Medieval Identity: A Sign and a Concept

BRIGITTE MIRIAM BEDOS-REZAK

In the two centuries following the turn of the first millennium, literate individuals in Western Europe rarely if ever resorted to mediated expression, to indirect communication by means of the written word, without expressing some sense of the absence of immediacy, that is, of personal presence. When Bishop Arnulf of Lisieux (d. 1181) could not attend a council in London, he sent a letter "so that the page might take the place of his person and the letter might faithfully bring his voice to life."1 Slightly earlier, Bernard of Clairvaux (d. 1153) sought to reassure his correspondents about the authenticity and representativeness of two letters to which he was unable to affix his seal. In one letter, he wrote: "I do not have my seal handy, but the reader will recognize the style because I myself have dictated the letter."2 The other letter states: "May the discursive structure stand for the seal, which I do not have handy."3 Bernard expects readers to notice his personal presence, however immaterial, within the fabric of the text, through its style and diction. His secretary and biographer, Geoffrey of Clairvaux (or of Auxerre, d. after 1188), emphasized this conflation of person and text by entitling Chapter 8 of his biography: "On St. Bernard's writings and the image of his soul expressed in them."4	1
Bernard's and Arnulf's letters reveal two closely related assumptions, that there is a symbiotic relationship between human presence and representation, one in which representation matches real presence, and second that the written text is an embodiment of its author and articulates a notion of authenticity revolving around authority and identity. Additionally, Bernard indicates that there was equivalence between his discourse and his seal, in that both had the capacity to signify his personality. Written texts, to be sure, were major instruments of the literate elite's effectiveness as personalities and public figures,5 but so too was the aura of their physical presence. Bernard and Arnulf lived at a time when it was still possible for them to deploy both media—body and text—equally in matters of authority, even though an irreversible movement had already commenced during the eleventh century that was to shift preeminence from personal to textual presence. Bernard, being literate, could both compose and write in Latin; his authorial identity might thus be vested just as well in his discursive style as in his seal. However, what became of such a form of personal identity if it had to be projected through texts that, produced by others in the names of non-literate individuals, necessarily lacked the authoritative imprint of authorial style and presence? The phenomenon I wish to consider in this essay involves the novel recourse to the written and sealed word by the lay aristocracy of northern France during the eleventh and twelfth centuries. At this time, French nobles were not yet literate; they lacked Latin beyond the modest requirements of liturgy, and as yet neither participated in modes of textual and iconic representation nor controlled the spheres of scribal and iconographic practice. I believe that the process of the French nobility's acculturation to such modes of representation as the sealed charter commenced in writing bureaus staffed by prescholastic clerics, who were actively involved in discussion on semiotics even as they wrestled with questions in sacramental theology.	2
Eleventh and twelfth-century lay elites came to be the subjects of representation in the explicit sense that, in situations requiring authority and commitment, they evolved from immediately present agents to represented actors. Persons absent in time or place were substituted by seals, which operated as alternates for those who were absent, acting in their place. It is intriguing that personal identity came to be signified just as people began to project their authority and accountability beyond their own actual, empirical presence. It is as if absence were required for the question of identity even to become conceivable.6 Since seals are evidence, in my opinion, of a more general and unprecedented shift toward mediation, representation, and the formulation of personal identity in the medieval West, questions arise about the conceptual origin, form, signifying modes, and agency of this new medium, the sealed charter.	3
My own method in exploring this matter has been to follow not the principles but the analytical agenda of Peirceian semiotic anthropology, which is critically presented below.7 By focusing on seals and on the institutions that produced them, I probe the effect of contemporary medieval theory on this sign's agency, assuming that seals' semiotic codes were dependent on a theology and an ontology that fostered their diffusion and interpretation. In this analysis, I do not seek to establish an absolute symmetry between semiotic theory and seal praxis. Rather, I examine how the seal was enabled by and how it encoded a specific set of ideas about signs and semiosis, and show how seal usage and metaphor contributed to contemporary reflection on and development of semiotic thinking.8 I ask what idea of semiosis must have been operative and what the place of ideas within semiosis was that enabled ideas about sign efficacy to create and shape material signs. Lastly, wishing to elucidate the social effects of seals as agents that performed and produced cultural works, I examine the action of seals as an innovative semiotic trope that, both in theory and in social practice, re-figured the categories of person, presence, identity, and authority.	4
I will argue that, in projecting personal distinction, seals acted through a system of identification, designation, and recognition in which representational identity rested on an ontological principle of likeness. The medieval seal was a serial object: seal iconography utilized a limited range of distinctive types, themselves established on the basis of a limited range of stereotyped personae, and the engraved seal-die (matrix) itself repeatedly projected its owner's identity by reproducing identical impressions. This technology of replication appears to have served as a model for the formation of medieval identity. Seal users thus came to develop an awareness of themselves in relation to an object whose operational principles as a sign were categorization, replication, and verification. As the elites who used seals came to depend on representation by signs, the concepts of both social and personal identity came also to be formulated in relation to such signs. This is not to say that such representation and such concepts were completely congruous with any definition of the self-as-an-individual as might then have existed, or that the notions of individuality and subjectivity were primarily generated by, or a construct subject to, cultural codes.9 I am not addressing here the entire postmillennial experience of selfhood or personhood, but I am exploring a new experiment in signing and signifying both person and personal identity within northern French culture and society.	5
In modern Western societies, while the term "identity" refers generally to those characteristics used to identify, define, and distinguish persons so that they can be individually recognized, it is also acknowledged that these characteristics as well as the very notions of identity and individuality may vary with time, place, and culture. In the medieval lexicon, the concept of identity did not address individual personality. Rather, identity in the eleventh and twelfth centuries centered on a logic of sameness and operated by assuming a model of similarity, referring to human beings as members of an identical species, or to the person as a psychosomatic whole, a social agent identical to itself with respect to number, essence, or properties. Since that particular sign, the seal, which accompanied, indeed articulated, the assertion of personal identity, participated in this same logic, conceptions of the sign and the human subject appear to be closely related. Indeed, they both operated on the basis of a newly elaborated premise of a dialogic connection between semiotics, theology, ontology, and anthropology.	6
Concern about mediation, signification, and representation pervaded the eleventh century. The whole of Western Europe was then agitated by the Investiture Controversy, a dramatic conflict between church and state in which the pope struggled with the German emperor to establish absolute ecclesiastical control over the appointment of church officials. Less emphasized in traditional historiography but central to this conflict were questions surrounding the effectiveness of certain signs, particularly material objects. The papal party believed that the symbols of ecclesiastical office, the ring and the crozier, possessed no intrinsic capacity to cause any effect but that the valid possession and application of them effectively and irrevocably established an ecclesiastic's right to both office and its associated power. The underlying sign theory thus held that material symbols were ordinary objects whose significance derived from a value ascribed to them by common agreement, by their recognized use in a particular ceremony. At stake here was the very nature of the operation of signs, the belief that their efficacy might be based on a contract or covenant and need not depend on any value inhering in the sign-object used. After a century of heated discussions about this semiotic issue, Thomas Aquinas (d. 1274) was to propose a reversal of this position, arguing that signs were effective on the basis of inherent or infused virtue.10	7
In eleventh-century northern France, however, the semiotic debate extended beyond a consideration of the efficacy of the signs of ecclesiastical investiture. The signifying modes at work in language, in writing, and in such fundamental signs of divine revelation as the sacraments, the Incarnation, and the Trinity came under intense scrutiny.11 The literate elites involved in this inquiry were prescholastic Christian churchmen who were active both in ecclesiastical schools, where they taught and directed doctrinal debates,12 and in chanceries, where they supervised the production of written documents.13 School and chancery shared not only the same location but, significantly, the same staff, which I have come to term chancery-scholars.14 In many cases, the theologically engaged scholars were themselves chancellors specifically in charge of the writing bureaus that produced charters, or else they were bishops or abbots responsible for the written output produced in their names.15 A complex skein of filiation, apprenticeships and training, associations, preferments, and marginalizations bound such scholars together over time and considerable distances. Archbishop Lanfranc of Canterbury (d. 1089), for example, was closely associated with Berengar of Tours (d. 1088) in the Loire Valley before settling in Normandy as master of the cathedral school at Avranches, founding master of the school at the abbey of Bec, and abbot of St. Etienne de Caen.16 At Bec, Lanfranc trained both his successor, Anselm of Canterbury (d. 1109), and Ivo of Chartres (d. 1115), who became abbot of a community of Augustinian canons near Beauvais before receiving the bishopric of Chartres.17 Anselm of Laon (d. 1117), who had been Anselm of Canterbury's student at Bec, became chancellor to the bishop of Laon, while gathering around him, in turn, such students as his brother Ralph (later to succeed him as chancellor, d. 1133), Peter Abelard (d. 1142), William of Champeaux (d. 1121), Alberic of Rheims (later archbishop of Bourges, d. 1141), and Gilbert of Poitiers (head of the Porretain school, chancellor at Chartres, and later bishop of Poitiers, d. 1154). William of Champeaux, at the time of his death in 1121 the bishop of Châlons, had been master in the cathedral school of Paris before founding the abbey of St. Victor in Paris.18 At the beginning of its existence, this abbey functioned as a virtual chancery for the production of royal diplomas while also evolving as a major doctrinal and spiritual center under the aegis of Hugh of St. Victor (d. 1141).19 A great admirer of the Victorines, the theologian Praepositinus of Cremona, who in his later years (1206–1210) became chancellor of the cathedral and university of Paris, is worth mention in this context, since homiletic materials he derived from his documentary and sealing functions on behalf of the bishop of Paris are the earliest of this genre to survive.20 Gilbert of Poitiers (d. 1154) became chancellor of Chartres, succeeding Bernard of Chartres, whose student he had also been.21 There was a break in scholarly activity after Gerbert d'Aurillac's tenure (later Pope Sylvester II, d. 1003) at Rheims, but the school reemerged from obscurity with master Herimann (d. ca. 1075) and his disciple Bruno (d. 1101), ultimately producing the controversial logician Roscelin of Compiègne (d. ca. 1125). Bruno also served as chancellor to the archbishop of Rheims, before himself founding the Grande Chartreuse, the mother house of the monastic Carthusian Order. Following the chancellorship of the learned humanist Godfrey (d. 1094), Alberic (d. 1141), who had trained under Anselm at Laon, became head of the cathedral school at Rheims in 1094.22 Boulogne, Arras, Cambrai, Amiens, Beauvais, Soissons, Senlis, and Rouen all had chancellors and bishops who were scholars, although many schools and their masters remain to be studied in detail.23 The map thus established of cathedral schools should be extended to include monastic establishments, for there was a fluid exchange of individuals and ideas between these two institutional worlds. Reform-minded bishops or their chancellors often founded or reorganized local abbeys; masters of schools not infrequently returned to cloisters (Bruno, William of Champeaux); indeed, some scholars produced most of their work in a monastic environment (Lanfranc and Anselm of Bec).	8
What was novel about these chancery-scholars and deserves our attention is the heightened semiotic sensitivity of their theological debates, their pronounced tendency to ponder the issue of presence and representation. Two of their constructs were unprecedented in the medieval West. First, they came to recognize presence and representation as essential to the structure governing the generation of identity, conceiving identity as dependent on sameness (that is, identicality) but necessarily involving interactions between the similar and the dissimilar. The identity they contemplated concerned both divine and human persons and sparked discussions on the very nature of personhood. Second, they objectified identity by using a new material sign: the seal. Thus the definition of identity that emerged in the eleventh century derived from specific concerns initially directed, later redirected, by the articulation of this definition within a theory of signs. In order to understand both the concept and the sign of identity, and their agency, it will be necessary to examine the domains that concerned chancery-scholars and led them to innovations in thought and social praxis. These domains included the relationships between language and reality, between the Eucharist and real presence, between the Trinity and the related subjects of person, image, and resemblance, and between writing and authority. Such issues were hardly new in the Christian culture of the West, but in their treatment as a set of related concerns they indicate a crisis in the dominant signifying system.	9
Seal of William, count of Nevers, 1140. Archives Nationales, Paris, D 859.
Discussions of linguistics pursued by prescholastics in wrestling with questions of sacramental theology involved a renewed study of the fundamental corpus of semio-linguistic theory that had been provided earlier by St. Augustine (d. 430).24 A resulting interpretive shift in the understanding of Augustinian theory brought an awareness of what may be termed the Augustinian paradox. Augustine's semiotics, presented in Book 2 of De Doctrina Christiana, are a confusing tangle of claims and doubts.25 Early church doctrine seems to have privileged the classical dualism between the sign and the thing referred to by the sign, whereby only the thing, though ideal and not of this world, has reality; the dualistic Augustine emphasized the lack of congruence between signifier (indicator) and signified (that which is indicated), privileging eternal ideal objects of reference over signs, and he deplored linguistic multiplicity and semantic obscurity as a condition of the Fall. Augustine recognized two classes of signs, signa naturalia, or natural signs, which he conceived as having a necessary and causal relationship with their referents (for example, "where there's smoke, there's fire"), and signa data, or given conventional signs (language, clothing, money), which signify by virtue of their givers' essentially arbitrary intentions.26 He seems never to have considered the possibility that conventional signs may function more like natural signs, because he did not believe that causal dependence or logical implications between signs and referents were possible models for language and culture.27 In Augustine's dualistic and idealistic theory, human language is an external imitation of a transcendental reality, lacking its necessarily ideal referent or object and thus fundamentally unable to express God's essence, God's identity as the perfection of self-reference. Understanding language as a form of alienation, since only God is Logos—the unique extra-semiotic guarantor of the adequacy of signs who resists capture by referential language—Augustine effectively deprived human knowledge of the possibility of stable notions and impeded the reification of human understanding.	10
Yet Augustine also wished to bridge the abyss between sign and thing that he himself had so effectively excavated, and this presented a paradox. The Augustinian solution for connecting word and thing, for circumventing the deferral and mediation inherent in text and language, is communion with pure presence, that is, incarnation. As God incarnate, the word-become-flesh, Christ bridges the gap between signifier and signified, for in Augustine's doctrine (as in the later dogma) of the Incarnation and the Eucharist, substance and its representation are one and the same. In this view, "the word of God [the Logos] suffered no change although it became flesh in order to live in us."28 Sacraments in this construction are different from other signs; they actualize the presence of that to which words merely point. In Augustinian theology, the incarnation of the Logos became a model that, while still limiting linguistic expression, promoted sacramental signification through presence.29	11
Thus, although Augustine reiterated the Platonic idea of a schism between sign and thing, he also left a semiotic legacy of reification, an escape from the mere referentiality of signs, a locus for unmediated presence. Worth noting here is the transition from signification to reification, from sign to thing to the silence of the word, which, made flesh, transcends the entire system of discourse.30 Augustine's desire for a communion with pure presence undermined the older representative mediation of signs, but it also provided, in the interpretive hands of prescholastic theologians, the seed for a new theory of representation. Identicality between sign and object came to inform a novel signifying process, during the twelfth century, when the Eucharist was firmly conceived as being, in and of itself, what it represents. It has been conventional to invoke a growing acceptance of Aristotelian thought as accounting for the appearance in prescholastic culture of the idea that a symbol partakes of the reality it expresses. However, it may well be that the Augustinian semiotic corpus was itself perfectly capable of inspiring the belief that immanence was central to the operation of symbolism.31	12
The governing, encompassing question was, therefore, that of the relationship between signs and the world, and the implications of this question were brought to the fore in the course of the controversy provoked by the ideas of Roscelin of Compiègne. During the second half of the eleventh century, Roscelin, the most famous teacher of dialectics in the schools of northern France, initiated what came to be known as nominalism and thereby launched a debate about universals. Opposing nominalists to realists, this debate bloomed into a pivotal controversy in medieval philosophy. Universals, for instance "man" or "animal," were general categories of properties shared by many particular entities. The discussion focused on the ontological status of these universal categories: what degree of reality did they possess, from what did they derive? For the nominalist Roscelin, these categories had neither objective nor subjective reality. They existed neither in the mind nor in reality, being simply spoken sounds or verbal expressions for mental constructs derived from experience with particular entities that exist in nature alone.32 On the other hand, following the Platonic and Augustinian tradition, the realists maintained that, although universal categories did not have corporeal existence, they nevertheless did exist outside the human mind: in God's mind, where, eternal and immutable, they were the source of forms for spatio-temporal things.33 Roscelin's nominalist denial of ideal realities (universals) and of any linkage between word and physical property contradicted Augustine's position on the reality of universal categories but not his distinction between words and referents. For Roscelin, however, referents were other words and not real things identical with divine ideas, as they ultimately were for the Augustinians. Supporters of Augustine's position, such as Anselm of Bec, Alberic of Rheims, and William of Champeaux, defended it by shifting from the earlier medieval accent on Augustine's radical dualism between word and thing to an emphasis on his theory of ontological immanence and participation.34 This theory argued that things guided the properties of signs, that, inhering in the spatio-temporal realm, universals created similarities among objects. To be sure, participation in the transcendent was not a matter of identicality but only of resemblance; only God, uniquely, possessed true identicality.	13
Disagreeing with both his teacher Roscelin and with Augustine and his followers, Abelard denied the existence of anything that is not a particular. While retaining the notion that the common nature inherent in things of the same species made them similar, he argued that such similarity fell short of constituting them as universals. For Abelard, words were universals, concepts of things, not images of things. Yet words functioned by means of images deriving their meaning, not from the things themselves but from the mode of signification at work in the human mind. Abelard held that the mind creates at will images or copies for configuring absent things. These images are the proper objects of thought and understanding, which thus operate on a likeness that the mind creates. This likeness has neither substantial reality nor the underpinning by transcendent universals that, for the Augustinians, accounted for the similarities between things. For Abelard, words apply the mind to the likeness of things, but words designate images, not their objects; words therefore signify an understanding of what they predicate rather than refer to the object itself. By pointing out that thoughts and understanding are not the same as their objects, Abelard displaced the Augustinian notion that divine realities are actually present in the human mind where they beget images of themselves. He located the act of understanding in the mind as the active inventor of universal concepts with its modus operandi of created images. From this initial controversy over universals, there emerged a reinforced vocabulary of "likeness," and an attendant notion of images as signs of absent things.35	14
Conflict over universals also permeated the argument raised by Berengar of Tours in the eleventh century, which prompted the northern French schools to reconsider the nature of the Eucharistic sign.36 For Berengar, as for Abelard, the issue was the relationship of logico-linguistic structures to the mind and to reality.37 The Eucharistic debate was couched in quasi-documentary terms: was the Eucharist "dispositive," that is, endowed with real presence and inherently potent, a belief that would equate the interpretive material supporting it (texts) with events in the real world? Or, as Berengar maintained, did the modality of the Eucharistic presence have to be consonant with a linguistic model of grammar and logic: in short, did the Eucharist have to be allegorized in order to be understood? Berengar refused to admit that which is denied by the evidence of the senses or by simple logic. He therefore stated that the bread and wine remain on the altar after consecration, and that it was interpretation, a getting beneath the surface senses, that gave the Eucharist genuine meaning. Though relying, as Abelard would later, on the tool of linguistic philosophy, Berengar argued from and for the older dualistic Augustinian distinction between sensible and spiritual, between symbol and reality.38 The rejection of physical symbolism that Berengar and his followers advocated was opposed both by monks (for instance, John of Fecamp) and by prescholastics like Lanfranc,39 Herimann of Rheims, and his student, the chancellor Bruno, whose scriptural exegesis and theories of the Eucharist exerted great influence on another chancellor, Anselm of Laon, whose teachings in turn influenced William of Champeaux and Hugh of St. Victor.40	15
Chancery-scholars were strong promoters of the notion of real presence, and it was indeed in defense of this concept that they engaged extensively in larger debates over sign theory, representation, and the authority and authenticity of the written word, both scripture and script. The earlier Augustinian semiotics buttressing Berengar's position had stressed the radical duality of signs as involving a negative dissimilitude: on the one hand, a mental, eternal signified, on the other, a physical, transitory signifier that refers to its object but is otherwise inessential to it. As they had done in their discussions of universals and of the referentiality of language, prescholastic theologians probed such dualism in developing their Eucharistic theology. In so doing, they scrutinized the economy by which an iconic sign might be similar to that which it denotes, and the mode involved and the extent to which it might itself partake of the object represented. As in the discussion about language and universals, attention was redirected away from Augustine's dualism toward Augustine's appreciation of a sign's tangible aspect.41 The incarnation of God was no longer to be elucidated by an image, as had happened when the notion of God as the original and of Christ as living image made it possible to see them as one and the same God, though not the same person. Rather, the image was now held to be the realization of form in matter and came to be understood as an actual incarnation. Images were promoted to quasi-personal beings.42 The language of analogy seeped into the language of ontology: "to be like" became "to be part of." The cultural content of the analogy, that is, the relationship between the object and its image, was altered so that an iconic representation might be seen as more real than the empirical experience. This is what occurs according to the doctrine of transubstantiation, where the consecrated bread and wine are the true body and blood of Christ. The Eucharistic debate produced the idea that reality was capable of being perceived through an iconic convention. In such a cultural crucible, the sign became representative less because of its relationship to a conceptual ideal than for its capacity to embody its referent's ontological characteristics. In semiotic terms, the represented object (the signified) became a constitutive part of the sign (the signifier), because for the sign to stand for its object, the sign had to incorporate a representation of that object; it was the expression of that incorporated representation that came to be seen as the sign's meaning. This newly elaborated semiotic doctrine—though it maintained a distinction between objects, the signifying functions of iconic signs, and their representative capacity—in fact sanctioned a conflation of signifier and signified so that immanence rather than transcendence came to govern the rapport between signifier and signified.	16
The discussion of the identity, whether divine, historical, or allegorical, of the Christic person present in the Eucharist had broached the question of the nature of personhood. This question received growing consideration as the relationship between universals and individuals, destabilized in the above-described debate over universals, was explored in a quest to understand, first, the persons comprising the Trinity and later the human person. Nominalism, by insisting on individuality, tended to fracture the divine unity of the Trinity into three separate entities. The nominalist Roscelin, wanting to address the problem of how the three Trinitarian persons of the Father, Son, and Holy Spirit might be of one substance yet not all incarnate in Christ, analyzed the terms Father, Son, and Holy Spirit in a logical sense, as names humanly imposed. Although nowhere did Roscelin actually state that these names signified separate things, his theology of the Trinity provoked attacks by Anselm of Bec, who accused Roscelin of being a dialectical heretic, one who thought universal substances to be nothing but the puff of an utterance. In The Incarnation of the Word, Anselm asserts that proper names designate different persons, indicating that, while persons bearing such names share a common nature, they are irreducibly distinct one from the other with respect to distinguishing properties: in his person, the Son assumed two natures so that the person of God and the person of man was the same, and that made the person of the Son different from that of the Father and of the Spirit, since different persons cannot be the same man.43 Roscelin was forced to defend his views on the Trinity at a council held in Soissons in 1092, where he evaded the accusation that he preached division in divine essence by affirming that his argument related only to names and nomenclature, not to God himself.44	17
Abelard pursued this debate in a treatise on the Unity and Trinity of God, and he argued that divine attributes were not fixed things but names predicated of God to signify certain properties of his being. The relationships between Son and Father, or Holy Spirit and Father, could be understood in terms of the relationships between these properties.45 Thus Abelard introduced the notion that members of a same species, such as men, may yet differ in their properties, or even by definition, when such properties remain intermixed. The example repeatedly used by Abelard to clarify this instance is that of the seal's waxen image. Both the waxen image made from the material and the material (wax) from which it is made are the same in essence and number, but they differ, not only by definition but by property, because the waxen image must be wax, and it comes from the wax and not from itself (the waxen image was not generated by the waxen image), but the wax may be joined as an image or as anything else, in the same way that if a man is a man, he must be an animal, but the species animal can be a man or any other animal.46 Abelard's reasoning illustrates how contemporary discussions of the persons of the Trinity were never far removed from inquiry into human personhood. Indeed, such debates fostered the creation and dissemination of the very term "person" (persona).47 Richard of St. Victor (d. 1173), a twelfth-century theologian who himself contributed greatly to the definition and acceptance of the term "persona" in the course of his work on the Trinity, commented that the noun "person" is regularly found "in the mouths of all, even of peasants."48	18
The prescholastic milieu of schools and chanceries had to consider human personhood in yet another context, the establishment of documentary authority. The authority of written documents, in a fundamental shift, moved away from immediate dependency on God and the supernatural, coming increasingly to derive from and depend on human persons.49 At issue in this shift was a need to project the authority and accountability of human beings beyond their actual, empirical presence, so as to impart to charters a level of permanence previously expected only of God. The solution achieved centered on the seal, a sign-object standing in, substituting, for its owner or user, and conceived and created so as to produce a duplicate presence, a presence not actual but nonetheless real.	19
Seal of Raoul, count of Vermandois, circa 1146. Archives Nationales, Paris, D 1010.
In the decades following the year 1000, the number of charters produced and preserved in northern France increased by several orders of magnitude, setting off a trend toward written documentation that was never reversed.50 These charters were issued in the names of the aristocrats responsible for the transactions being recorded in writing, such transactions typically involving gifts of land made to religious houses and to their saints for the salvation of the donors' souls. However, the actual production and subsequent control of such written records remained a monopoly of the ecclesiastical beneficiaries who drafted them and maintained them archivally.	20
Both donors and benefactors were interested in ensuring textual and transactional permanence; the most reliable traditional agency for this purpose was God. Documentary writing derived much of its power from a visible affinity with Holy Scripture, an affinity established both by graphic logic and by liturgical manipulation. Graphic logic involved such methods as the inscription of a Chrismon, a trinitarian invocation, the use of Latin, biblical preambles (arenga), and divine maledictions and threats of excommunication against anyone who might challenge the gift being recorded.51 Liturgical manipulations included the charters' production by priestly scribes and their placement on altars or in Gospels.52 A manuscript charter was kindred to Scripture and, as such, was a space of sacred and secure inscription.	21
The charter's text, however, was formulated in the first-person voice of the individual who was making the donation, and it conveyed the will, the intention, of an individual donor. The religiously designed charter located the ego, the "I" of diplomatic discourse, within the rationale of Christian ethics and salvific eschatology. Therefore, with respect to those charters given in his own name, and to which he was entrusting the fate of his soul and of his kin, the donor remained a problematic author. First, he had not himself created the manuscript document, which was rendered in Latin, a language he hardly knew. Second, the written text itself was fundamentally impersonal because its actual scribe, who remained anonymous, wrote in an official or a technical capacity, as a fictive person, persona fictiva, in the name of someone else. Writing in the name of a donor, representing him as author, the scribe introduced motivations, decisions, and gifts, repeatedly using the word ego. Utilizing this first-person form, the scribe, though semiotically entering the subjectivity of the donor, in fact maintained the reference of third person. Hence the locus of subjectivity transcended the individual, and what presented itself as individual subjective discourse was actually suffused with multiple voices.53	22
Diplomatic discourse thus incorporated a cultural "self" that was quite distinct from an individual body. Yet the postmillennial charter required a first-person-singular pronominal category, an ego to function as index, to indicate the originator of the utterance. It may be helpful to point out here that this method of written documentation developed in mid-eleventh-century northern France at the end of a period of a century and a half (900–1050) during which transactions had normally been accomplished by the oral statements of principals and witnesses, usually made under oath and publicized by symbolic gestures and rituals.54 Set within such oral and visual modes, the empirical presence of the subject, of ego, had of course been immediate and undoubted.	23
With the growing importance of writing in the ceremony that sanctioned land transactions, the ego of written diplomatic discourse—a linguistic category differing from the person uttering the words—could not provide referentiality through actual contiguity with the charter's author. The issue became how to reconcile ego, the linguistic category, and ego, the physical individual, the actual subject of the enunciation. It is because the postmillennial charter long remained part of a ceremonial format, in which the charter's operations hinged less on its legibility (as text) than on its visibility (as scripture), that the charter's contextual apparatus long continued to derive from and to parallel the ambient oral modes. The oral and the written did not stand in opposition, then, but operated jointly within a single framework of intelligibility.	24
This framework rested on the primacy of empirical presence in the assertion of authority; it construed power to emanate from character, to be the effluence of personality. Thus, when gifts of land were contested and resolved by charter, as often happened, such disputes were not settled by considering the parties as donors and recipients, and by applying legal rules appropriate to these categories, but rather by an agreement through which the status and self-esteem of both parties as particular individuals might be saved and a social relationship between them created or renewed. Behavior was remembered and inscribed in the form of statements about particular persons and their actions.55 The attention was to individual will and responsibility, to a personal examination of the implications of one's actions, which were understood as involving, beside terrestrial and social consequences, merits capable of saving (or losing) one's soul in the afterlife. In short, the legal realm conjured up by the charters was equated with the realms of ethics and theology. The "subjective" and "personal" in the law, far from diminishing legal authority, in fact constituted it.56	25
The individual person encountered in prescholastic charters is, pace Jacob Burckhardt,57 an autonomous, voluntary, and empirically present agent, situated within a set of social relationships arising out of consent and contract. The sources of empirical presence within the charter were, necessarily, the event recorded and its author. Thus the focus on the individual coalesced around two related requirements—the need to anchor the written charter within the concrete ceremony of gift giving, and the need to embody the determinant elements of this context: the transaction itself and the actual speakers (empirical, physical persons who had performed and witnessed the transaction). How to achieve this incarnation? The answer was a system of signs. Thus the evolution of the charter's format from sacred inscription to sealed deed occurred as an attempt to incorporate within the charter the actual nature of personal authority rooted in being, soon to be obsolescent.	26
At first, transactions and their authors loomed equally large. Evidence of the transaction, such as a symbolic rod or knife, was initially either manipulated together with or even affixed to the charter.58 By the late eleventh century, however, only donors' and witnesses' names appear in the charters, preceded by an inscribed cross, a sign not necessarily autograph, although it might be.59 The signatory who marked a cross on a charter, or had it inscribed, would also have made a ceremonial sign of the cross across his or her body, an act usually described in the charter's text.60 The manual and manuscript crosses were signs both of identity and commitment, typically accompanied only by the name received at baptism. Such crosses indicated individual filiation as son, or daughter, of God, and hence individual commitment as God's child; they recorded Christian filiation and the subscription of a solemn oath made in the presence and in the name of Christ crucified. Thus the person engaged by the charter was responsible both for the content of the charter, the gift of land, and for his or her soul, since the land had been given, on oath, to enter the economy of salvation.	27
The initially tangible symbols of rod and knife gave way to the scripted cross on charters. From signs of conveyance to signs of the authors and the witnesses to conveyance, the focus moved from the action and its object to the actors. This shift occurred in conjunction with the increased concern for salvation characteristic of postmillennial aristocratic spirituality.61 The nascent hermeneutics of personal identity tended to merge with the theology of the soul, but this fusion did not last long. For all its potentially powerful symbolism, the cross functioning as a sign-signature on a charter marked identity only in the broadest possible terms: membership in a Christian society. Although the cross might emanate directly from the author of the charter, from the "I" of diplomatic discourse, it more often was actually traced by the scribe. In either case, the cross denoted that the authority for the enforcement of the charter, its ultimate warrantor, was God; indeed, the cross both signed and signified God.	28
When seals began to be affixed to documents during the course of the eleventh century, the manuscript textual cross was still a standard appurtenance of charters. The charter by which Robert, son of the count of Flanders, made a donation for the salvation of his soul to the abbey of Wheaten in 1093 reads: "In order that these dispositions may remain firm and untouched through eternity, I have had this charter confirmed and signed with the victorious symbol of the holy cross, and with the sign of my authority and the seal of my highness."62 While both cross and seal signified a sacred undertaking, and both have rhetorical presence within the charter, there are two major distinctions between them as documentary signs. First, the cross remains a written sign, and in this case non-autograph, whereas the seal is both a material object and a figural presence that emanated directly from the author of the charter. Second, as the fairly standard clause within the document makes clear, the cross symbolizes Christ victorious, while the seal signifies the authority, as it is the image, of its owner. The cross signaled Christian kinship and invoked God's authority; the seal marked and invoked personal identity and authority.	29
Seals have a long history.63 Originating alongside if not actually preceding the invention of writing, sealing remained, in most civilizations, a valuable mechanism for marking and protecting ownership, signing commitment, designating identity, representing authority, and authenticating documents. In parallel to their roles in the sphere of practice, seals also served as a metaphoric focus. Mesopotamian and biblical texts, Platonic and Aristotelian treatises, patristic and early medieval commentaries all incorporate the imagery of seals as a conceptual tool.64 Such historical longevity, however, does not necessarily imply congruence between the cultural and modal significance of the seals. Most historian-sigillographers have simply assumed continuity of seal usage between very different societies as a category of historical explanation, thereby promoting interpretation of the seal as a single "historical," and thus ahistorical, object. In addressing the new appearance of seals on charters issued in the name of French non-royal elites between 1000 and 1200, my earlier work grounds this diffusion laterally, within the very circumstances of its occurrence, rather than approaching it vertically, as an event somehow predicated or determined by historical continuity.65	30
Seals were not entirely novel in eleventh-century France.66 From the seventh century onward, seal impressions had been affixed to royal documents exclusively, by a chancellor who had custody of the royal seal and was responsible for the production and validation of royal diplomas. In early Frankish times, under the Merovingian kings (sixth to mid-eighth centuries), the royal seal affixed to diplomas did not function as a means of documentary validation. Its use imitated the usage of the Byzantine imperial chancery; to seal a document was, for the Merovingian kings, to behave as a ruler. The seal was a sign of the king, the only person who could issue documents thus marked. This linkage between documentary seal usage and royal status was maintained upon the ascension of the Carolingians (751); royal seals came to operate as an apparatus of the office of kingship. They spread in parallel to the Carolingian multiplication of kingships. Each Carolingian offspring used a seal upon his assumption of a royal office, but as soon as a king was functionally replaced by a non-royal official, such as a duke, the use of seals disappeared, although the political entity might still be called a kingdom (Lat. regnum), and the overall administrative structure remained otherwise intact. Seal usage thus related to kingship and not to the territoriality of power, a characteristic further emphasized on Carolingian seal legends, that is, the inscriptions surrounding the images, which included only the name and title of the ruler; territorial designation was absent. The same seal might be, and was, affixed if and when a ruler changed kingdoms, as happened during the tumultuous partitions of Charlemagne's empire from the mid-ninth century onward. A particular seal might and often did serve for different rulers so long as the names and titles matched. The accent was on the function and on the nature and degree of ruling authority (king, emperor) as defined by the title. It was specifically along titular lines that seal matrices were systematically engraved afresh, such as when a king became emperor. More official than personal, Carolingian royal seals projected an order of reality grounded in permanence, obscuring the contingency inherent in any individual ruler by reference to the continuing symbolic activity of statehood. By early Capetian times (early eleventh century), however, the royal seal was used only sporadically and had lost some of its standing as a formula and prerogative of kingship.67 During the second half of the eleventh century, the royal chancery resumed the systematic sealing of diplomas, while French non-royal elites began, for the first time in the medieval West, to seal charters issued in their own names. Non-royal sealing demonstrated and articulated the loss of a royal prerogative. This may indicate the establishment of competing comital, episcopal claims to authority, a desire to share the aura of royal status by the emulation of royal chancery practices, or both. However, since the appearance of non-royal sealing and the revival of royal sealing are contemporaneous, it is worth considering the possibility that eleventh-century usage of royal seals was a part of, rather than the model for, the new spread of sealing to non-royal elites.	31
Seal of Pierre de Courtenai, count of Nevers, 1184. Archives Nationales, Paris, D 863.
In analyzing the diffusion of seals along the axes of regionalism, politics, and gender, I have come to rethink four previous assumptions that have long, almost axiomatically, dominated the field of medieval sigillography, or sphragistics, that is, the study of seals,68 and have in my view obscured the actual historical significance of medieval seals, relegating them to the world of antiquarianism and connoisseurship.	32
The first assumption is that the seal's function, between the eleventh and fifteenth centuries, was to authenticate documents. This notion, which was first articulated at the end of the twelfth century and received its prescriptive formulation in the thirteenth, cannot account for the early pattern of non-royal seal usage. In fact, when late twelfth-century canon lawyers began to reflect on documentary validation, they assigned the power of authentication only to the sigillum authenticum, the authentic seal. The meaning of authentic here does not derive from a concern about counterfeits but from the desire to establish the capacity for authentic seals to confer full validity on documents devoid of witnesses. However, if from its very inception the concept of the authentic seal involved a precise understanding of the seal's effect, this effect was specifically understood to emanate from the public authority of popes and rulers. Thus the authenticating power of seals was conditional on the status, conceived as public, of their owners. By the late thirteenth century, when jurists attempted to provide the authentic seal with a broader sociopolitical conception, they insisted that for a seal to function as an authenticating device it must be well known. Even when finally producing such imprecise and relative definitions, jurists did not conceal the fact that viewpoints in the matter of seal validation differed widely; they recognized that the meanings and agency of seals depended on local custom.69 In short, medieval legal discussions of seals, not to mention actual sealing practices, far from displaying consensus about the seal as a validating device, testify to the difficulty legal scholars had in articulating the values and beliefs implied both by the authenticating and by the sealing processes. The question for these scholars was the very nature of the authority underlying the seal's efficacy, since non-royal seals did not base their owners' authority on royal grant and affiliation, nor did seals invoke the political hierarchy as party to the act they witnessed.	33
A second hypothesis holds that seals spread because of the concurrent revival, in the twelfth century, of trade and urbanization, the growth of bureaucracies, the reintroduction of Roman Law, and the spread of literacy. Enabling conditions should not be mistaken for explanations, nor do the circumstances lend themselves readily to a chronology indicating the precedence of one phenomenon over the others. There had been a moderate continuum of unsealed charters given in the name of lay authors since early medieval times, but only when they came to be sealed did such texts lead to that generalized and irreversible social dependence on the written word that has continued up to the present day. Thus it seems that seals furthered rather than resulted from literate modes. This suggests that seals played a unique role in fostering, to borrow M. T. Clanchy's expression, medieval trust in writing. Basing his argument on English records (1066–1307), Clanchy gives the most satisfactory account to date of the seal's ability both to encompass and to translate the meanings of the symbolic objects and gestures that had previously validated written deeds, or indeed had entirely substituted for them.70 This scenario elegantly situates the seal as a bridge between the literate and the non-literate, deftly bypassing a polarized historiography that had either associated seals with the growth of literacy or labeled them a technique for illiterates. However, these theories did not adequately confront the fact that there is no systematic or even necessary relationship between seals and the growth of literacy. Charlemagne (d. 814), for example, reinforced the dependence of his administration on the written word by turning to a system of notaries to impart authority to non-royal documents even though his own chancery was sealing the royal diplomas.71 Then, too, areas of Southern Europe that had retained a higher rate of documentary practice throughout the early Middle Ages also used the notariate, adopting sealed charters comparatively late.72 Thus there clearly were successful medieval literate practices independent of sealing.	34
A third presumption maintains that seals were icons of power. Interpretations of seals as evidence both for and of social and political processes assume a causal relation between the function and meaning of seals and the status and authority of their owners. I have reconsidered this putative causal relationship in the light of the documentary practices discussed above. To repeat: there is no extant evidence of literate intercourse between lay individuals for the purpose of transacting land in eleventh and twelfth-century France. Such operations of lay society not involving churchmen appear to have been accomplished primarily by means of oral and gestural modes. In this period, as before, churchmen held a scribal monopoly and were responsible for both the production and the conservation of charters. The ecclesiastically scripted but lay-sealed charter thus indicates secular participation in and acculturation to documentary modes, which were fostered as much by the preexisting churchmen's scribal and scriptural culture as by sealing. The new category of sealed charters must therefore be analyzed within the contextual framework of their originating scriptoria. The sealed charter, heretofore interpreted solely as an act of individual or familial will, must now be reconsidered as a text and artifact articulating cultural and ideological models ambient in specific scriptoria.	35
A fourth prejudice, indeed a paramount force that has focused seal scholarship on artistic considerations and legal functionality, and has, more generally, shaped traditional sigillography as an antiquarian discipline, is its heuristics, based on casts. Archival collections usually consist of casts made from molds directly taken from single originals, the best impression of a given seal type. Seal catalogs typically describe such casts, thus discounting information inherent in original seal impressions and subsuming such variants into a single archetype. The use of casts as standard objects of study tends to transform the seal into a fixed type, undermining its most fundamental signifying and operative principle, reiteration, by which medieval seals produced identity through identical devices. Aftercasting denatures seals by altering their locus (from documents) and their status (as signs), abstractly recasting them as separate objects of knowledge, removed from their original cultural sphere of discourse and practice.	36
Restoring seals to their historicity, as agents within the culture that produced and used them, extends our understanding of the instrumentality of seals well beyond their long-recognized documentary function. A semiotic approach to seals enables them to be reexamined as signs and symbols, and redirects the analysis toward their modes and areas of signification (rhetoric, sign theory, authority, personification, identity), toward the assumptions encoded by seals about the nature of their operation, and toward the effect of seals in and on the society that manipulated them.	37
It is noteworthy that medieval authors themselves explicitly defined seals as signs. This might earlier have suggested a semiotically informed study of them to historian-sigillographers, yet it was not the sigillographic literature that first proposed a semiotic analysis of seals.73 Rather, broader works on modern semiotics and anthropology have helped bring to my attention both the nature and implications of seals as semiotic agents and processes and the extreme sensitivity to semioticity during the period in which seal usage spread. Of particular relevance for the conceptualization of such an approach is semiotic anthropology, and it is at this point useful to review the extent and limits of this discipline, which arose at the University of Chicago in the mid-1970s.74 Building on insights of Ferdinand de Saussure (1857–1913) and Charles Sanders Peirce (1839–1914), semiotic anthropology encourages considerable attention to language without confining culture to the single reductionist model of a linguistic code. This approach holds that cultures are particularized by semiotic processes and semiotic systems specific to them. Semiotic processes involve mediation, reification, naturalization, metaphorization, and emblematization; these are the means by which material features of the sensible world (color, stone, bone, blood, flesh) are made to perform as vehicles of signification. Semiotic systems incorporate the various ideologies that organize these processes and enable their signification by providing interpretive codes and principles (interpretants). Encased both in social action and in theoretical discourse, the ideologies of semiosis are themselves cultural phenomena subject to textual forms, pragmatic rules, and semiotic processes. Stipulating ways to interpret, such ideologies also suggest how the signs themselves signal the way they are to be interpreted. Thus signification is always dialogic, at once representative and interpretive of the practices and conventions of a culture. Semiotic anthropology insists that signs are historical, since both their actualization and their performance are matters of contextual appropriateness.75	38
Current trends in semiotic anthropology are more indebted to Peirce than to Saussure. Saussure's structuralist theory of language assumed fixed abstract codes that underlie the utterances of a language. Such codes, shared by the members of a given culture, depend on a correspondence between the word (the signifier), not so much with the existence of its object in the real world but with its constructed cultural equivalent (the signified). Meaning is seen as intentional and definite, independent both of the presence of a "real object" and from any particular interpretive human subject. Saussure posited an equal exchange between the univocally corresponding signifier and the signified, which bracketed both the external world and the interpreting self. His approach has little interest in the historical causes behind meaning.76	39
Peirce, on the other hand, proposed the necessity of an equivalence between representation and reality, which was suggested to him by the workings of scientific rationality. Far from assuming the preeminence, abstraction, arbitrariness, and permanence of a linguistic code at any given point, Peirce argued that signification occurs through a sign's real relation with its object. A sign exists as vehicle (or representamen) in relation to another sign that acts as interpretant (or meaning), so that both signs represent the same object (or referent). Peirce thus posited a triadic relation between sign, object, and meaning. These three elements may pertain equally to such diverse classes of phenomena as single objects, human actions, or natural laws. A sign relation, that is, the unique semiotic bond linking a particular triad together, is made of two reciprocal vectors: determination, which points from the object toward both the sign and the interpretant, and representation, which points from the sign and the interpretant toward the object. The dynamic between determination and representation implies that each element in the sign relation modifies its roles as further determinations and representations are accomplished. While semiosis is a limitless process of interpretation, representation remains guided by determination, for "the object of a sign must resist in some measure any tendency it may have to be as the thinker thinks it is."77 This real relation between sign, object, and meaning is itself predicated on Peirce's metaphysical notion that the signs used to represent mental and external realities also share substantial identity with these realities, and on his ontological view that all knowledge at a given historical moment must to some degree relate to something with which the knower is already acquainted. Peirce thus produced a relationship of dialogic adequacy between signs and objects, between meaning and experience, and between thought and reality.78	40
This insight is not popular with social scientists, most of whom dismiss it as a vestigial Western attachment to the Renaissance theory of signatures,79 but it is precisely what recommended Peirceian semiotics to semiotic anthropologists, and what should make such semiotics valuable to historians, for it locates meaning within history by suggesting that it is only in social practice that the sign is used and its sense determined: semiosis as enacted social practice. Meaning, therefore, is not disembodied but acts and is enacted within sign systems (linguistic and nonlinguistic alike), which, embedded in context-specific purposive behavior, are at once socially grounded and socially creative. Contextual parameters are concrete realities—time, space, matter, and sign operations all require physically manifested sign vehicles, experienceable over time. Complementary to this concrete contextualization is the tracking of a given culture's meta-semiotic understanding through theoretical discourses, ideological assumptions, and social actions. The material aspect of semiosis does not deprive contextually grounded signs of meta-level correlates regulating a further range of acceptable meanings. Thus semiosis, as a multidimensional process sensitive alike to the formal properties of signs, the material circumstances of context, and the influence of meta-semiotic anchors, opens up ways to study social action seen both as emergent in real time and projected from meta-semiotic representation.80	41
The European Middle Ages, the twelfth century in particular, have recently been examined through the lenses of semiotic anthropology by Richard Parmentier, a distinguished practitioner of this methodology. Parmentier considers the twelfth century to have been primarily governed by Platonic Realism, which affirmed the reality and transcendent referentiality of abstract ideas or universals.81 However, he also comments on the diversity of twelfth-century symbolic discourses, though without specifically identifying such discourses, as earlier in this essay where I analyzed the conflicts that erupted over the reality of universals, about the Christic presence in the Eucharist, over the nature of the persons in the triune God, and over the expression of authorial presence in script. Based on his perception that twelfth-century semiosis had multiple, often contrasting, operative mechanisms, Parmentier draws the general conclusion that specific semiotic structural elements, such as Realism, do not necessarily correspond to particular types of societies but are widespread phenomena acting cross-culturally. For instance, since universals are conceived as templates of the spatio-temporal realm, social regimentation and political permanence can be ascribed to any society that proclaims the transcendent immutability of universals.82 Realism, in this conclusion, has been essentialized, that is, assumed to have systematic effects independent of context.	42
Parmentier's conclusion involves a radical epistemological shift, as several aspects of semiosis are decontextualized and universalized, while the explanatory power of semiotic anthropology is directed away from categories of cultural order and toward types of cross-cultural experience. Such an approach raises several problems. First, it denies a particular semiotic theory, in this case Realism, its own power as sign and as engine rather than as mere determinant of semiosis. Second, although synchronic but conflicting semiotic systems are recognized, those medieval sign operations not accounted for by contemporaneous theoretical discourses are subordinated, identified by reference to anachronistic modern interpretive norms, or conceived as having been so pervasive and axiomatic as not to have been recorded in texts. Third, medieval semiotic processes themselves are translated directly into analytical tools of modern research, when, in fact, such processes were constitutive of twelfth-century culture and operated in a vastly different interpretive context than the tools of current social studies. Consequently, actual medieval semiotic processes, though initially contextually anchored and/or meta-semiotically correlated, become, once reinscribed (and reified) within modern epistemologies, nonfunctional, since the ways in which they originally enabled specific and new signifying forms, new meanings, new forms of meaning, and new chains of interpretations remain unretrieved. When semiotic practices are seen simply as presupposed habits, objective discourses, or analytical tools, the interpretive creativity of signs within historical societies cannot become a proper subject of historical inquiry. Parmentier's comprehensive review of semiotically informed studies on the twelfth century reports interactions between different media (images/texts, heraldic emblems/agnatic discourse), different discourses (monastic/prescholastic), and different esthetics (romanesque/gothic). All of these media, discourses, and esthetics, however, are analyzed only from the viewpoint of their engagement with explicit medieval semiotic systems. Such analyses tend to impute a meta-semiotic, superstructural dimension to those medieval systems, whereby they are conceived to be external to the very reality they constitute. The actual semiotic nature of the heraldic emblem or the gothic cathedral, both new forms in the twelfth century, their signifying modes and locations within processual chains of interpretation, and their force in producing specific cognitive and external realities remain unaddressed in this treatment. Semiotics of this sort, from the viewpoint of the historian, merely serves to reinforce the well-known chronicle of innovations.	43
Thus charting the zones left in shadow as the spotlight of semiotic anthropology sweeps across the field of medieval history reveals the paradox at the very heart of semiotic anthropology. On the one hand, Peirce insists on the necessity of context for semiosis to take shape and to make sense, and semiotic anthropologists advocate careful examination of the particular sociohistorical setting within which signs, as contextually informed material instances, operate. On the other hand, the anthropologists have abstracted Peirce's sign theory into an analytical tool held to be applicable to any sign system, on the explicit assumption that his theory is, like calculus, "the indispensable mathematical tool for modern scientific research, [which] makes no claims in itself about the laws which govern the physical universe."83 It is somewhat inconsistent for semiotic anthropology to claim contextualization for semiosis even while essentializing its working definition of the sign. Implicit in the projection of Peirceian sign theory onto all times and cultures is a universalization, a hard-wiring, of signification. Do signs signify independently? Is semiotics a new form of historical determinism, on a par with, say, materialism? Such theoretical essentialism is especially problematic for medievalists who, at least in some cases, feel prompted by Peirce's own knowledge of and reliance on medieval sign theory to make medieval systems "fit" Perceian semiotic trichotomies.84 In my view, the systematic application of Peirce's theory runs contrary to his own notion of the contingency of signification on context, and to his own implication that all knowledge is relative; for, while knowledge is tested by objective reality (and thus is not merely subjective), reality itself remains dependent on thought (that is, on signs and representation, in other words, on culture) in order to acquire knowable meanings and qualities. Peirce's semiotics ought to inspire historians with a sense of the historicity of both sign theory and sign agency, and of their dialogic ability to encode and articulate specific ontological and metaphysical views.	44
The tendency of semiotic anthropology to universalize Peirceian sign theory only partially undermines the utility of its insights for historians. Semiotic anthropology is particularly valuable in exploring material objects as signs because it eschews the systematic application of a logocentric model of meaning, and thus does not reduce culture to the single model of a linguistic code. It calls for the study of material culture beyond technical reductionism or linguistic symbolism and for an understanding of material culture's functions beyond the instrumental or the practical, providing, among other things, a useful agenda for research by which to explore the nature of the signifying material, the agents and purposes of its interpretation, and the status of the relationship between this material and surrounding cultural traditions, social organizations, and cosmological powers.85	45
It is because I have found politics, law, orality, and literacy inadequate as contexts to account for the diffusion of seals and the newer formulation of personal identity in the eleventh and twelfth centuries that, inspired by semiotic anthropology and its programmatic directions, I have come to consider sealed charters from the viewpoint of the writing bureaus that originated them, situating the conception and production of these charters within the scholarly world described at the beginning of this essay as being in the throes of a semiotic crisis.86 The scriptoria and writing bureaus that initiated the sustained production of sealed charters appear to have been located in abbeys or cathedrals that either currently had in residence or had trained those Schoolmen who were active participants in debates about signs and signification in relation to three theological concerns. To reiterate, these were the Eucharist and the related subjects of presence and representation; the Trinity and the related issues of person, identity, image, and resemblance; and the authority of script(ure) and the issue of the referentiality of language. Since the maps tracing seal diffusion and prescholastic theological reflections on sign theory are largely coterminous, it may be the case that the seal derived its new means of signification, especially its capacity to present and represent, from the discourses of semiotics and theology. I propose to interpret the extension of sealing as a manifestation of a new semiotics in which, as already discussed, immanence rather than transcendence governed the rapport between signifier and signified, thereby making possible new forms for the representation of reality. This new semiotics emerged from the context of an increasing, though initially contested, acceptance of God's incarnation as a hermeneutic axial point. The Eucharistic motif had now become the foundation of a representational model articulated around the theme of "real presence." While the extent to which seals became effective in representing their owners owed much to this Eucharistic debate, the principles and modes of their operation as a sign of identity may also be situated in prescholastic ideas about the nature of personhood, since it was part of the new semiotic conception that a sign be representative through its capacity to embody the ontological characteristics of its referent.	46
Among the conceptual tools chancery scholars used to address the issue of personhood was the seal as metaphor. I find it suggestive that the same prescholastic milieus that promoted changes in semiotic thinking, that entertained concerns about representation, authority, and personal identity, and that produced the novel medium of the sealed charter as a solution to these concerns are the very ones that resorted to the seal metaphor to clarify these concerns. There apparently was no precedence of the metaphorical seal over the documentary seal, and there may be little advantage in trying to explicate one by reference to the other, but it is undeniable that both cover the same semantic territory, organizing and thereby elucidating contemporary views of identity. In the spheres of both discourse and practice, the seal, linking the divine and the human, was centered precisely on persons, their agency and representation, and their personal relationships to others, to God, and to script.	47
The seal metaphor was not new in Christian discourse and liturgy of the eleventh and twelfth centuries,87 but its semantic range was now extended. Seal metaphors facilitated discussions on the relational presence of the divine persons—Father, Son, Holy Spirit—within the Trinity, of the Son in Man, and of the Son in God the Father. Such metaphors were used particularly in discussing image and resemblance, first between the Creator and his Son, who was engendered and not created, and second between the Creator and his creature, the human being. As the body of seal metaphors is vast, I will present and discuss here only a few representative examples.88	48
When Abelard wished to demonstrate that the Trinity can be discussed in logical terms, he identified as a principal conundrum the question of unity (the Godhead) in diversity (the three persons of the Father, Son, and Holy Spirit), and in proceeding to address this question articulated the main thrust of his theological argument through a seal metaphor:	49
Identity and diversity may be described in five, and perhaps more, ways. There is identity if a thing exists entirely with another thing, that is, by essence and number. There is identity secondly, in property; thirdly, by definition; fourthly, by likeness; and fifthly, by incommunicability, when a thing never changes into anything else. We can say things are identical in these five ways, and by contrary we can say that they are diverse in these five ways; that is, if the conditions of identity are not fulfilled then the things are diverse . . . Things may be identical in essence and number, but not identical in property or proper character. This may be the case even when their substance is the same, their proper functions alone making a fundamental distinction between them. A wax image, for instance, may be identical in essence and number with the wax of which it is made. But there is no interrelation between the proper character of wax which is one thing, and the proper character of an image, which is another thing.89
Building on his demonstration that the wax and the waxen image are essentially the same but not the same by property and definition, and reusing the same metaphor, Abelard demonstrates the simultaneity of the identity of the triune God and of the difference between the persons of the Father and his begotten Son.	50
Look at a waxen image. Consider that in it is the mixture of wax: that is, the wax itself as substance. From this wax, the image becomes, in philosophical language, materialized out of material. The same essence is both the wax itself and the wax image. We can predicate of the wax that it is the image, and of the image that it is the wax. Nonetheless, it is also true to say that the waxen image is from the wax. But the wax is not from the waxen image. The wax itself is, however, the material of the image. The waxen image is not the material either of the wax or of itself. Again, we can assert that the image was realized out of the wax of which it is composed. Yet neither the wax itself nor the image itself were composed simply out of the image. Now if we take these names of wax and waxen image absolutely, not relatively to one another, we can assert anything of them that will be true of both because the substance is identical. I mean, for instance, if the wax is yellow and the image an upright figure, then the thing is yellow and upright throughout. If, however, we take the names relatively, in respect, that is, of the generation or composition of the waxen image, thinking of them as the material and the thing materialized from this material, as cause and effect, or the begetter and the begotten, then we cannot link them in respect of their particular functions by a predicating adjective. We cannot say that the material is the same as the thing materialized from it. Apply this comparison to the divine generation and my position is clear. God, the Father, is the divine Power; God, the son, is divine Wisdom. Now divine Wisdom is a kind of power, since it is the ability to discern and foresee and deliberate aright against anything that may deceive God. Hence divine Wisdom coming from divine Power is a sort of waxen image out of wax. Philosophically, it is a species of genus. The species is the same as the genus, as a man is the same as an "animal," or a waxen image the same as wax. The wax image is from wax as man is from animal. I mean that, in so far as it is a wax image it must be wax, just as in so far as a man is a man, he must be an animal. But the contrary is not true. Power, therefore of discernment and doing all kinds of things may be considered like wax which has potentially either to be a wax image or anything else: or, as the animal species, which may be a man or any other animal. This is my illustration to show that, when the son is begotten of the Father, I mean that divine Wisdom is from divine Power as I have explained.90
Both these passages clearly articulate a concept of identity as a principle of sameness and also a product of the polarization between similar and dissimilar, a concept of property (that is, definition, or proper character) as that which both characterizes and distinguishes the person. The seal metaphor in these passages specifically addresses two points. First, there is priority of the material (or substance or essence) over the image. Second, there may be diversity by virtue of definition (or property) when things are identical in essence and number. The prescholastic semiotic of mimetism afforded not only an economy of signification but also a differential principle of being. It defined a human person as existing by virtue of relationships of origin, as identical in the sense of its similarity to humanity (species) but distinct with respect to properties in relationship to others. Yet it was neither perfect identicality nor absolute distinctiveness but rather comparative likeness—difference in essence, number, and properties—that was emphasized. Human personhood and identity were thus formulated both in relation to God (essence) and to other human beings (number and properties). As such, the concept of the person that developed in the twelfth century, modulating likeness to reveal heterogeneity, was of a unique psychosomatic unit expressing a distinct identity as both flesh and spirit, capable of representation for the purpose of activity in the world.91	51
Prescholastics, in their ontological exploration, privileged an exegetical approach that, borrowing from Neo-Platonic readings of Genesis, presented the human being as created in the image of God so as ultimately to be transformed into his resemblance.92 In this sense, identity consisted of a God-like image within the human fabric. Here, the metaphor of sealing was recurrently used to evoke the imprint of the divine archetype on the human raw material. Commentaries on Genesis 1: 26 (God made man in his image and likeness) from the School of Laon, from Abelard, and from the canons of St. Victor contemplated just how human beings might be said to be "in the image and likeness of God" when they have no common property with God. Using the seal metaphor, the commentators determined that the human soul in God's image is different from the Son who is in God's image, in proportion to the difference between the king's image on a seal and the king's generated image in his son.93 Only the engendered image (the Son), which shares properties and is consubstantial with its model, may be equal to it: only the Son is the image of God. The created image (Man), on the other hand, bears only an analogy to its model: the human being is in the image of God. Abelard and the School of Laon were concerned, however, to reconcile transcendence and immanence, and so insisted on the presence of God within the begotten Son and, through the Son, within the created human being as well. Here again, Abelard and the Laon scholars resorted to another seal metaphor, this time involving the die, its image, and its imprint. God is the seal's inherent material (the substance of its die or matrix); the Son is the figure of God's substance, the image of God engraved in the matrix, which in turn imprints itself on the human soul (reason, heart, memory),94 enabling that soul to be configured as the Son. In this sense, the human being is created as an image, imprinted through the medium of divine substance but sharing no substantial affinity with it, unlike the Son, whose image is consubstantially figured of divine substance. The human creature, conceived as sealed and therefore as replicated image, is ontologically constituted to participate in its informing prototype, capable of tending toward the prototype's realization. In terms of seal metaphors, human identity is about creation, impression, oppression, and reformation. Creation is the process by which Man is made in the image and likeness of God (Genesis 1: 26). Impression, that is, the soul formed and signed by the seal of God, expresses the human capacity for good. Oppression, that is, an opposition to or the breaking of God's seal through Man's sinfulness, involves dissimilarity and alienation.95 Reformation presents the hope that likeness to God is an end capable of human accomplishment. Personal formation and reformation are fundamental processes of human identity that Hugh of St. Victor, among others, discussed, resorting frequently to the seal metaphor, as in this striking passage from the De institutione:	52
In good men the form of the likeness of God is engraved, and when through the process of imitation we are pressed against that likeness, we too are molded according to the image of that likeness. But you must know that unless the wax is first softened, it cannot receive the form, and this also, a man can not be kneaded to the form of virtue through the hand of another's actions, unless he is softened and all pride and stiff-necked contrariness removed . . . Why do you think we are enjoined to imitate the life and conduct of good men, unless it be that by imitating them we are reformed to the likeness of a new life? For in them the form of the likeness of God is expressed, and when we impress ourselves on them through imitation, then we too are reshaped according to the image of that same likeness.96
Paralleling their seal metaphors, the prescholastics who were fostering the new semiotics displayed in their own chanceries a predilection for visibility centered on the concept of an imprinted image at once generated by the principles of likeness and linked to a model. In non-royal charters, the motif of visibility had previously engaged only a single modality of representation, the symbolic, constructed by linguistic signs arranged as a discourse. With seals, a second, iconic modality was introduced, where representation was achieved by lines and figures arranged as images. In fact, the linguistic and iconic modes were both present on the seal itself—the legend (text) and the type (image)—but the essence of their representative power came from their being produced as imprints. That a seal represents by being an object whose marked matter has become graven form is crucial in terms of prescholastic semiotics. The seal metaphors previously discussed suggest that an imprint, by virtue of containing the trace of an origin in its very matter, is a sign forever indicating a radical presence, for instance, that of God in human beings. The very act of seal imprinting both articulated and dramatized these principles of marking origin and materializing presence. Sealers sometimes went so far as to impress parts of their own bodies on the waxen seal: toothmarks, fingerprints, bits of hair or beard.97 In the very act of impressing die on wax, the seal blended with its referent (the sealer), the written text with its enunciating subject (again, the sealer). In terms of prescholastic ontology, both seals and sealers were imprints carrying within their very matter the mark of an original. The seal, thereby participating in an existential relation with the sealer it represented, became an efficacious sign, a power. Thus was the seal enabled to confer on the document its own authority, transforming the document into a monument, which is the name by which sealed charters came to be known during the twelfth century.98 In a manner analogous to Hosts imprinted with a cross, the letters IHS, and, from the twelfth century onward, a crucifixion scene or the lamb of God, seals not only mediated but embodied the real presence of the individuals who affixed them. Seals allowed simultaneous presence and representation. Their mode of signification was through incarnation. The ritual process of sealing also involved a transformation of substance: it fused two quite different spaces, the locus on the parchment where the affixed seal affirms that ego was there and the physical location where the documentary sealing took place in the presence of witnesses. Above all, sealing changed a written leaf of parchment into a monument. This occurred by authorizing writing, that is, by incorporating the author into the text. Seals were the incarnation of the ego of diplomatic discourse, marking the charter so that it acquired substance and body. However, although seals and the Eucharist participated in a common semiotic logic, seals fell short of sacrality. Their relationship to script occurred at the lower margin of the page: the ego of the author-donor-sealer and his mark are not so much within the text but in consubstantial relationship to it.	53
Seals represented individuals and, by personifying their owners, personalized the written word. From a graphic viewpoint, however, there is a tension in seals between individualization and categorization. The text of a seal's legend contains the individual's baptismal name but also both a title (king, count, bishop) and the entity or group ruled, underscoring the fact that identity was articulated primarily around function and its territorial or ethnic circumscription. The legend is obviously the part of the seal that individualized its owner. The image of the sealer placed on seals was anthropomorphic, though not a realistic portrait. I described earlier how the ritual of sealing, of imprinting, was itself significant in achieving presence and representation and was often enhanced by bodily marks as part of the imprinting process. Yet it was also true that a donor might utilize another person's seal to seal a charter given in his own name; the text in such cases would duly record the act of borrowing the seal, whereupon the document produced was considered properly sealed and authorized.99 This specific manipulation indicates, in my opinion, that the generic gesture of sealing was also effective in committing and representing an individual, as might be expected of a bodily participation within a ceremonial culture. It also points to the importance of the spirit, that is, of the intention to seal, the animus signandi; for intention was the seal's intellectual and spiritual element, an important part of both seal and sealing. Intention was made explicit in a clause within the document's text announcing the affixation of the seal, as well as through the personal gesture of sealing.100 The referential category engaged by seals and sealing is, therefore, a physical person who is ethical and accountable, and endowed with personal intentionality.101	54
Depictions of the body on seals are, as I have noted, nonrealistic, which is not to say that they did not function as a form of portrait within the medieval rules of figuration. Realism is, after all, simply a convention, and one that the Middle Ages did not equate or associate with physiognomic likeness.102 In the charters themselves, authors refer to their seals as their own image, imago noster, which reveals that seals and their depictions incorporated elements meaningful to self-representation.103 Realistic physiognomy was not privileged; emblems of function and symbols of kinship were. Kings were shown in royal garb and posture, nobles as warriors, and bishops in episcopal array. Heraldry, from the mid-twelfth century onward, served as an iconographic rhetoric that expressed the identity of a kindred in relation to other groups, to its own land, and to its separate sub-branches. From an iconographic viewpoint, seals may be said to display abstracted figures and iconic types. Abstracted figures on seals refer to a conception of the individual as exoteric, someone who must be seen and decoded. As iconic types, seals display a severely limited, barely differentiated repertoire. Seal iconography thus affected the formulation of personal identity in that, through modulated differences of posture, costume, and emblems, it established and published a lexicon of images that classified and limited the contingencies of individual identity. By linking each individual to a formulaic icon, seals tended less to designate singularity than generic conformity to a group; indeed, they functioned as an index of shared membership in specific groups.	55
King Louis VI confirms an exchange of slaves between the abbeys of St. Geneviève and Notre Dame cathedral of Paris, 1118. Archives Nationales, Paris, AN K 21 n° 13 4-AE II 132.
Formulaic icons thus suspended individual referentiality, conferring on seals the status of a system. The text of the legend particularized a given seal, giving it the status of instance. Thus seal graphism generated personal identity through a grammar that articulated the organizing principles of society. In this way, personal identity was defined and produced as an instance of social order, and thus produced itself as the verifier of the system it substantiated. The medieval sense of identity was about resemblance: the person as sign signaled that signs of representation were in conformity with social reality. This sense of identity parallels what is conveyed by the seal metaphor: the self as seal impression. The seal was the form, and the resultant personalized individual was a likeness.	56
Seal metaphors and seal graphism were not alone in projecting this concept of identity. The element of likeness was intensified by the technique of sealing, which involves duplication. Every seal impression in wax from a specific matrix was identical. The seal's competence and significance was, indeed, predicated on replication. Seals, bearing conventional images and acting through replication, did not emphasize distinction so much as likeness. The element of likeness was also heightened through the very modes by which seals presented themselves as representative of their owners: the seal bore and was his owner's image, his imago. And the seal owner, as the object of representation, himself became an image of sameness, a warranted replica.	57
The identities of the individual and his seal depended on their capacity to resemble a model. In its operating and metaphoric principles, the seal was associated with transcendency (God) and at the same time also partook of the properties of its referent, an individual. The seal—operating through the medium of its progeny (the impressions), through its creative capacity, through its power of becoming (the impression), as well as simply of being (the matrix)—was experienced in analogy to the life process. On the mechanism of seal operation, the individual could project the autonomy of his conscience (we have seen the importance of intention), his ability to control the idea of his person. Mechanization and personalization are not contradictory. Individuals and seals became reciprocal models. Seals, conforming to and informing the logic of prescholastic semiotics, derived their capacity for signifying from their perceived affinity to, and agency within, human biography. Thus seals were successful as objects denoting both identity and authority. They produced identity as a foundation for documentary authorship, authority, and, ultimately, authentication. The notion of identity as likeness and replicable resemblance, as it came to be conceptualized and realized through seals, was to affect more generally the fabric of social life.	58
With the diffusion of sealed charters in the eleventh and twelfth centuries, human beings bounded by flesh and consciousness were now engaging in strategies of deferred representation so that, where they had previously operated as their own empirical self-representing agents, they now came to coexist with, indeed relate to, a "double," their representative image (imago). This double, which functioned as if the other (the human absentee) were both present and identical, was an object, the seal; reciprocally, the seal signified the individual, who thus came to be newly mobilized as a locus for imparting permanence and authority to the written word. Such mobilization was therefore achieved by means of representation conceived both as replicate presence and as objectification. These two processes had radical effects on the notion of the individual.104	59
In the course of embodying the linguistic ego of a charter together with the physical presence of its individual referent, seal and imago veered away from personal expression and toward stylization. Seals empowered not the individual as particular being but the person as category, the person as representative. The graphic logic of seals established a crucial distinction between the individual of flesh and character and the individual as an impersonation of social roles specified by codes. The particular living individual of earlier oral ceremonies came to be increasingly abstracted as an incarnation of a particular soc