Corps de l’article

I am honoured to be able to contribute the two texts below to this journal’s survey of semiotic approaches within the discipline of anthropology. Delivered originally as “performance” pieces, these two texts reflect on the methodological implications of semiotic analyses from two other fields of inquiry – classics and medieval studies – that continue to have enormous relevance for anthropology. The first was presented as a formal response to Brigitte Miriam Bedos-Rezak’s lecture titled “Imprint : Ontology and Christian Theology in the Western Middle Ages”, which was given as the keynote address at the symposium “(Re)constructing Religions : Evidence, Methods, and Disciplines” held at Brandeis University on October 23, 2003. Ever since her groundbreaking 2000 essay on the semiotics of medieval identity, Bedos-Rezak, now at New York University, has almost single-handedly advanced a “semiotic anthropology” of the high Middle Ages, particularly the period just prior to the systematization known as “scholasticism”. An authority on the theological and political dimensions of the practice of “sealing”, she has produced nuanced readings of the historical trajectory of various sorts of practices and ideologies that constitute a veritable “culture of the imprint”. My response to her lecture focuses on a Boasian question, namely, the difficulty of discovering a history of social practices from the writings of elite or “esoteric” agents of these same practices. Readers of this journal will want to study with care Bedos-Rezak’s most recent expression of her semiotic approach to the Middle Ages in her 2010 book When Ego Was Imago.

The second text is a “welcome song” performed at the beginning of a highly productive meeting of the Symbolic Form Study Group at Brandeis University on May 3, 2006, when our “semiotic circle” was joined by Gregory Nagy of Harvard University. Thanks to the collegial intervention of Leonard Muellner, Nagy agreed to lead a general discussion of the relationship between text and society, with particular reference to the Homeric epics. A scholar of all aspects of ancient Greek poetics, Nagy has developed a number of concepts – especially his notions of “diachronic skewing”, “multiformity”, and “poetry as performance” – that have important consequences for recent anthropological efforts – particularly by Richard Bauman, Michael Silverstein, and Greg Urban – to expand the notion of “textuality” to include forms of oral discourse. The key “political” point of my remarks was to bring into conscious articulation Nagy’s brilliant work on the semiotic qualities of the Greek epic tradition and anthropological work in the tradition of the “natural histories of discourse”.

The two texts below are presented unchanged from their original delivery. I hope that, being “entextualized” here for the first time, anthropological readers will be informed that significant semiotic work in classical and medieval studies speaks profoundly to our own subdisciplinary concerns.

First Encounter : Methodological Reflections on Medieval Metaphors of Semiotic Mediation

Robert Fitzwalter, an imperious early 13th century English earl, attempting to lease a valuable wood from the Abbey of St. Albans, colluded with one William Pigun, a resident of the abbey. William noticed that the abbey’s official seal, kept in a box with various documents and charters, was carelessly looked after. And so when Robert sent him a forged charter William “furtively and rapidly sealed it under his sleeve; though done hastily, yet [the impression] was amazingly clear, elegant and exact, though the seal had been snatched away by trickery”, so reports the chronicler Matthew Paris (1986 : 17). Fortunately the equally clever Abbot John deduced that the charter was a forgery and that it was an inside job; William, his trickery being revealed, was banished to a solitary cell in Tynemouth. From then on the seal, now kept under several keys, was brought into the chapter whenever deeds were drawn up and sealed in everyone’s presence, and then returned to a strong chest in the safest possible place.

In Abbot John’s dying days a delegation of monks (including Alexander of Appleton, the “bearer and guardian of the seal”) brought the charter into his presence to argue against the abbot’s practice of arbitrarily exiling monks to distant cells. But the abbot refused to seal the charter; and the members of the delegation took his silence for assent and sealed it themselves. The abbot died three days later, and so the charter seal was broken and then resealed by no less than Stephen, the Archbishop of Canterbury. Alas, the monks of St. Albans enjoyed their chartered protection for only a short while, since almost immediately after his installation as the new abbot, William of Trumpington – a buddy of the King – refused to honour its terms, prompting a visit by the Cistercian papal legate Nicholas. Upon reading a copy of the charter the legate denounced the monks for renouncing their vows of obedience. “And with these words […] he tore the charter into pieces with his front teeth, and, smashing the seal attached to it, threw it onto the floor” (Paris 1986 : 37).

I retell these tales to reinforce the point that seals were principal semiotic mediators in the life of medieval elites and that, by the beginning of the 13th century, the normative role of seals as “signs of history” had already become manipulated as “signs in history”, to use phrases I have applied to the engagement of historical markers in the dynamics of social life (Parmentier 1987). That is, seals not only represented the identity and authority of the absent sealer but were also focal objects of social praxis.

Reflecting on these stories, stories that describe events over a hundred years after the Pre-scholastic period focused on in Bedos-Rezak’s papers on medieval semiotics, suggests several issues related to the general themes of forms of evidence and methods of reconstruction in historical semiotics. Sealing is a social practice that leaves little evidence of its contextual and processual aspects; what remains physically are seal matrices, wax imprints, and copies or casts of both, each providing information about the substances involved, such as gold, bronze, wax, etc., and the inscribed words accompanying the figure. Seals are additionally, as Bedos-Rezak shows, the object of metasemiotic discourse, as writers refer to seals and sealing directly in historical narratives and metaphorically in theological texts. In her scholarship over the past twenty-five years Bedos-Rezak (1993, 2000) has rightly insisted on the necessity of studying the iconography of seals and the discourse about sealing in order to produce a semiotic account of medieval society as the “culture of the replica”.

Thanks to Bedos-Rezak’s research, medieval sealing joins the ranks of material technologies that have been identified as providing semiotic metalanguages in culture : weaving for ancient Greeks, metallurgy for the Inca, pottery for Native Americans, and printing for early modern Europe. In each of these cases, aspects of the technology stimulate metaphorical association and philosophical speculation that, in turn, offer the analyst a privileged “ethnosemiotic” window for reconstruction.[1] While not doubting the importance of seals, weaving, pottery, metallurgy, and printing, I want to suggest that the analyst should always be careful not to overlook alternative metasemiotic metaphors[2], especially alternatives that seem to form a paradigmatic set at a given period of cultural history. In the medieval period alternatives would include “covering” (integumentum) (Stock 1973 : 50), macrocosm/microcosm analogies, the mirror image, organic bodily metaphors (Le Goff 1989; Bynum 1991 : 254), hierarchical orders (Duby 1980 : 66-69), the “Porphyry’s Tree” (Vance 1987 : 86; Piltz 1981 : 56), light as divine emanation (Panofsky 1979 : 21-24; Eco 1986 : 43-51; Duby 1981 : 97-135), the book (Gellrich 1985 : 41; Petrucci 1995), and typological transfer (translatio) (Nichols 1983 : 20). And sealing seems to be a member of a paradigmatic subset having to do with molding relatively plastic substances, such as minting coins, pottery, sculpture (Morrison 1990 : 65), model construction (Abelard in Tweedale 1976 : 171), and stamping on an anvil. So the twin methodological tasks become, first, to find contextual regularities governing the use of all these alternative metaphors and, second, to deduce the “intensional” core (that is, the specific meaningful features) of paradigmatic sets. (Note that Bedos-Rezak argues that “imprinting” is the key to the plasticity subset.)

Standing back from these objects and texts we can see that the political use of sealing is a perfect exemplification of the triadic process of Peircean semiosis – whether or not anyone in the 12th century realized it as such! – in which, initially, the sealer (normally a person with some authority who directs that a seal be affixed) creates a seal matrix (consisting of a metallic substance and some unique inscribed design) capable of impressing into wax (or some other soft medium) an impression of the matrix design (often leaving some material residue of the very sealing process), in order that the document so sealed will be interpreted (at some removed temporal and spatial context) by an interpreter as having authenticity and, thus, as conveying authority grounded in the transmittal of the sealer’s identity and, thus, power. In more technical Peircean vocabulary, a medieval seal (matrix) operates as a dicent indexical legisign generating identical imprints, which are rhematic indexical sinsigns of the special class called “replicas”. (The matrix potentially generates many identical replications, so, while as a real thing it is not a true Peircean legisign, it operates like one.) The semiotic “ground” between seal matrix and original sealer is indexical because the sealer actually presses (or has pressed on his authority) the seal into the wax. And the sealing process is dicent – that is, intended to be interpreted as authenticating a specific document authorized by the presence of the sealer at some previous definite moment – although this is made possibly largely on the basis of the similarity of the seal impressions stemming from a single authority[3].

That seals simultaneously engage iconic and indexical modes of signification provided medieval thinkers with material for more general theological and philosophical reflection. The seal, for example, mirrors the triadic commonplace that God creates the human soul through the imprint of the Son, who is the figure of God’s substance[4]. Abelard then brilliantly realized what Bloch (1983 : 35) calls “imbrication of signification and generation” in postulating a parallel triad for language : verbal signs (verba) standing for signified objects (signatum or res) and producing significant concepts (dicta or sermones) or what are “said about things” in the mind of the listeners.

Critical to Bedos-Rezak’s account of the medieval semiotic landscape is the argument that the generalization of this triadic semiotic model provides elites with an intellectual explanation for the “presencing” of divine realities in human experience. In the context of any dualistic religious worldview that strictly separates transcendent, eternal, and unknowable divinity from the mundane, temporally bound, and changeable world, mediating processes and phenomena have been regularly noted – Lévi-Strauss’ (1974) study of rain and mist as mediators in Pueblo mythology is the classic example. In the specific context of the medieval period the problem of mediation was often expressed semiotically as bridging the gap between experienced signifiers and divine signifieds. In other words, the issue of the immanence or presencing of transcendence parallels the issue of what Saussure calls the “motivation” of signification and what Peirce calls the “ground” of the sign. Anthropologists are justly wary of crude typological generalizations about cultures grounded in some “mystical participation” in which the levels of sign and referent are blurred because of some evolutionary cognitive deficit (Berger 1968 : 282; Parmentier 1997), but the problem I want to raise here is the historical dimension of mediation.

If it is true that mediation was especially problematic in the 12th century due to the simultaneous expansion of monastic spirituality and the birth of a science of nature, then several hypotheses can possibly explain the proliferation of imprinting metaphors (Grant 2001; French and Cunningham 1996). Given that these metaphors occur regularly in the context of theological reflection on the Trinity, the Incarnation, and the Eucharist, it might be, first, that the seal metaphor spread because it provided a metalanguage to understand these three privileged mediational moments as a paradigmatic set, as evidence for the normative “presencing” of the signified in the world of experience. Second, it might be the case that all this theological focus on semiotic mediation reflects a crisis in the religious life of elites of the period, as a defence against the growing spiritualization of natura (called a “genitive force”, a “crafting fire”, the “mother of all things”, and the “instrument of divine operation” [Gregory 1988 : 64]), that is, a mediation from the ground up. Alternatively, it spread as a way to counter the declining power of the Platonic “dialectic of participation” (Gregory 1988 : 74) or the Augustinian linkage of likeness and participation (Bell 1984); or as a hyper-rationalized account of divine presence that maximally distances the Schoolmen and monks from the illiterati (“unlettered”) or idiotae (“simpletons”) (Gurevich 1992 : 97) for whom relics and images – and miracles and magic – provided a steady stream of opportunities for “blending the spiritual with the physical world of events” (Gurevich 1985 : 85; Brooke and Brooke 1984), like a child, notes Anselm, fearful of a sculpted dragon, who “does not yet know how to distinguish clearly between a thing and the likeness of a thing” (Anselm quoted in McKeon 1930 : 1, 161).

Still another possibility might be that the expansion of sealing practices and metaphors signals the actual distance or even absence of the sealer (God or king) whose presence, transparent in an earlier world of “orality”, can only be created though semiotic projection (Vance 1986 : 52). As Jaeger (1994 : 347-348) has recently argued, the plasticity and humanism of Gothic art might be a nostalgic representation of a period of moral discipline that has already passed. The methodological consequence of this argument from “envy” is frightening : if there can be such a temporal disjunction between iconography and mentality, how can the analyst ever trust any kind of semiotic data as positive evidence for the operation of cultural categories?

Many scholars have pointed out the increasing reflexivity and systematicity of 12th century religious thinking and artistic programs as a prelude to the monumental summative works of the next century (Southern 1995 : 49-58; Colish 1988). At this time everyone seems to be asking a “superfluous novelty of questions” (Goswin of Mainz quoted in Jaeger 1994 : 367) about everything. “Even on public streets”, writes Stephen of Tournai, “the indivisible Trinity is taken apart and wrangled over” (quoted in Chenu 1968 : 294). Anselm even asked the self-reflexive question : whether a grammarian is a substance or a quality! (Colish 1983 : 76). Systematicity anywhere in culture outside language immediately summons the spirit of Franz Boas, who used religious phenomena to illustrate the difference between fundamental or primary cultural categories and secondary elaborations generated by esoteric specialists. What a people consider independent objects and what they consider aspects or attributes of objects varies cross-culturally; some people, for example, view the luminosity of the sun to be an object the sun can put aside. What if 12th century theological reflection on the doctrine of the Trinity is really just a secondary elaboration attempting to explain the independence and/or identity of attributes of divinity?

Furthermore, Boas argued that it is often the historical fact of breaking or violating relatively unconscious or deeply patterned cultural associations – often stimulated by acculturation or borrowing from neighbors – that results in the coming into consciousness of these patterns, making the categories and associations open to rationalization and eventually to systematization. Boas seems to describe a medieval Schoolman and his seal as he “ransacked the entire field of his knowledge until he happened to find something that would be fitted to the problem in question giving an explanation satisfying to his mind” (quoted in Stocking 1974 : 254).

Two methodological paradoxes are implied by the Boasian argument : first, the more substantial the textual evidence of, for example, the seal metaphor, the greater the probability that this consciously articulated discourse conceals the fundamental and integrating classifications of a culture; and, second, the greater the systematicity of, for example, theological doctrines, the greater the probability that this achievement reflects the work of an “esoteric specialist” far removed from the operating principles of the culture – described reflexively by Abelard, whose “excellent knowledge […] cannot be attained by long study, but only by genius”.

Second Close Encounter : A Welcome Song for Gregory Nagy

Let us now turn our attention back about twenty-eight hundred years to the period of the entextualization of the Homeric poems and, specifically, to the contributions of Gregory Nagy. In 1998 another Homeric scholar Richard P. Martin wrote, “Cultural anthropology will mold the shape of classical studies for this generation[…]. To be fully and honestly philologists, we must now learn our Geertz along with Greek, absorb Lienhardt as well as Latin, undertake ethnography after epigraphy”(1998 : 108). Those of us here who share cultic identity with Geertz and Lienhardt certainly appreciate Martin’s comment, but an inverse remark by Nagy is perhaps more directly appropriate for our discussion : “Ironically, the field of anthropology has as much to benefit from the currently construed field of Classics as the other way around” (1992 : 23). Let me try to briefly sketch at least one of these benefits.

Nagy’s account of the diachronic stages of the entextualization of the Homeric oral epic is the best exemplification I know of the analytical model for the anthropological study of texts proposed by the authors in Natural Histories of Discourse, a collection edited by Michael Silverstein and Greg Urban (1996) – although I am not aware of any reciprocal citations. As heir to the Parry-Lord discovery of the orality still evidenced in the fixed Homeric canon, Nagy honours that intellectual heritage by exposing its fundamental weakness : not taking oral performance seriously enough to realize that the texts of Homer that we have don’t just evidence the contexts of their performance but are fully constituted by those contexts. Parry and Lord saw vestiges of orality in the systematicity of formulaic expressions fixed in specific metrical contexts; and they accounted for the internal coherence and Panhellenic uniformity of the two poems by postulating an enregistering moment of dictation whose “external impetus” is the exact opposite of performance in context. Nagy, on the other hand, rethinks the whole matter. He locates the coherence and uniformity of “texts” and “textuality” in the diachrony of their contextual enactments and in the centrality of their focal performance sites. The textual uniformity of the Homeric poems even at the spatial extremities of their diffusion is the result of parallel performance contexts, not of the replication of text-artifacts (Nagy 2001). That these texts were eventually written down in a Phoenican-derived alphabet and then subjected to “cultural consolidation” (Nagy 1997 : 178) does not compel us to take this historical contingency as an all-powerful retrodictive interpretive key.

Alas, Nagy’s position actually makes the whole analytical task much more difficult, for the simple reason that the two epics don’t seem to reflect in any obvious or direct way either the societies or the contexts of archaic eighth-century Greece. The world represented in the texts is the Mycenean period four hundred years earlier, and the texts’ very monumentality makes it unlikely that wandering singers went around performing the complete Iliad and Odyssey at festivals and competitions. But Nagy turns these challenges into weapons in a set of precisely interlocking theoretical moves involving what Silverstein names the “pragmatics” of the “metapragmatics” of texts, and helping to re-insert what Bauman (2004a : 128-158; cf. Nagy 2004 : 43) labels “mediational relays” as the cultural mechanisms for the constantly changing enactment of tradition.

While it is both true and interesting that these poems are richly self-reflexive – that is, contain numerous examples of explicit and implicit metapragmatics or discourse about discourse – Nagy (2003) postulates in his concept of “diachronic skewing” an encompassing disjunction between the texts’ reflexive metapragmatics and the pragmatics of texts as performed. In many publications Nagy (1990) brilliantly explores, for example, the ethnosemiotics of the conventional index (sêma), signs requiring mental acts of recognition (nóēsis). Note, by the way, that the Homeric sêma is a perfect Peircean Third : Odysseus is the semiotic Object, the scar is the Sign, and the recognition is the Interpretant. More relevant for the study of textuality more globally are the metapragmatic references to weaving (Nagy 2002 : 79), sewing, the chariot-wheel and carpentry in which the poet creates the cohesive and unified fabric of song (Nagy 1995 : 179, n. 122). The poet himself is metaphorized as a varied-throated nightingale; the resourceful hero is likened to the performer of poetry who adjusts each performance to the needs of the local context; and even the name “Homer” etymologically exemplifies a semiotics of coherence. Furthermore, embedded in the epics are accounts, “mirroring” real-world performance events, of singers taking turns at song, storytellers reciting epic tales to enthralled audiences – indeed, telling the very tale of Troy – and other contextually-anchored speech-acts such as praising, blaming, prophesying, lamenting, and praying.

It would be a serious mistake, Nagy implies, to read off from these textual metapragmatics the “natural history of discourse” – that is, the actual pattern of mediational relays (e.g., rhapsodic sequencing) that enables these oral poems to crystallize into the Homeric epics – since in almost every moment and at every level the Homeric testimony “belies” (1990 : 24) both the history of their entextualization and the contextual parameters of their performance. The creation of the poems is “viewed as happening at a remote point in time, not over time” (1996 : 76). As Nagy (1979 : 16) puts it, “The immediacy of performance […] is counterbalanced by an attitude of remoteness from composition”. In a related discussion of sympotic songs he echoes, “The song proclaims that its own unchangeability is a prerequisite for its own perpetuation” (2004 : 29). (The comparativist John Miles Foley [1999 : 20] makes a related point : “the referent of the concrete signs in the performance or text lies outside the immediate performance or texts”.) Perhaps the fact that the Iliad and the Odyssey stand in complementary distribution might be seen as a global refusal of intertextuality.

Now this strategy of radical decentring – the ideological regimentation of a text as being distantiated or decontextualized – is typical of many authoritative cultural texts in the archaic world and, again ironically, seems to correlate with an “ideology of mimesis as re-enactment” (Nagy 2004 : 27), with an esoteric hermeneutic in which power is demonstrated in the privileged discernment of hidden signs (Ford 2002 : 74), and with a mythological surround that presents a “coherent system” (Muellner 1996 : 51) precisely at those moments of normative violation requiring immediate recalibration. Ritual performance is culturally powerful to the degree that contingencies of performance are masked. As Nagy cites Homer, “the poet only hears” (Iliad ii.485).

Clearly the political function (that is, the pragmatics) of performing epic poems changes through time, especially in this period of massive social upheaval prior to the emergence of classical Greece. Songs about Troy sung as hero tales before royalty in Mycenean citadels, sung in dispersed and autonomous city-states for aristocratic pleasure and reassurance, sung by professional rhapsodes in Panhellenic competitive festivals on behalf of tyrants “diligently searching through the legendary past to find precedents” (Snodgrass 1980 : 70), and sung by school children as feats of pedagogical memorization are all radically different in their encompassing pragmatics. We might say, then, that there is a diachrony to diachronic skewing! And this vital lesson about the changing pragmatics of texts reinforces the related (and entailed) methodological necessity of being constantly attuned to the revaluation of marked asymmetries in the cultural categories within texts. As Nagy (2003 : 40), following the master Kuryłowicz, writes, “A diachronic perspective reveals a shift in meaning from category to subcategory or from subcategory to category”.

“Benefit” to anthropology? Context; diachrony; comparison; textuality; tradition. What’s not to like?