraves of Tuscany and for the safety of the margrave Boniface (pro salute et remedio animarum ducum seu marchionum Tusciae et pro salute et incolumitate clarissimi marchionis Bonifatii) [Falce 1927, II.7, 28]. As Anton notes, this document clearly distinguished between the dukes or margraves of Tuscany on one hand and Boniface on the other, therefore Boniface could not have received the duchy of Tuscany by this point [Anton 1972, 537]. This means Boniface began using the title after Conrad had left Italy (Conrad appears in Brixen on 31 May 1027 (DD K2; 101)) and implies that Rainer was not deposed in 1027 but remained ruler of Tuscany until July 1028 at least.

Rainer’s retention of his position in spite of his rebellion is consistent with Conrad’s broader behaviour towards his former opponents. Acts of reconciliation in the form of rituals of clementia and iustitia formed a core part of Conrad’s strategy of rule: by embracing and formalising these Carolingian and Ottonian rituals Conrad was able to present himself as the rightful heir to these earlier emperors [Althoff 2008, 77-79]. This strategy is evident in Conrad’s willingness to restore his opponents to imperial favour even after extended periods of revolt. Despite his repeated attempts to secure a French king of Italy, Ulric Manfred was not directly punished and by the end of Conrad’s life he was closely tied to the emperor [Bertolini 1971, 103]. Pavia ultimately reconciled with Conrad despite the city’s lengthy revolt [Wolfram 2006, 102]. Conrad welcomed Boso and Wido, the sons of Arduin of Ivrea, Henry II’s great opponent for Italy, back into the imperial fold through a charter confirming their rights in 1026 (DD K2; 67). North of the Alps this forgiveness and reconciliation was also common practice. For example, Conrad’s main rival for the German throne, Conrad the Younger, unsuccessfully intrigued to overthrow Conrad II [Wolfram 2006, 73-74]. He retained his titles and was later granted the duchy of Carinthia [Wolfram 2006, 75-76]. It is therefore completely in keeping with Conrad’s method of rule for Rainer to retain his position.

If, as suggested above, Boniface did not receive Tuscany in 1027, there is no reason to assume he was a particularly active supporter of the emperor. None of the narrative sources identify Boniface working in concert with Conrad at this point – even Donizone is silent. Meanwhile Arnulf and Wipo placed Aribert, the archbishop of Milan, at the head of the pro-German contingent but made no mention of Boniface. [21] Boniface’s absence from Arnulf’s account is particularly noteworthy. Arnulf’s desire to strengthen the connection between Milan and Canossa and his subsequent references to Boniface alongside the archbishop of the city, it is surprising for him to miss an opportunity to praise Boniface here. These accounts are supported by a charter produced on 23 March 1026 by Conrad in Milan which confirmed the foundation of a monastery in Milan by Aribert and demonstrates an immediate link between the two (DD K2; 58). Aribert’s support of Conrad is clearly visible, while Boniface cannot be seen at all.

Although Boniface cannot be placed firmly among Conrad’s supporters, he can be connected to figures who acted against the emperor. From the text of Ulrich Manfred’s correspondence with William of Aquitaine, it is clear that Manfred was the ringleader of a group of Italian magnates: in a letter of mid 1025 conveying his desire to abandon the expedition, William refers to Manfred as “the most illustrious Margrave”(M[aginfrido] marchioni clarissimo). [22] the identity of the other members of this group is less clear. The Otbertenghi counts of Luni were connected to Manfred through marriage, his wife Bertha was the daughter of Otbert II [Previté-Orton 2013, 166], and the Otbertenghi were subsequently connected to the rebellion of Pavia [Wolfram 2006, 99] so their involvement here is likely. Boniface held a pair of close family connections with Manfred: Boniface’s aunt, Prangarda, was Manfred’s mother [Bertolini 1971, 98]; Boniface’s first wife Richilde was the daughter of Manfred’s sister [Bertolini 1971, 97]. Boniface had also campaigned alongside the Otbertenghi towards the end of the reign of Henry II and was connected to them further through his marriage to Richilde [Bertolini 1971, 97-98]. While this is certainly not enough evidence to demonstrate Bonfiace’s direct opposition to the emperor it does highlight the complexity of the Italian relationship networks and illustrates that his connection with Conrad is less certain than is usually assumed.

The imperial charters produced around the time of this expedition provide further insight into Conrad’s actions and goals. A large proportion of these documents adjusted the balance of power in Italy in order to counter the strength of Conrad’s opponents. An early example of this is a document issued on 10 June 1025 which granted extensive lands and rights to the bishop’s Church of Novara including control over the nunneries of S. Salvatore and S. Felix in Pavia (DD K2; 39). This served as a statement of Conrad’s authority within Pavia and formed part of his broader conflict with the city. Further, the bishop of Novara had been involved in a recent conflict with Ulrich Manfred of Turin [Bertolini 1971, 98]. By emphasising the imperial connection with the bishopric and empowering this institution, Conrad sought to balance the power of Ulrich Manfred in the region.

There are several further examples of Conrad using his charters to counter threats to his power between 1025 and 1028. Most notably, Conrad produced an unusually large number of charters for institutions and individuals in Tuscany including monasteries in Lucca (DD K2; 25, 55, 76), Capolona (DD K2; 63), Amiata, (DD K2; 79) and Perugia (DD K2; 85), the cathedral chapter at Arezzo (DD K2; 74), a landholder near Pisa (DD K2; 77), and the bishops of Fiesole and Lucca (DD K2; 78, 83). This surge in production was connected to the conflict between Conrad and Rainer of Tuscany: even before Rainer barred the gates of Lucca to Conrad, Conrad attempted to undermine his vassals position by empowering other figures in the area and stating the imperial right to intervene. Conrad likewise reaffirmed links with a series of religious institutions within Pavia in 1026 and 1027 as a counter to the rebellion of the city (DD K2; 59, 63, 75). Against Ulrich Manfred, the monastery at Fruttuaria, a traditional and powerful balance to secular power in the region, received confirmation of its rights (DD K2; 70, 88), and bishop Leo of Vercelli, had the rights of his cathedral confirmed (DD K2; 84). Throughout Italy, but especially in Tuscany, Conrad took steps to assert his authority and undermine that of the figures who had opposed him.

However, Rainer, Ulrich Manfred and the city of Pavia were not the only targets of Conrad’s statements of power and reorganisation of political networks. Conrad also used his charters to emphasise his right and ability to intervene across Canossan territory. A key example of this is Conrad’s empowerment of the bishops of Reggio and Modena. On 1 May 1027 Conrad conferred the role of legate (missus) within a radius of four miles of Reggio on bishop Teuzo (DD K2; 89). This was a jurisdictionally important and prestigious role: Teuzo was empowered to act with the authority of the emperor in legal disputes within this area. The earlier confirmation of the rights and protections of the bishop’s church in Modena on 19 June 1026 (DD K2; 65) was primarily a reiteration of older rights, but did include a pair of innovations. Firstly, this document provided a new list of the property of the church which was now under the emperor’s protection. This list included chapels, lands and other possessions across Canossan territory. Not only was this document a statement of Conrad’s right to intervene in one of Bonfiace’s key holdings, but it underlined his ability to do so across Canossan lands. Secondly, the charter extended the bishop of Modena’s rights of jurisdiction to a three mile radius of the city. Both of these documents represent major statements of imperial authority at the heart of Canossan territory and mirror the strategy used against the Canossans by Henry III and Henry IV. [23] In both of these documents, Conrad asserted control of public rights over and above those held or assumed by Boniface.

Beyond the core Canossan lands, Conrad issued a number of charters to bishops within Boniface’s sphere of influence. Conrad confirmed and extended the lands and rights of the bishop of Bergamo (DD K2; 56, 90) and the canons of h. Vincenzo in the same city (DD K2; 60). He likewise supported the monastery of Leno (DD K2; 100), located to the South-West of Brescia, an area of Canossan expansion. Conrad’s charters of 24 May 1027 confirming the rights and property of the monastery of S. Zeno in Verona (DD K2; 95) and the cathedral chapter of the same city (DD K2; 96) displayed his ability to act in another area of Canossan interest. In a similar manner, his grant of lands in Verona and Ferrara to bishop John of Verona on 8 June 1031 (DD K2; 167) was produced as a statement against Boniface’s expansion in these areas. This is particularly notable as John and his family (including his brother, the count of Verona, and father, the count of Garda) had held a strong connection with Henry II and their empowerment by Conrad’s predecessor formed part of an earlier imperial effort to balance Boniface’s growing influence [Houghton 2015, 404].

Conrad’s early charters also demonstrate his attempts to empower some of Boniface’s main rivals. This included the bishop of Parma who had his rights confirmed and extended in a pair of charters in 1027 (DD K2; 98, 99) and the archbishop of Ravenna who received a confirmation of his rights and property in 1028 (DD K2; 119). This pair were among the most powerful bishops in Italy, having received extensive lands and rights from previous emperors. They also represented the main obstacles to Canossan expansion to the east and west respectively. In 1029 Conrad issued a further pair of charters to the bishop of Parma. The first of these, produced on 12 June 1029, simply confirmed the existing rights of the bishop’s church, including those granted in 1027 (DD K2; 142) demonstrating the ongoing connection between the bishop of Parma and the emperor. More significantly, the second document, issued on 31 December 1029, agreed that the bishop of Parma would receive the comital jurisdiction (comitatus) throughout the diocese on the death (without male heir) of Bernard, the current count (DD K2; 144). This was a major and unprecedented concession to the bishop and represents an attempt by the emperor to produce a figure capable of withstanding Canossan expansion. The goal of creating a prominent power in Parma is underlined by the installation of Hugh, a former imperial chancellor, as bishop of the city in 1027 [Pelicelli 1936, 106-130; Santifaller 1964, 27-43; Schumann 1973, 97]. Conrad empowered a potential rival to Boniface while tying this figure and his office more closely to the imperial court.

Boniface cannot be identified as a clear supporter of Conrad II during the period 1026-1032. None of the narrative sources identify his involvement in Conrad’s campaigns and there is no evidence that Boniface was given control of Tuscany at this point. Instead, Boniface can be tied to Conrad’s opponents through marriage connections and previous alliances. Moreover, the charter record demonstrates the careful construction of a complex imperial relationship network designed to counter and restrict Canossan expansion. Within his charters, Conrad’s treatment of Boniface was most similar to his treatment of Rainer of Tuscany: in both cases although the emperor made no overt moves against his vassals, he did nothing to favour them and much to counter their interests. Although there is no sign that Boniface actively participated in rebellion or intrigue against Conrad, it is clear that Boniface’s position within the imperial relationship network was not exceptional: he showed no unusual loyalty to Conrad, was not a particularly favoured vassal and his lord took active steps to undermine Boniface’s position.

Conrad’s Burgundian Campaign (1032-1034)

Conrad’s conquest of Burgundy between 1032 and 1034 was the culmination of long standing diplomatic efforts by the German emperors which had secured the agreement of Rudolf III of Burgundy that on his death the kingdom would pass to Conrad [Wolfram 2006, 239]. However, when Rudolf died in September 1032 Odo of Blois invaded the kingdom to press his own claim. Conrad, after concluding his campaign in Poland, entered Burgundy with German and Italian forces, defeated Odo and incorporated Burgundy into his empire. This expansion of the empire was significant for Italian politics as it secured new routes into the kingdom from Germany and represented the first time that Italian troops were used en masse in an imperial campaign outside the peninsula.

Donizone presents Boniface in a leading role in this campaign. [24] Although the poet incorrectly places the expedition in the reign of Henry III, [25] he associates Boniface with the siege of the last stronghold of Odo of Blois at Miroaltum. [26] Arnulf provides only a brief account, but is equally enthused by Boniface’s role. He identifies Aribert of Milan and Boniface as the two leaders of the Italian forces, describing them as “two lights of the kingdom” leading the other Italian magnates through the pass at Bard in the Aosta valley: «Vicino autem Italie cum optimatibus ceteris electi duces incedunt presul Heribertus et egregius marchio Bonifacius, duo lumina regni, explorantes accessus illos, quos reddunt meabiles precisa saxa inexpugnabilis opidi Bardi». [27] These two accounts form the basis for the presentation of Boniface as a leading participant in this campaign, a narrative which is often supported through Boniface’s installation as duke of Tuscany by 1032. Superficially Boniface appears to emerge as a leading Italian supporter of Conrad early in the 1030s and this loyalty seems to be associated with the allocation of an important title by his lord.

However, Wipo provides a rather different account of the leadership of the Italian contingent. He places “the archbishop of Milan, Aribert, and the other Italians under the leadership of Count Humbert of Burgundy” archiepiscopus Mediolanensis Heribertus et cateri Italici ductu Huperti comitis de Burgundia [Wipo 1978, chap. 32]. In contrast with Arnulf’s account, Boniface is not mentioned at all and Aribert is reduced to a position below that of Humbert. Wipo, with his connections to Burgundy and the imperial court [Mommsen, Morrison 1962, 42], was better placed than either Arnulf or Donizone to describe this campaign. Wipo had his own agenda, which did not include the aggrandisement of Boniface, but his account suggests that Boniface did not occupy a particularly prominent position in the Italian army. At the least, Boniface’s absence from the Gesta highlights the fact that his position was not viewed as extraordinary within the German court: Conrad did not see his relationship with Boniface as special.

Furthermore, Arnulf’s reference to the optimates indicates the presence of other Italian magnates on this expedition, identified by Bertolini as the lords of Esarcato, Gebeardo of Ravenna and Ugo count of Bologna [Bertolini 1971, 101]. Given his powerbase in the north-west of Italy and his improving relationship with the emperor, it is probable that Ulrich Manfred of Turin was also present on this campaign [Bertolini 1971, 101]. Arnulf’s promotion of Boniface fits with his desire to improve Milan’s relationship with Canossa and this goal also explains the references to Boniface as “outstanding” (egregious). [28] The reference to Boniface as a light of the kingdom was likewise used to flatter his heirs. The phrase should not be taken to indicate his fidelity towards the emperor as it was used to describe not only Boniface but also Aribert who became Conrad’s main opponent in northern Italy in the last years of his life. Arnulf inflated the role of both Aribert and Boniface in order to further his own goals.

While the grant of Tuscany to Boniface does demonstrate a connection with the emperor, the significance of this appointment should not be overstated. There is no indication that Rainer, Boniface’s predecessor, enjoyed any connection with Henry II beyond his receipt of the duchy. His promotion was simply part of broader imperial attempts to balance the power structures of the region. Furthermore, as demonstrated above, Conrad’s charters during his first expedition into Italy had done much to decay the authority of the duke of Tuscany. Numerous institutions had been empowered and placed beyond the jurisdiction of the duke. The title was still important, but Boniface and his successors would expend a great deal of effort attempting to reclaim the ducal rights which had been dispersed during Conrad’s reign: in combination with the family’s relatively small allodial possessions in Tuscany [Fasoli 1978, 65], this factor contributed towards Matilda’s difficulties in controlling the region [Fasoli 1978, 66; Golinelli 2001, 515]. In the absence of other convincing evidence for Conrad’s support of Boniface, the Canossan acquisition of the duchy of Tuscany should be seen as part of the broader imperial strategy of balancing power and asserting authority in Italy rather than evidence for a unique bond as described by Donizone.

Further doubts about Boniface’s personal connection with Conrad during the Burgundy campaign are raised through a consideration of the imperial charters. A pair of documents suggest that Conrad continued his programme of balancing Boniface’s power during his campaign in Burgundy. On 30 April 1034, Conrad granted the archbishop of Ravenna the comitatus in Faenza (DD K2; 209). This strengthened the position of the archbishop and countered Canossan expansion east along the Via Emilia. This was followed on 6 June 1034 by the confirmation of the foundation of a hospice at a monastery in Florence by its Abbot Peter (DD K2; 210). The document shows a continued relationship between the Abbot of this monastery and the emperor, the continued empowerment and protection of the institution by the emperor, and the exclusion of Boniface from this relationship. The monastery was a fairly powerful entity in a strategic position within Boniface’s domain and so its close and ongoing relationship with the emperor represents a balance to Canossan power.

This period in the middle of Conrad’s rule may represent a strengthening of his relations with Boniface, but this is far from certain. Boniface was probably involved in the Burgundian campaign: the inclusion of German and Italian elements in the imperial army suggests one of the largest mobilisations of Conrad’s reign [Bertolini 1971, 101], and Boniface’s absence from the host as a prominent, if not necessarily particularly favoured, magnate would be unusual. However, his position within the army was almost certainly inflated by both Donizone and Arnulf for their own ends. Conrad’s charters suggest a continuation of his earlier policy of balancing Boniface’s power by building relationships with figures who could oppose the margrave and strengthening the political, military and jurisdictional position of these individuals. The evidence for this period is less conclusive than that for Conrad’s first expedition into Italy, but on balance it seems that the situation did not change significantly during this campaign. Boniface was still a powerful and important figure in Italy and was now integrated into the imperial relationship network, but he did not enjoy a position of particular favour with the emperor and the emperor continued to take steps to counter Boniface’s power. This is not to say that Conrad was overtly hostile towards Boniface, but this more nuanced relationship is very different from the vision presented by Donizone. Even as he expanded Boniface’s rights in Tuscany, Conrad sought to maintain a balanced relationship network in Italy, not to promote a single figure to a position of dominance.

Conrad’s Second Italian Expedition (1035-1039)

In the last years of Conrad’s life the political balance in northern Italy changed dramatically. Aribert of Milan, a dominant power in the region and the traditional supporter of Conrad, became the main opponent of the crown [Anton 1972, 541; Reynolds 1997, 199; Salvatorelli 1940, 21]. This conflict caused profound political shifts, including the formal confirmation of the rights of the valvassores across Italy through a charter of Conrad on 28 May 1037 known commonly as the Constitutio de Feudis (DD K2; 244). Beyond the events in Milan, the emperor was involved in other conflicts including a riot in Parma at Christmas in 1037 and a brief campaign into the south of the peninsula. It is no coincidence that this is the point at which Boniface appears closest to the emperor not only in Donizone’s narrative but also in the charters: the crisis with Aribert led Conrad to rebalance the structures of power in Italy extensively and Boniface, alongside several other individuals and institutions, benefited from this reorganisation [Anton 1972, 542-543; Salvatorelli 1940, 21]. It was only at this late point in Conrad’s reign that strong evidence suggesting a prominent role for Boniface in the imperial relationship network finally emerged. Nevertheless, there are several indications that this relationship was still not as close or exceptional as Donizone and those who follow his narrative imply.

Donizone devoted a chapter to Boniface’s involvement in Conrad’s second Italian expedition. This section of the poet’s work focuses on the riot in Parma at the end of 1037 while Conrad celebrated Christmas in the city. After detailing the arrival of Conrad and the start of the uprising, [29] Donizone brings Boniface into his narrative:

The madness was purified through the wisdom of Caesar,
He immediately sent his armed and helmeted guards to the walls,
He ordered our experienced lord Boniface,
As far as he could, to break the foolish city. [30]

Boniface, here distinguished as the friend of the king, is then described assaulting the city:

Having swiftly gathered his squadrons of cavalry and infantry,
The friend of the king hurried to the ruin of the city. [31]

The approach of Boniface terrified the rioters, who submitted to Conrad:

Having seen this, the citizens were afraid and reflected on their destruction;
They soon approached the feet of the king,
And together they opened up the gate to the city. [32]

Boniface is presented as the leader of the army, the friend and saviour of the emperor, and a vassal of special importance.

The charter evidence, for once, broadly supports Donizone’s account of a close relationship between Conrad and Boniface. Shortly before Conrad’s entrance into Italy Boniface made his first definite appearance in the imperial charters. On 5 July 1036 (DD K2; 231), Boniface is listed alongside the chancellor, Pilgrim, and Hermann the archbishop of Cologne as a petitioner requesting that Conrad take the monastery of S. Sisto in Piacenza under his protection. Boniface is described here as ‘our beloved margrave” (nostri diecti marchionis) claiming a close bond to the emperor.

This connection was continued after Conrad entered Italy. On 10 July 1037 Boniface again appeared as a petitioner for a charter requesting the confirmation of the rights of the cathedral chapter of Florence (DD K2; 246), this time alongside Conrad’s wife Gisela. Boniface is again presented in glowing terms, this time as “our most faithful margrave” (nostri fidelissimi marchionis). Then, on 22 February 1038, Boniface appeared in two court sessions both held in his casa at Uiuinaria (DD K2; 258, 259) both dealing with the church of Lucca. These two documents present Boniface in the role of a loyal vassal and position him in close proximity to the emperor. The documents go out of their way to highlight Boniface’s role as host for these sessions. This role for Boniface is particularly notable as Conrad typically made use of ecclesiastical possessions for his adjudications and on his itinerary. In combination, these factors demonstrate a very visible display of the connection between Boniface and Conrad.

Boniface’s sudden appearance and prominence within Conrad’s charters coincided with two key developments in his connection to the emperor. Firstly, following the death of Richilde by the end of February 1036, Boniface married Beatrice of Lorraine, a member of a family with close ties to the emperor. This was one of a series of marriages linking prominent Italian magnates with German families close to the emperor undertaken with Conrad’s approval if not support [Anton 1972, 543; Bertolini 1971, 103; Fumagalli 1978, 31]. Secondly, Boniface was present at the wedding in Nijmegen of Cornad’s son Henry to Gunhilda of Denmark in June 1036 [Anton 1972, 540]. This is significant as it demonstrated a link between Conrad and Boniface which went beyond Italian affairs. These two events fit the broader pattern of closer ties between Conrad and Boniface described in the Vita and the charters.

However, even at this point of crisis in Italy Boniface’s proximity to Conrad should not be overstated. Nobili has convincingly suggested that Donizone’s account of Boniface’s rescue of Conrad was dictated by the poet’s integral presentation of the Canossans as loyal vassals, questioning whether Boniface was involved in the incident in the way Donizone suggests [Nobili 1978, 272]. Boniface’s role in the riot at Parma is mentioned only by Donizone and the sources which follow him. Wipo, despite a vivid account of the event, does not indicate that Boniface was present [Wipo 1978, chap. 38]. Arnulf is completely silent with regards to the event, maintaining a focus on events in Milan. The other sources to mention the riot, the Annales Hildesheimenses [1839], the Annales Augustani [1839], and the Annales Parmenses minores [1863] also emphasise the imperial role and do not mention Boniface. While Boniface’s absence may be explained by a lack of interest in Italian affairs on the part of the German annalists and chronological distance in the case of the Parmese chroniclers, Donizone’s version of events is nevertheless isolated. Even if Boniface was present at the riot, Donizone surely elaborated his role.

Donizone’s own account of the riot raises some questions regarding the position of Boniface in relation to the emperor. Boniface is placed with the imperial host, hence camped outside the city. This means he was absent from the feast and celebrations themselves. Donizone spins Boniface’s guard duty into an honourable and valiant position, but the fact remains that this is not where a loyal and prominent supporter of the emperor would be expected to be found. Court sessions, feasts and other public events provided the opportunity for the maintenance and development of relationships between lords and vassals, but this relied on the physical presence of these vassals [Nelson 1987, 166-172; Althoff 1993, 27-28; Arnold 1997, 170]. Moreover, the riot in Parma was not the pivotal event of this expedition. Boniface’s participation in the siege of Milan or campaign in the South is not recorded in the Vita, any of the other narrative sources, or in the charters. Boniface’s apparent absence from these central elements of Conrad’s expedition illustrates the limits of the connection between the two.

That Conrad chose to celebrate Christmas in Parma is also telling. As demonstrated above, Conrad repeatedly empowered Hugh, bishop of Parma, and Hugh remained a key balance to Boniface in the late 1030s. By spending Christmas, one of the key festivals of the Christian calendar, in Parma Conrad made a very visible statement of his support for the bishop of the city. This was designed to cement the bishop’s position against the archbishop of Milan, but also underlines Conrad’s ongoing support for one of Boniface’s traditional rivals.

Although Boniface’s appearances in Conrad’s charters in these years demonstrate the development of a closer relationship, these documents also show that Conrad continued to empower groups and individuals within and around Canossan territory as a counter to Boniface. On 15 February 1036 Conrad reiterated the transfer of the comitatus to the bishop of Parma (DD K2; 226), highlighting his continued support for this figure. This was followed by confirmations of the rights of the monastery at Leno (DD K2; 227), monastery of S Zeno in Verona (DD K2; 247), bishop’s church in Brescia (DD K2; 248), cathedral chapter of Pistoia (DD K2; 256), cathedral chapter of Arezzo (DD K2; 263) and monastery of S Nazaro and Celso near Verona (DD K2; 274). All of these institutions were located in areas of Canossan landholding, many of them had received charters from Conrad earlier in his reign and Boniface is not mentioned in any of these documents. While Boniface was now positioned closer to the imperial court, Conrad still sought to keep Canossan power balanced.

This imperial intervention was not restricted to the peripheral areas of Boniface’s holdings. On 31 March 1037 Conrad produced two charters for the benefit of bishop Hildolf of Mantua. The first of these (DD K2; 235) confirmed the bishop’s existing rights, but went into greater detail than previous documents in specifying the property of the Church which was under imperial protection. The second (DD K2; 236) reiterated the imperial protection for the bishop and his church and extended this protection to the abbot of S Ruffino in Mantua. Through these two documents Conrad again underlined his right and ability to intervene in the very heart of Canossan territory [Gardoni 2006, 221-224]. Mantua was the most important of the Canossan cities and most central to their power [Fumagalli 1987; Gardoni 2006; Marani 1987]. Conrad’s intervention here demonstrates that even as Boniface began to play a more central role in imperial rule in Italy, he was by no means afforded unchecked power within his core lands.

In the last four or five years of Conrad’s life, and in sharp contrast with the rest of his rule, there is an undeniable political connection between the emperor and Canossa. Boniface appeared frequently and prominently within imperial charters. He travelled to Conrad’s court in Germany. He married into a powerful German family with strong ties to the emperor. Conrad held court within Boniface’s palace. However, even at this zenith, Boniface’s position within the relationship networks of northern Italy and the empire more generally were complex. Conrad continued to build connections with groups and individuals within and around Canossan territory who could counter Boniface and who had traditionally opposed him. Boniface was empowered to a certain extent and started to play a more prominent and visible role within the imperial relationship network, but this is far from the unilateral image of designated responsibility presented in the Vita Mathildis. Donizone exaggerated Boniface’s involvement in Conrad’s expedition and the participation of the margrave in key sections of the campaign cannot be determined. Boniface was certainly of importance to the emperor in this period, and the glowing portrayal of him in the charters demonstrates a desire to at least present a close relationship on the part of the imperial court, but his prominence should not be overstated.

Conclusion

Boniface’s position within the imperial relationship network in Italy during the reign of Conrad II has been overstated. Before 1032 there is no evidence of any remarkable connection between the two, much less a unique and strong relationship of the type described by Donizone. By the time of the Burgundian campaign a link may have developed, but there is little evidence that Boniface enjoyed a particularly unusual relationship with the emperor at this point. It was only during the preparations for Conrad’s second Italian campaign that Boniface emerged as an important figure with close ties to the emperor, and this was only ever in the context of a carefully balanced network of relationships. Our understanding of this connection has been oversimplified by a reliance on Donizone’s work, but by consulting the other narrative sources and the charter record it is possible to identify a much more nuanced relationship. Boniface was simply one element in the complex system of political connections maintained by the emperor. He was not, and was never intended to be, an imperial viceroy in Italy.

Boniface’s sudden prominence at the end of Conrad’s life was prompted not by a special relationship between the two, but by the shift in the political environment in Italy in the 1030s. Aribert of Milan’s fall from favour was the most visible of these changes. Aribert’s power and his traditional support for Conrad left the emperor with the need for supporters in the region in order to bring his unruly archbishop to heel. To this end the emperor empowered several bishops and the valvassores. However, the death of Ulrich Manfred in 1034 left Boniface as the only powerful secular magnate who could support Conrad. Boniface’s marriage to Beatrice of Lorraine in 1036 may represent an early start in this shift – Conrad appears to have been preparing to move against Aribert prior to his second Italian expedition. In any event, it was these changing circumstances which led to Boniface’s empowerment, not a dramatic show of loyalty in the form of a rescue at Parma.

This alternative reading, combined with previous analysis on the relationship between Boniface and Henry II [Houghton 2015], and long standing studies into the connection between Boniface and Henry III, [33] demonstrates that Boniface was only close to the imperial court for a rather brief period at the end of the life of Conrad II. Boniface’s overt support for the German emperors was therefore the exception rather than the rule. This is in direct opposition to the narrative presented by Donizone, who went to great lengths to present Boniface as a consistently loyal and prominent vassal of Henry II, Conrad II and Henry III in order to portray Matilda’s conflict with Henry IV as an isolated incident completely justified by the failures and illegal actions of the emperor.

This reassessment also raises questions about when and how Boniface became margrave or Duke of Tuscany. Boniface appears with the title duke in 1032, but is only acknowledged in the imperial charters in 1036. Even then, in Conrad’s documents Boniface only ever appeared as margrave and was never associated with Tuscany. The issue needs more detailed consideration, but the transfer of the title was certainly more complex than is usually allowed. The nature of this transfer is of particular importance for the study of the control of Tuscany by the later Canossans. If this grant was not widely acknowledged it could explain the difficulty Matilda had in claiming the ducal rights in the region. Further, it would partly explain the apparent refusal by Henry V to confirm Matilda in these rights, despite their reconciliation in 1111 and the restoration of her authority in Lombardy [Golinelli 1994, 2001, 517-519].

More generally, this reconsideration underlines the complexities of the political networks of northern Italy and Europe as a whole. It is insufficient to present individuals and groups as loyal or rebellious vassals. Boniface is a striking example of a figure immersed in a complex and changing political relationship network whose allegiance cannot be simplified in this manner. Rainer of Tuscany receives only token consideration in most accounts of this period. He is presented most typically as a rebellious foil for Boniface’s loyalty to the Emperor. However, the charter evidence and Conrad’s typical manner of rule suggest a briefly rebellious figure returned to the imperial fold with only typical sanction. Ulrich Manfred of Turin receives more attention, but is likewise almost universally presented as a rebellious contrast to Boniface. His family ties with Boniface and his reconciliation with Conrad and subsequent empowerment by the emperor are generally ignored. Even Aribert of Milan whose complex and changing connection with the imperial court has been considered in some depth is nevertheless often presented in an oversimplified manner. In particular, the archbishop’s connection with Boniface before and during his conflict with Conrad needs further consideration. It is possible to gain a better understanding of their goals and motivations through a closer investigation of these systems.

Finally, this analysis highlights the necessity of moderating chronicles and other narrative sources with charter evidence. Donizone, Arnulf and Wipo all present very clean narratives, simplifying, modifying, ignoring and fabricating events to produce an account which supported their goals. Accounts such as these are very attractive as they present a graphic and detailed description of their subjects. However, these accounts are almost invariably incomplete and overreliance on them can lead to critical misunderstandings of the political and social systems in which they wrote. While charters are also vulnerable to distortion by the goals of their authors and patrons, they provide an invaluable balance to these more dramatic sources.

Abbreviations

• “DD K2” = Die Urkunden Konrads II, Mit Nachträgen zu den Urkunden Heinrichs II, Breslau H. (ed.), Monumenta Germaniae Historica, Diplomata, 4, Hanover, 1909.

Primary Sources

• Annales Augustani, Pertz G. H. (ed.), in Annales, chronica et historiae aevi Saxonici, Monumenta Germaniae Historica, Scriptores (in Folio) 3, Hannover, 1839, 124-136.
• Annales Hildesheimenses, Pertz G. H. (ed.), in Annales, chronica et historiae aevi Saxonici, Monumenta Germaniae Historica, Scriptores (in Folio) 3, Hannover, 1839, 90-115.
• Annales Parmenses minores, Pertz G. H. (ed.), in Monumenta Germaniae Historica, Scriptores (in Folio) 18, Stuttgart, 1863, 662-663.
• Arnulf, Liber Gestorum Recentium, Zey C. (ed.), Monumenta Germaniae Historica, Scriptores Rerum Germanicarum in Usum Scholarum Separatim Editi 67, Hannover: Hahnsche Buchhandlung, 1994.
• Donizone, Vita Mathildis, Golinelli P., Fumagalli V. (eds.), Biblioteca di cultura medievale 823, Milano: Jaca book, 2008.
• Fulbert, The Letters and Poems of Fulbert of Chartres, Behrends F. (ed.), Oxford Medieval Texts, Oxford: Clarendon Press, 1976.
• Wipo 1978, Gesta Chuonradi II Imperatoris, in Trillmich W., Büchner R. (eds.) 1978, Fontes saeculorum noni et undecimi historiam ecclesiae Hammaburgensis necnon imperii illustrantes, Darmstadt: Wissenschaftliche Buchgesellschaft, 507-615.

Bibliography

• Althoff G. 1993, Demonstration und Inszenierung: Spielregeln der Kommunikation in mittelalterlicher Öffentlichkeit, «Frühmittelalterliche Studien», 27: 27-50.
• — 2008, The Variability of Rituals in the Middle Ages, in Althoff G., Fried J., Geary P. J. (eds.) 2008, Medieval Concepts of the Past: Ritual, Memory, Historiography, Cambridge: Cambridge University Press, 71-88.
• Anton H. H. 1972, Bonifaz von Canossa, Markgraf von Tuszien, und die Italienpolitik der frühen Salier, «Historische Zeitschrift», 24 (3): 529-556.
• Arnaldi G. 1962, Arduino, re d’Italia, in Dizionario biografico degli Italiani, 4, Roma: Istituto della Enciclopedia italiana, 53-60.
• Arnold B. 1997, Medieval Germany, 500-1300: A Political Interpretation, Toronto-Buffalo: University of Toronto Press.
• Bartoli Langeli A. 2002, Private Charters, in La Rocca C. (ed.) 2002, Italy in the Early Middle Ages: 476-1000, The Short Oxford History of Italy, Oxford-New York: Oxford University Press.
• Bernhardt J. W. 1993, Itinerant Kingship and Royal Monasteries in Early Medieval Germany: C. 936-1075, Cambridge Studies in Medieval Life and Thought, 4th ser., 21, Cambridge: Cambridge University Press.
• Bertolini M. G. 1971. Bonifacio, marchese e duca di Toscana, in Dizionario biografico degli Italiani, 12, Roma: Istituto della Enciclopedia italiana, 96-113.
• — 1981, Note di genealogia e di storia canossiana, in I ceti dirigenti in Toscana nell’età precomunale, Atti del 1o Convegno di studi sulla storia dei ceti dirigenti in Toscana, Firenze 2 dicembre 1978, Pisa: Pacini Editore, 111-149.
• Bresslau H. 1967, Jahrbücher des deutschen Reiches unter Kaiser Konrad II, Berlin: De Guyter.
• Castagnetti A. 1985, Società e politica a Ferrara dall’età postcarolingia alla signoria estense (sec. X-XIII), Rome: Pàtron.
• Colorni V. 1959, Il territorio mantovano nel Sacro Romano Impero. I, periodo comitale e periodo comunale (800-1274), Milano: Giuffré.
• Cowdrey H. E. J. 1968, The Papacy, the Patarenes and the Church of Milan, Transactions of the Royal Historical Society 18: 25.
• Falce A. 1927, Bonifacio di Canossa, padre di Matilda, Reggio-Emilia: Bizzocchi.
• Fasola L. 1980, Arnulf von Mailand, Chronist, in Lexikon des Mittelalters, 1:1020, München: LexMA.
• Fasoli G. 1971, Rileggendo la Vita Mathildis di Donizone, in Bertuzzi G. (ed.) 1971, Studi matildici, Atti e Memorie del II Convegno di Studi matildici (Reggio Emilia, 1-3 maggio 1970), Modena: Aedes Muratoriana, 15-39.
• — 1978, La realtà cittadina nei territori canossiani, in Bertuzzi G. (ed.) 1979, Studi matildici, Atti e Memorie del III Convegno di Studi matildici (Reggio Emilia, 7-9 ottobre 1977), Modena: Aedes Muratoriana, 55-78.
• Fliche A. 1912, Arnoul, chronographe milanais (XIe siècle), in Dictionnaire d’histoire et de géographie ecclésiastiques, in Encyclopédie des sciences ecclésiastiques 4, Paris: Letouzey et Ané.
• Fumagalli V. 1971, Le origini di una grande dinastia feudale Adalberto-Atto di Canossa, Tübingen: M. Niemeyer.
• — 1976, Terra e società nell’Italia padana: i secoli IX e X, Torino: Giulio Einaudi.
• — 1978, I Canossa tra realtà regionale e ambizioni europee, in Studi matildici, Atti e Memorie del III Convegno di Studi matildici (Reggio Emilia, 7-9 ottobre 1977), Modena: Aedes Muratoriana, 27-37.
• — 1979, Il potere civile dei vescovi italiani al tempo di Ottone I, in Mor C. G., Schmidinger H. (eds.) 1979, I Poteri Temporali dei Vescovi in Italia e in Germania nel Medioevo, Bologna: il Mulino, 77-86.
• — 1987, Mantova al Tempo di Matilde di Canossa, in Golinelli P. (ed.) 1987, Sant’Anselmo, Mantova e la lotta per le Investiture: atti del convegno internazionale di studi, (Mantova, 23-24-25 maggio 1986), Bologna: Pàtron, 159-167.
• — 1994, Discorso di Apertura, in Golinelli P. (ed.) 1994, I Poteri dei Canossa, da Reggio Emilia all’Europa: atti del convegno internazionale di studi (Reggio Emilia-Carpineti, 29-31 ottobre 1992), Bologna: Pàtron, 15-20.
• Gardoni G. 2006, Vescovi e città a Mantova dall’età carolingia al secolo XI, in Andenna G., Brogiolo G. P., Salvarani R. (eds.) 2006, Le origini della diocesi di Mantova e le sedi episcopali dell’Italia settentrionale, IV-XI secolo, Trieste: Editreg, 183-246.
• Goldberg E. J. 2009, Struggle for Empire: Kingship and Conflict under Louis the German, 817-876, Ithaca: Cornell University Press.
• Golinelli P. 1990, Una prerogativa dei Canossa: il “Paparum ducatus”, in Canossa prima di Matilde. Atti del Convegno di Reggio Emilia, Milano: Camunia, 199-214.
• — 1991, Matilde e i Canossa nel cuore del medioevo, Milano: Camunia.
• — 1992, Donizone, in Dizionario biografico degli Italiani, 41, Roma: Istituto della Enciclopedia italiana, 200-203.
• — 1994, Matilda ed Enrico V, in Golinelli P. (ed.) 1994, I Poteri dei Canossa, da Reggio Emilia all’Europa: atti del convegno internazionale di studi (Reggio Emilia-Carpineti, 29-31 ottobre 1992), Il mondo medievale, Bologna: Pàtron, 455-471.
• — 1999, Le origini del mito di Matilde e la fortuna di Donizone, in Golinelli P. (ed.) 1999, Matilde di Canossa nelle culture europee del secondo millennio: dalla storia al mito: atti del convegno internazionale di studi (Reggio Emilia, Canossa, Quattro Castella, 25-27 settembre 1997), Bologna: Pàtron, 29-52.
• — 2001, L’Italia dopo la lotta per le investiture: la questione dell’eredità matildica, «Studi Medievali», 42 (2): 509-528.
• — 2008, Donizone e il suo poema per Matilde, in Golinelli P., Fumagalli V. (eds.) 2008, Vita di Matilde di Canossa, Milano: Jaca Book.
• Gualazzini U. 1933, Per la storia dei rapporti tra Enrico III e Bonifacio di Canossa, «Archivio storico lombardo», 19: 67-83.
• Guyotjeannin O. 2001, Ecrire en chancellerie, in Zimmermann M. (ed.) 2001, Auctor et auctoritas: invention et conformisme dans l’écriture médiévale : actes du colloque tenu à l’Université de Versailles-Saint-Quentin-en-Yvelines, 14-16 juin 1999, Paris: Ecole des chartes, 17-35.
• Haverkamp A. 1988, Medieval Germany, 1056-1273, Oxford: Oxford University Press.
• Houghton R. 2013, Donizo, in Thomas R. F., Ziolkowski J. M. (eds.) 2013, The Virgil Encyclopedia, Chichester: Wiley Blackwell.
• — 2015, Reconsidering Donizone’s Vita Mathildis: Boniface of Canossa and Emperor Henry II, «Journal of Medieval History», 41 (4): 388-408.
• Keller H. 1998, Zu den Siegeln der Karolinger und Ottonen: Urkunden als “Hoheitszeichen” in der Kommunikation des Königs mit seinen Getreuen, «Frühmittelalterliche Studien», 32: 400-441.
• — 2011, The Privilege in the Public Interaction of the Exercise of Power: Forms of Symbolic Communication Beyond the Text, in Mostert M., Barnwell P. S. (eds.) 2011, Medieval Legal Process. Physical, Spoken and Written Performance in the Middle Ages, Turnhout: Brepols, 75-108.
• Koziol G. 1992, Begging Pardon and Favor: Ritual and Political Order in Early Medieval France, Ithaca: Cornell University Press.
• Lazzari T. 2008, Aziende fortificate, castelli e pievi: le basi patrimoniali dei poteri dei Canossa e le loro giurisdizioni, in Calzona A. 2008, Matilde e il tesoro dei Canossa: tra castelli, monasteri e città, Milano: Silvana, 95-115.
• Marani E. 1987, Topografia e urbanistica di Mantova al tempo di Sant’Anselmo, in Golinelli (ed.) 1987, Sant’Anselmo, Mantova e la lotta per le Investiture: atti del convegno internazionale di studi, (Mantova, 23-24-25 maggio 1986), Bologna: Pàtron.
• Mommsen T. E., Morrison K. F (eds.) 1962, Imperial Lives and Letters of the Eleventh Century, New York: Columbia University Press.
• Mor C. G. 1964, Dalla caduta dell’Impero al Comune, Verona: Istituto per gli Studi Storici Veronesi.
• Morselli M. 1992, Le origini di un comune cittadino: analisi di alcuni documenti del liber privilegiorum comunis mantue, Modena: Università degli studi di Modena.
• Nasali Rocca E. 1955, Il patriziato piacentino nella eta del comune e della signoria, in Visconti A. 1955, Scritti storici e giuridici in memoria di Alessandro Visconti, Milan: Ist. ed. Cisalpino, 185-306.
• Nelson J. L. 1987, Carolingian Royal Ritual, in Cannadine D., Price S. (eds.) 1987, Rituals of Royalty: Power and Ceremonial in Traditional Societies, Cambridge: Cambridge University Press, 137-180.
• — 1992, Charles the Bald, London-New York: Longman.
• Nobili M. 1978, L’Ideologia Politica in Donizone, in Studi matildici, Atti e Memorie del III Convegno di Studi matildici (Reggio Emilia, 7-9 ottobre 1977), Modena: Aedes Muratoriana, 263-279.
• Pelicelli N. 1936, I vescovi della chiesa parmense, Parma: Officina Grafica Fresching.
• Previté-Orton C. W. 2013, Early History of the House of Savoy: 1000-1233, Cambridge: Cambridge University Press.
• Reynolds S. 1997, Kingdoms and Communities in Western Europe, 900-1300, Oxford: Clarendon Press (2nd ed.).
• Rinaldi R. 1997, Da Adalberto Atto a Bonifacio. Note e riflessioni per l’edizione di un Codice Diplomatico Canossano prematildico, «Bullettino dell’Istituto Storico Italiano per il Medio Evo e Archivio Muratoriano», 101: 13-91.
• Riversi E. 2013, La memoria di Canossa: saggi di contestualizzazione della Vita Mathildis di Donizone, Pisa: Edizioni ETS.
• Rosenwein B. H. 1996a, Friends and Family, Politics and Privilege in the Kingship of Berengar I, in Kline Cohn S., Epstein S. A. (eds.) 1966, Portraits of Medieval and Renaissance Living: Essays in Memory of David Herlihy, Ann Arbor: University of Michigan Press, 91-106.
• — 1996b, The Family Politics of Berengar I, King of Italy (888-924), «Speculum», 71 (2): 247-289.
• — 1999, Negotiating Space: Power, Restraint, and Privileges of Immunity in Early Medieval Europe, Ithaca, N. Y.: Cornell University Press.
• Salvatorelli L. 1940, L’Italia Comunale dal secolo XI alla metà del Secolo XIV, Milan: Mondadori.
• Santifaller L. 1964, Zur Geschichte Des Ottonisch-Salischen Reichskirchensystems, Vienna: Hermann Böhlaus Nachf. (2nd ed.).
• Schumann R. 1973, Authority and the Commune, Parma 833-1133, Parma: Presso La Deputazione Di Storia Patria Per Le Province Parmensi.
• Sergi G. 1994, I poteri dei Canossa: Poteri delegati, poteri feudali, poteri signorili, in Golinelli P. (ed.) 1994, I Poteri dei Canossa, da Reggio Emilia all’Europa: atti del convegno internazionale di studi (Reggio Emilia-Carpineti, 29-31 ottobre 1992), Bologna: Pàtron, 29-39.
• Simeoni L. 1927, La “Vita Mathildis” e il suo valore storico, Modena: Società tipografica modenese.
• Stock B. 1983, The Implications of Literacy: Written Language and Models of Interpretation in the Eleventh and Twelfth Centuries, Princeton, N. J.: Princeton University Press.
• Tabacco G. 1989, The Struggle for Power in Medieval Italy: Structures of Political Rule, Cambridge: Cambridge University Press.
• Tincani A. 1994, Le corti dei Canossa in area Padana, in Golinelli P. (ed.) 1994, I Poteri dei Canossa, da Reggio Emilia all’Europa: atti del convegno internazionale di studi (Reggio Emilia-Carpineti, 29-31 ottobre 1992), Bologna: Pàtron, 253-278.
• Vasina A. 1990, Tedaldo di Canossa e Ferrara, in Canossa prima di Matilde: atti del Convegno internazionale di studi Canossa prima di Matilde: origine della potenza dei da Canossa, svoltosi a Reggio Emilia il 19-20 giugno 1987, Milano: Camunia, 167-176.
• Vecchi G. 1964, Temi e Momenti di Scuola nella “Vita Mathildis” di Donizone, in Studi Matildici. Atti e memorie del Convegno di Studi Matildici (Modena e Reggio Emilia, 19-20-21 Ottobre 1963), Modena: Aedes Muratoriana, 210-217.
• Violante C. 1952, Aspetti della politica italiana di Enrico III prima della sua discesa in Italia (1039-1046), «Rivista storica italiana», 64: 157-176.
• — 1971, Arnolfo, Proniponte del fratello dell’arcivescovo di Milano Arnolfo (I), in Dizionario biografico degli italiani, Rome: Istituto della Enciclopedia italiana.
• Weinfurter S. 1999, The Salian Century: Main Currents in an Age of Transition, Philadelphia: University of Pennsylvania Press.
• Wolfram H. 2006, Conrad II, 990-1039: Emperor of Three Kingdoms, Kaiser D. A. (transl.), University Park: Pennsylvania State University Press.
• Wood I. N. 1994, The Merovingian Kingdoms, 450-751, London-New York: Longman.
• Zimmermann H. 1990, Canossa e il matrimonio di Adelaide, in Canossa prima di Matilde. Atti del Convegno di Reggio Emilia, Milano: Camunia, 141-155.
• — 1994, I Signori di Canossa e l’Impero (da Ottone I a Enrico III), in Golinelli P. (ed.) 1994, I Poteri dei Canossa, da Reggio Emilia all’Europa: atti del convegno internazionale di studi (Reggio Emilia-Carpineti, 29-31 ottobre 1992), Bologna: Pàtron, 413-419.
• Zumhagen O. 2002, Religiöse Konflikte und kommunale Entwicklung: Mailand, Cremona, Piacenza und Florenz zur Zeit der Pataria, Köln: Böhlau Verlag.