Erasmus and the hermeneutics of linguistic praxis.

Erasmian hermeneutics are notoriously difficult to describe clearly because Erasmus is always looking in two directions at once -- both toward the ideal, perfectly expressive Word and toward the multitude of imperfect, human words caught in the tumult of history and transmission. In the Enchiridion (1503), a relatively early work, he argues that words inevitably fall short of their task of miming the Logos, that the smallness of the manna rained down on the Israelites in Exodus 16 "signifies the lowliness of speech that conceals immense mysteries in almost crude language."(1) Erasmus believes in an essential connection of some kind between res and verbum, but it is clear that he holds as well to the Platonic view that this connection is always necessarily inadequate, that there can be an approach but never an arrival at complete meaning through human language.(2) Thus the multiple levels of meaning present in Scripture should be understood as a function of its immeasurable fecundity rather than a token of any ambiguity.(3) And for the present, until the end of history unveils the full meaning of the Word, Erasmus requires that we be content to "adore with a simple faith and from a distance" what cannot be understood no hermeneutic practice, he implies, can wholly rehabilitate the split between res and verbum.(4)

At the same time, the lofty linguistic goals put forth in Erasmus's pedagogical works imply a formidable confidence in how much language can do if the proper techniques are achieved. Despite his persistent reservations about the insufficiency of language in relation to the divine, textbooks ranging from the programmatic De ratione studii (1511) to De pueris statim ac liberaliter instituendis (1529) became the basis of a humanist education that largely determined how readers approached Latin reading and composition for the rest of the century and beyond.(5) Common to all of these texts is an emphasis on performance and the building blocks of performance: practice is preferred over rules; method over specific content; and an accessible, manageable organization of resources over memorization. Other Erasmian works, such as the Parabolae sive similia (1514), Apophthegmata (1531), and the various editions of the always expanding Adagiorum collectanea (1500) and Colloquia (1518) provide endless grist for the linguistic mill that the first group of texts sets in motion. The popular De duplici copia verborum ac rerum (1512), at once a program of exercises and a catalogue of copiousness, seems to fall somewhere in the middle by demonstrating the best method for generating one's own linguistic grist. Overall, an Erasmian curriculum created a (Latin) speaker among other (Latin) speakers, a speaker qualified to interpret texts, produce commentary, translate, and speak and write extemporaneously -- a man (usually) of considerable linguistic action.(6)

In Erasmus this apparent contradiction between two very different views of language becomes a source of strength and even, perhaps, a kind of axiomatic expression of a philosophical view. Indeed, at the very moment that the Enchiridion was being printed, with its imprecation "not to linger over the sterile literal sense, but to hasten on to more profound mysteries, assisting the inadequate efforts of human industry with frequent prayer" (36), Erasmus was working to perfect his Greek, a scholarly attainment that later enabled him to embark on the most ambitious philological, hermeneutic, and literary project in the Jeromian tradition since Jerome himself: the textual restoration, retranslation, and annotation of the New Testament.(7) Erasmus may have deprecated the results of human industry, but no scholar was ever more magnificently or ambitiously industrious. He repeatedly argues that reading and understanding scripture requires hard-won linguistic skills such as a familiarity with the three biblical languages; a knowledge of figures, tropes, and idioms; and even a broad factual knowledge of the natural world.(8) Somehow the abundance and exuberance of Erasmus's own linguistic practices and the energy with which he describes the appropriate methodologies for acquiring linguistic facility always seem to belie any formal or theoretical reservations that he expresses about the limits of those practices.

Over the last twenty years, guided by a growing general interest in the philosophy of language, a number of scholars have looked anew at the way that Erasmus brought the linguistic techniques and sensibilities of humanism to bear on the objectives of Christian piety. Margaret Boyle (1977) has attacked the problem by analyzing the linguistic and theological issues involved in Erasmus's daring decision to translate the Vulgate's verbum as sermo. Jacques Chomarat's Grammaire et rhetorique chez Erasme provides an exhaustive study of the tradition and sources of Erasmus's linguistic practices as well as his general philosophy of language. Others, such as J.B. Payne and T.F. Torrance, described Erasmus's hermeneutics in a systematic way. Payne, who sees Erasmus as trying to achieve a balance between the principles of the letter and the spirit, between the historical and philological exegesis of Jerome and the more "spiritualistic, platonizing" tradition of Origen, describes Erasmus's hermeneutics as a kind of negotiation between these two very different traditions. Manfred Hoffmann's recent book Rhetoric and Theology: The Hermeneutic of Erasmus takes an integrated approach by looking at how Erasmus frames his theological hermeneutics and homiletics in notions about language derived from the rhetorical tradition.(9) Something about the way that Erasmus urges his readers toward eloquence, while persistently reminding them of its limitations, seems to require explanation.

Yet another approach to some of the apparent contradictions produced by Erasmus's attitudes toward language might be to put aside temporarily any effort to erase these inconsistencies and instead focus on the practices that seem to generate them and on the hermeneutic consequences that they entail. For what language actually does -- the kind of effects that it produces -- is what interests Erasmus most in his pedagogy and theology. De copia, for instance, is fundamentally a study of the effects produced by different word choices and the various techniques of linguistic elaboration. And both the Ratio verae theologiae (1518) and Ecclesiastes sive de ratione concionandi (1535) discuss preaching largely in terms of style and rhetorical technique, clearly aiming more at spiritual and ethical results than at theological meaning for its own sake. The goal of theologians, instructs Erasmus, should be to set forth the Scripture, give an account of the faith, and discuss spirituality, but always with the objective of eliciting tears of contrition and inflaming the hearts of their listeners.(10)

In short, Erasmus's linguistic practices add up to a functional treatment of language as praxis. A common denominator of humanism, Hanna Gray has argued, was a concern for the practical application of eloquence, especially for its utility in being able in principle to move people to act according to their understanding of the moral precepts communicated to them. In Erasmus's work, the concern for practical application becomes paramount. Linguistic praxis Erasmian-style marries form to content and meaning to utility. Language as praxis is language in action, language always in the process of exchange and negotiation, language that aims primarily at performance, not representation.(11) Its inherently incomplete and processive nature tends to produce a double motion of simultaneously doing and undoing that proves itself profoundly instrumental even while it presents new challenges. For instance, language as praxis may seem to give up on properly delineating the whole of the real -- there can be no true congruence between res and verbum -- but it can do real work in the world. Erasmus also discovers along the way, I believe, that linguistic praxis cannot guarantee hermeneutic closure. On the contrary, it negates any clear distinction between representation and effect or speech and its speaker. What it offers instead is the limited human authenticity of a system grounded in the actual usage of its speakers and the very real effects that those speakers have on one another through language.

Making Language Work through Rhetorical Variation and Allegory

Richard Waswo has argued that the dichotomy between words and ideas presumed in Erasmus's De copia and De ratione studii represents what he calls the "dilemma of the age . . . the desire for the transparent cognition of things that is frustrated by the awareness of the opaque fluidity of words."(12) But the kind of frustration that Waswo describes is seldom apparent in Erasmus. Rather, Erasmus treats the semiotic vagueness of a discourse caught up in history and contingency as a kind of linguistic felix culpa at the generative heart of communication -- the place where the same is turned away and version meets version, making more versions (and more mediations) possible. The multiplicity (and copiousness) that marks where one meaning intersects another makes this place recognizable. And there, where meanings mix and mingle, we find the possibility for the spiritual accommodations and moral transformations that ethically legitimize Erasmian linguistic praxis.(13)

In the Adagia, for instance, a plenitude of voices gradually unfolds the wisdom encapsulated in the verbal hieroglyphs that seem to offer themselves (at least in Erasmus's hands) as topics for ongoing conversation and linguistic excavation.(14) Erasmus maintains that it is a combination of the antiquity of the adage (often an authority spoken through a sequence of different voices) and its metaphorical resistance to any simple and singular interpretation that gives it a special status -- multivalence enhances the practical utility of an authority supposedly grounded in another kind of priority.(15) An analogous situation exists in the realm of the sacred, where multiple and even disparate expressions of faith can be considered productive: "variety does not truly disturb the harmony of Christ; on the contrary, just as a delightful medley is rendered by a diversity of voices, so the variety of Christ makes the harmony more complete."(16) In other words, the abundant nature of Christ as the Word makes our many different approaches to divine meaning through human language an appropriate and positive response. Amplification, points out Hoffmann, is for Erasmus a kind of "theological paradigm for spiritual freedom."(17)

Erasmus's praise for variety -- and differences -- in secular rhetoric is not unique, but its passion and degree may be:

Variety is so powerful in every sphere that there is absolutely nothing, however brilliant, which is not dimmed if not commended by variety . . . Just as the eyes fasten themselves on some new spectacle, so the mind is always looking round for some fresh object of interest. If it is offered a monotonous succession of similarities, it very soon wearies and turns its attention elsewhere, and so everything gained by speech is lost all at once. This disaster can easily be avoided by someone who has it at his fingertips to turn one idea into more shapes than Proteus himself is supposed to have turned into. (De copia, 302) Erasmus argues similarly that amplificatio has a profoundly persuasive effect in preaching.(18) But the rhetorical force that Erasmus aims at here is not the same as representation understood as a straightforward rendering of res through verbum. Linguistic variety in both secular and sacred discourse may have the inevitable consequence of generating additional meaning, but it seems plain that its primary role is to effect change, and that precise signification remains subordinate to that function.

Signification also takes a back seat to rhetorical vigor in regard to metaphor, the figure that Erasmus considers the most compelling of all figures, whether one is "looking for abundance, distinctness of language, credibility, vehemence, brilliance and dignity, or pleasantness and grace."(19) Described in the Ecclesiastes as contributing greatly to amplificatio through the figure of emphasis, metaphor has the same kind of powerful effect on the minds of its audience as other species of amplificatio -- it stimulates our interest. And given the exemplary efficacy of metaphor, it should not be surprising that allegoria, classically defined as a kind of extended metaphor, is also the figure most characteristic of scriptural rhetoric. Moreover, the willingness of much of Scripture to submit to and even to demand an allegorical reading demonstrates for Erasmus that allegory is a primary means through which God accommodates the Word to our limited understanding, just as his Son (also his Word) accommodated himself to our need for salvation. Erasmus explains in the adage "Silent Alcibiadis," that Scripture, like one of those small statues of the Greek satyr Silenus, appears amusing, ridiculous, even contemptible on the outside, but opens up to reveal the image of the divine. "Pause at the surface, and what you see is sometimes ridiculous; were you to pierce to the heart of the allegory, you would venerate the divine wisdom."(20) This penetration of the allegorical nature of the text, argues Hoffmann, is the process through which the breach between verbum and res begins to be ameliorated by the "parabolic" connection of metaphorical similitude.(21)

But as Hoffmann notes, not only does allegory bring about an accommodation of the divine to the less than divine; it simultaneously moves the reader to try to accommodate himself or herself to the Word.(22) "The divine Spirit has his own peculiar language and modes of speech, which you must learn through careful observation," insists Erasmus. "Divine wisdom speaks to us in baby-talk and like a loving mother accommodates its words to our state of infancy . . . But you must hasten to grow so that you may receive solid food. It lowers itself to your lowliness, but you on your part must rise to its sublimity" (Enchiridion, 35). Most notable in Erasmus's instructions in the Enchiridion, however, is the persistent emphasis Erasmus puts on the need for the reader to act: to "pierce" through to a more spiritual meaning, to "break through the husk and extract the kernel," to "search out the spiritual meaning" -- then to grow and to rise (ibid., 35, 32). Accommodation (and allegory as its most effective instrumentality) is the process through which the Word becomes most successfully performative, not by making divine meaning wholly present to the reader's understanding, but by drawing the reader to itself through a kind of allegorical provocation.(23)

Although scriptural allegory in the text of Scripture may move the reader into a greater proximity to meaning, Erasmus always seems more interested in the multiple rhetorical functions that allegory performs. And even these functions are not consistent with one another; only the effectiveness of allegory remains constant. Christ, the perfect exemplar of the perfectly efficacious Word, is a kind of Proteus, adapting to all those persons whom he wishes to attract to himself, at once diverse and yet "never unlike himself."(24) On one hand, explains Erasmus in the Ratio, Christ employed allegories, similes, and parables to stir up our spiritual sluggishness.(25) Christ also spoke allegorically because allegory strongly affects the minds of readers, firmly implanting the text in their memories and leading them step by step to a more perfect knowledge.(26) On the other hand, allegory can just as easily work to hide the true meaning of Christ's actions from the impious.(27) Functioning very differently according to different applications, allegorical figures are the "uneven places in the road" that the divine wisdom puts in our way to press us in the right direction along an always variable spiritual via.(28) They sometimes lead the reader toward the divine; at other times they seem to block the way.

Hence malleable inconsistency is that which is most profoundly instrumental in linguistic praxis. Even the New Testament, in which the presence of Christ theoretically renders the text capable of perfect representation, seems to privilege the active accommodation of praxis over the precision of pure signification. Indeed, the very need for accommodation assumes that the presence of an interpreter of some kind is an unavoidable part of the relation between res and verbum. Allegory, copiousness, and accommodation belong not to a Saussurean semiotic defined by the arbitrary relation of signifier and signified within the sign, but rather to a Peircean model in which signs always signify "to somebody for something in some respect or capacity."(29) Allegory (and the hermeneutic difference epitomized by allegory) is the protean, limitless, functional aspect of language that both ameliorates the split between verbum and res and affirms that there can be no unmediated relation between the two.

The Effect of Praxis and Performance on the Relation between Word and Meaning

But what kind of access to meaning does this notion of language allow us? More specifically, what kind of relation actually exists between word and meaning in Erasmus's hermeneutic practice?(30) Critics arguing that Erasmus had a conventional perspective on this issue often cite a passage from his pedagogical work De copia that evokes the classical topos of language as clothing the body of thought: "But to return to the main point, style is to thought as clothes are to the body. Just as dress and outward appearance can enhance or disfigure the beauty and dignity of the body, so words can enhance or disfigure thought" (306).(31) Words only marginally affect content in this largely conventional formula. But a closer look discloses that Erasmus sometimes elaborates this topos in unexpected ways; that, indeed, he uses it in a way that effectively blurs the traditional relation between meaning and the word that discloses it: "The practice of giving variety to expression is exactly like changing clothes. Our first concern should be to see that the garment is clean, that it fits, and that it is not wrongly made up. It would be ludicrous to have a man go out in public dressed like a woman, and objectionable to see a person wearing his clothes back to front or upside down" (ibid., 306). The cleanliness of the garment, Erasmus tells us, alludes to a refined Latinity, a grammatical pumas. But what precisely is the body that this garment must fit? That of the subject or that of the speaker? Supposedly, Erasmus continues to refer to thought and the expression that clothes it. But the way he develops the simile pushes the whole matter so firmly into the context of the socially appropriate or inappropriate that the distinction between thought and speaker becomes increasingly indistinct. Thought, expression and social norms are stirred into a single mix: the subject of speech and the context of its speaking seem inseparable from the speaker and his or her position among these circumstances.

Once we puzzle out Erasmus's description, however, the familiar argument and its particulars seem to make pretty good sense. For a man to go out dressed as a woman would indeed violate the norms of social behavior. For him to wear his clothing upside down would be ridiculous in the context of social norms and absurd in terms of the functional relationship between body and clothing. A certain fit between meaning and the form of expression seems a reasonable criterion for speech. Nevertheless, the identity that may or may not exist between meaning and words, or even how to arrive at the particular words that will most closely stand in for a meaning, is not really the question. Images of social propriety are used for the good reason that Erasmus is committed to looking outward toward the social context in which words are spoken; he is interested in effect, usually, moral and spiritual effect.

Later, in Erasmus's instructions for variation in theological speech, it becomes clear that purity of heart and grammatical purity are both prerequisites for comprehending the pure meaning of Scripture. It is the puritas on which Erasmus insists, ideally shared by all elements of the equation, that lends the speech act its greatest efficacy. Hoffmann argues that this emphasis on puritas is founded on the Aristotelian principle of similia similibus -- like attracting like.(32) But it is also true that an ethical subtext inherently distracts from any abstract hermeneutic relation between thought and word by introducing a notion of similarity that is not necessarily one of identity. In the passage cited above, expression works with content, but also against it simply by showing that there is no place where words and ideas find themselves in an isolated relation to one another. Meaning may be theoretically accessible (and practically so in a equivocal fashion), but words in the world are so deeply contextualized that any exact communication of that meaning seems impossible.

Given the overriding concern with performance that characterizes linguistic praxis, the shift in focus from meaning in the abstract to what goes on in the transaction that language at once brings about and mediates seems a natural one. But it leads to another essential feature of what I am calling linguistic praxis -- the dialogic aspect of discourse as it has been defined for us in the work of M.M. Bakhtin.(33) To insist on the dialogic property of language means to privilege multiple readers and contingent readings over the idea of an objective (unvoiced) discourse in some ahistorical relation to an ideal truth. The dialogic word is a word coming from somewhere and someone, already existing in a fabric of other, alien words; addressed to somewhere and someone; and intercepted by another word similarly enmeshed in yet more words. This attention to a multitude of reading and speaking subjects interacting with one another over time characterizes the copiousness and flexibility that Erasmus values in linguistic training, theological philology, and hermeneutics. Yet Erasmus's methodology does not ignore meaning in favor of social contingencies. Rather, it is designed to put more meaning into the world and to make that meaning more active by showing readers and writers how to take account of those contingencies and use them to their advantage. Nonetheless, the emphasis on dialogue and multivalence does give his work an open-endedness that can be mistaken for imprecision.

The fit between words and matter is already a little hazy in De copia -- and thus the precision of meaning compromised -- where Erasmus describes linguistic copiousness and compression as complementary rather than opposing qualities.(34) First, he cites a "number of famous sophists" who in showing how to abridge a matter simultaneously demonstrate how to expand it (ibid., 297). A little further on he declares that "the purpose of these instructions is to enable you to so include the essential in the fewest possible words that nothing is lacking, or so to enlarge and enrich your expression of it that even so nothing is redundant; and to give you the choice, once you understand the principles" (ibid., 301). Finally, he demonstrates the potential of copia by turning a given sentence into over 200 different "Protean variety of shapes," from "Always as long as I live, I shall remember you" to "You are too deeply hidden in this breast of mine to be driven out by any means, as long as the fates shall not grudge me life" (ibid., 354-64).(35) But does this astonishing performance merely create new versions of a pre-existing meaning or does it generate new meaning? Either way, what meaning is that and how do we name it? Nothing gives the sentence with which Erasmus began any real epistemological priority or necessarily makes it anything more than another version.

Such a narration may not directly challenge the distinction between expression and content, but it does preclude the possibility of a static or transparent relation between them. It implicitly substitutes for any distinct epistemological object what I want to call a fuzzy set, a delineation of the signified that dissipates beyond a certain extension of the set and likewise fades away within a certain compression of the set's parameters.(36) Erasmus is not questioning the existence of a meaning troth is distinct, diverse and bounded. But his practice suggests that it may be difficult to get at this kind of meaning, since merely its entry into language seems to render it less precise and certainly unsusceptible to any independent verification.

The parameters of meaning also seem somewhat fluid in Erasmus's Paraphrases of the New Testament, where he takes on the job "of bridging gaps, smoothing rough passages, bringing order out of confusion and simplicity out of complication, untying knots, throwing light on dark places, and giving Hebrew turns of speech a Roman dress."(37) All this in order to help the New Testament text act more effectively on its readers. The challenge, Erasmus maintains, is to "say things differently without saying different things, especially on a subject which is not only difficult in many ways, but sacred, and very near the majesty of the Gospel" (ibid., 3). Although his procedures for accomplishing this clarification are various, more often than not he ends up supplementing the text in some way, adding clauses, sentences, and whole paragraphs that in turn provide historical detail, explain motivation, elaborate on theologically key words, or place the immediate narrative m the larger context of salvation history. In short, a process of amplification, very similar to that described so enthusiastically in his secular pedagogy, functions as the underlying principle of clarification in Erasmus's scriptural paraphrases.(38) If nothing were added to Scripture, Erasmus maintains, his paraphrases would not be paraphrases at all.(39)

Aptly calling on the etymology of the word explanatius in the Erasmian phrase "immo omnibus crassius et explanatius loquentem" ("in terms quite plain and clear"), Robert Sider argues that Erasmus's intention in the paraphrases is to "unwrap layers of meaning hidden in the language of the Holy Writ, to unfold before our eyes and extend, as it were upon a plane surface, the abundant truths of scripture."(40) If so, a familiar question arises. How can we decide whether this meaning is an unfolding of meaning already present or the production of new meaning appended to the old? What means do we have for judging what is said -- and judging whether the same things are said differently but remain the same -- except by what is said? In his paraphrase of John 2, for instance, Erasmus sets the scene by explaining what brought Jesus, or more specifically, his mother Mary, to the marriage at Canal He spends considerable effort explaining Jesus's rebuke of his mother ("not that he, who loved the whole human race, did not have a singular affection for her as his mother" in terms of the need to separate the authority of miracles from human desire.(41) However, none of this material is present in Scripture itself. It is commentary that has been made a part of the biblical narrative in order to clarify the text and to make it work better.

Erasmus himself admits that the process of paraphrasing creates certain troubling and unavoidable distortions in the literal sense of the text. He found, for instance, that elucidating the figurative language of Jesus in order that the ordinary reader might better understand it had the effect of making it seem odd that persons in the scriptural story sometimes responded as if they could not understand what Jesus meant. Moreover, as Erasmus complains to the young Archduke Ferdinand, since "it is the paraphrast's business to set forth at greater length what has been expressed concisely, it was equally impossible for me to observe the limits of time" in the composition of the Paraphrase on John.(42) Laying out the abundance of meaning and the richness of Christ's discourse on that last night seemed inevitably to misrepresent the time frame of the narrative. When Erasmus contends that paraphrase differs from translation in that paraphrase rephrases what the author says in the words of the paraphraser, whereas a translation attempts to reproduce a version of the words themselves, he seems to be claiming that his paraphrases successfully render the res if not the verbum of the scriptural text.(43) But in practice, attempting to produce a more compelling and comprehensible expression of content, an expression thereby truer to the true res, also has the unanticipated effect of introducing new indeterminacies into the relation between words and matter.

Bringing Pure Meaning into the World of Speakers

In Erasmus's De recta latini graecique sermonis pronuntiatione (1528), a treatise written as a dialogue between a paternal Lion and a garrulous Bear, the Bear seems to suggest that words share the nature of the things to which they refer. Here would seem to be a strategy that will stabilize the relation between words and meaning by making them kissing cousins, each tied independently to a common point of material reference. A syllogistic relation is set up with the "thing" as the middle term:

If meaning is a discrete idea of a particular thing,

And that thing in turn influences or even determines the word which

properly refers to it,

Then meaning must be concretely related to the words which express it.

The problem this time turns out to be the indefinite nature of meaning itself. The Bear, contending that the letter "r" represents a sound universally associated with "a rattle or a whirr like a stone flung with great impetus from a sling or a javelin leaving a ballista, or a cane or rod swished through the air," argues that it therefore naturally forms parts of words having to do with speed. But when the Lion raises the very common sense objection that the logical extension of such a theory would be that "every object ought to have the same name in every language," we see that the Bear is really aiming at something a little different. To the contrary, he replies, a name represents only a particular aspect of the "whole meaning" and consequently "the same totality can be represented in many different ways."(44) Meanings are not congruent with things at all. Although the Bear argues for a kind of natural connection between words and things, the meanings that he evokes are much more elusive, lack distinct boundaries, and are never fully present through the word. A particular linguistic performance makes present one aspect of meaning, not its totality. Another performance (and another) may summon another aspect of that meaning.

Erasmus seldom interests himself in meaning independent of speakers and their performances. For him the speaker speaking seems to be an essential factor in the efficacy of a discourse generally, and as we have seen, particularly important in relation to the degree to which discourse can successfully perform an ethical function. This precept is not unfamiliar to classical rhetoric. In De oratore, for instance, Cicero's Crassus argues for the importance of natural genius and character in the successful orator.(45) Quintilian, a critical influence on Erasmus's pedagogical works, underscores this implicitly ethical concern by insisting that not only should the orator be a good man, but "no one will be an orator unless he is a good man."(46) Here the puritas that Erasmus demands of the theological orator is shown to have a very practical foundation. One should, he maintains in the Ratio, come to theology "purified" from transgressions, not simply because of the sanctified nature of the subject, but also because in that way one can best act as a "placid river or a smooth and unblemished mirror in which the image of eternal truth can shine" and be communicated to one's listeners.(47) Hence, when Erasmus declares "let us hear the man himself, speaking like a Roman to Romans, no, rather to all, in terms quite blunt and clear," he means to present Paul's epistle in the most effective possible way.(48)

Arguing against the perception that Erasmus has "no clear awareness of the differences between Paul and himself as speakers in the paraphrase," John Bateman points out that Erasmus had pressing rhetorical reasons for maintaining Paul's voice in his Paraphrase on Romans. Allowing Paul to speak directly enhances the ethical force of the text, he argues. And removing any mediating voices gives the text a greater immediacy and authority.(49) Both characteristics can be explained as part of Erasmus's concern for accommodatio. Nonetheless, Erasmus's claim to render the voice as a man speaking "to all" is not quite accurate. Erasmus's linguistic task is much more specifically defined by the nature of his sixteenth-century Latin readers, and in trying to evoke a Paul that speaks with rhetorical decorum to these readers, Erasmus inevitably makes Paul speak differently. In the very process of accommodating the text to the understanding of its contemporary readers, the humanist goal of ad fontes is necessarily compromised by a movement in the other direction, toward an always new historical moment.

Erasmus's declared goal of trying to try to understand the apostle's intent in the same way that the ancients had done is also compromised by his attention to rhetorical effectiveness, since part of his strategy consists in enclosing historical and interpretive differences among the Fathers within the constructed integrity of a single voice speaking continuously.(50) Such a technique allows no opportunity for more than a single interpretation of the text and no room for indecision, notes Chomarat.(51) The Erasmian Paul speaks clearly and relatively unambiguously to his readers. However, this straightforward lucidity of this voice is achieved only at the price of fictionally suppressing the dialogism which frames it, the underlying concert of voices with whom Erasmus actually speaks. The paraphrase of the apostle's epistles actually consist of a translation and a reading of that translation, a reading whose status as a reading is disguised by a continuous narrative and lack of annotation. Although this dialogue lends authority to Erasmus's interpretation, the dialogic element of that authority remains enclosed within a monologic narrative.(52)

Erasmus's claim is that the multiple voices subsumed within the single voice that speaks Paul clarify the difficulties of Paul's thought and language. But how can they do so without in some way speaking for him or through him and thus appropriating his speaking and his voice? Here again we are concerned, perhaps, with the way expression shapes matter, with the always mediated relation between verbum and res. Erasmus's paraphrases are offered to us as a via by means of which we can penetrate the sometimes hard surface of the scriptural text and draw closer to its true spiritual meaning. However, in order to produce this simple and accessible version of the Word, it has already been necessary to subject it to the contingencies of usage and exchange. Paul's speaking in the Paraphrase on Romans is already a negotiation of diverse voices and diverse moments, a discourse embedded in praxis.

A theoretical model of signification can offer hermeneutic closure and the possibility of a complete meaning only when the words are considered independently of the speaker and other speakings. Indeed, the most fundamental premises of theory inherently involve a separation between subject and object -- or the reader and the text -- an Aristotelian distance that the syllogistic logic of scholasticism took for an essential attribute of truth. This distance is palpable in the Greek theoria, a word meaning contemplation, speculation, or sight, and even more so in the Greek theoro's, meaning spectator.(53) A theoretical grammar and logic may offer a kind of practice, but it is a practice that seeks to refine itself out of existence because logical operations intentionally aim at a formal stasis based on an elimination of the contingencies of practice.(54) In contrast, Erasmian praxis turns away from the distances of theory and toward the endless mediations and performances of usage. Just because an Erasmian text always lives on what Bakhtin calls "the boundary between two consciousnesses, two subjects"(55) does not mean that it fails to mean, only that the kind of meaning adduced from praxis cannot (until humanity is again brought face to face with the Logos) be authorized by any verifiable appeal to something outside of praxis.

Looking for Origin: The toss of the Transparent Text

In 1516 Erasmus published Novum instrumentum, the first of five editions of his New Testament. Each subsequent edition contained significant changes: the second edition of 1519 made the controversial substitution of sermo for verbum at John 1:1; the 1522 edition included a more completely revised Latin translation. Along the way, Erasmus developed a significant group of philological techniques, some of which are still admired by contemporary practitioners of textual criticism. Erika Rummel's volume Erasmus' "Annotations" on the New Testament (1986) provides an excellent study of these techniques; and Rummel, Bentley, Rabil, and others have noted the examples I cite below for a variety of reasons. Rummel is particularly interested in the theological implications of Erasmus's choices; Rabil emphasizes the effects of humanism on Erasmus's techniques and exegesis; and Bentley looks closely at the details of Erasmus's methodology, especially the principle of the "harder reading," in order to analyze how this methodology actually works.(56) But it is also interesting to observe the way in which the hermeneutic goal of Erasmus's project is at once aided and made to seem doubtful through the very same methodology.

Erasmus undeniably saw the task of purifying the material text of Scripture from the contaminations of history as an act of piety, a step toward restoring the full capacity of Scripture to act on its readers.(57) He seems to look, sometimes quite literally, for a purer text that dwells beneath the corruptions and ill-conceived corrections visited on the palimpsests with which he worked.(58) At 1 Corinthians 8 note 8, for instance, he supports his own translation by referring to a Pauline codex lent to him by Colet where he could actually trace an older reading through a correction.(59) Similarly, at Matthew 27 note 42, he comments that "in the Constance codex an erasure witnesses to a distortion."(60) Although Erasmus may not claim to be able to restore the New Testament to a condition of absolute referentiality, the implication remains that this is at least the ideal goal of his project. But this kind of textual archaeology, carefully expounded in a growing body of annotations, seems to have the practical effect of supplementing the text without ever providing final closure. Unexpectedly, it leads Erasmus back to the linguistic uncertainties and Pauline theology of the Enchiridion.

This is not to say that Erasmus's scriptural renovation fails to remove from the Vulgate many of its spurious accretions, nor that the art of discourse is fruitless for Erasmus, simply that the notion of an original text promised by a philological renovation according to the humanist goal of adfontes remains elusive, just as the plenitudinous meaning of the Gospel remains inexpressible through the application of a strictly literal hermeneutics. Privileging the historical and the literal senses, and boldly exploiting the discrepancies and disagreements among various historical manuscripts, these sophisticated techniques constitute a thoroughgoing praxis and a brilliant strategy, but one that hopelessly entangles the text in history. Precisely the attention paid to the renovation of the material word points to the inability of language caught up in other language ever to yield a text of total authenticity. In pursuit of an ideally transparent text -- or a text that approximates that text -- Erasmus unintentionally ends up staging a loss of textual origin in which technique and practice point back to themselves (back to praxis) and away from the end that they are meant to achieve. Erasmus, with a poet's love for words and a deeply held conviction that they continue to be our only possible approach to meaning, remains committed to the possibilities for effective discourse.(61) Nevertheless, his practice suggests that language itself, and the texts in which it is embodied, is the glass through which we see darkly.

Noting that the errors are not consistent from one manuscript to another, Erasmus's overall strategy is to reconstruct a more perfect reading by collating a variety of imperfect and corrupted manuscripts. He especially exploits the semantic resistance between manuscripts, the places where textual discrepancies make a true congruence of meaning between them impossible. For instance, whereas conservative theologians were deeply suspicious of the alternate readings found in the early Greek manuscripts, Erasmus was only critical of those Greek manuscripts that had come into contact with early Latin manuscripts. These later Greek manuscripts were obviously more consistent with the Latin, but their antiquity and independence (their differences) had been compromised. "Texts of this kind are like a white line on a white stone" he protests in his second note on 2 Corinthians 2.(62) In other words, the semantic resistance and tension necessary for a useful negotiation of meaning had been eliminated. Erasmus understood that textual differences may beget ambiguities, but he also understood that technically it is at the confluence (the dialogic crossroads) of such differences that ambiguity can be best negotiated and a new reading obtained.

Among the more interesting philological techniques that Erasmus employs is that of the "harder reading," a method in which this resistance becomes the sole principle used to determine the earlier, and thus more authentic, reading.(63) "And whenever the ancients note a variant reading," argues Erasmus, "the reading that appears absurd at first glance always tends to be the more suspect one [meaning the more likely one], in my opinion; for it stands to reason that a reader who lacked either education or concentration was offended by the absurdity of the expression and changed what was written here."(64) Jerry Bentley offers several such examples from Erasmus's annotations, describing how Erasmus analyzed the different motives that might have inspired particular distortions of the Scripture. At John 7:1, 1 John 5:7-8, and the well-known Trinitarian verse at 1 John 5:78, for example, Erasmus argues that fear of the Arian heresy prompted scribes to alter the text. Bentley also shows how elsewhere Erasmus found passages where scribes had made changes simply in the name of rendering the text less rigorous or more consistently readable. Matthew 5:22, where the passage "whoever becomes angry at his brother will be held accountable in the judgment," had been altered to read "whoever becomes angry at his brother without cause will be held accountable in the judgment" seemed to be an example of just this kind of modification.(65)

But whether the case is plainly ideological or more one of literary judgment, "difficilior lectio potior" only works to free the text from an erroneous reading by recognizing the resistance of an earlier reading that is itself a function of meaning already caught up in the dialogism of words. Erasmus's explanations attempt to free the text from the vagaries of transmission in order to release it into some purer philological realm that allows for an unfettered rational analysis of textual problems. But philology can never wholly throw off hermeneutic presuppositions. The earlier reading appears as the "harder reading" because it belonged to someone writing in some particular time and place and in response to other meaning and other challenges. Hermeneutic grounds of some kind are essential if only to be able to determine what constitutes a "harder" reading. A lectio impossibilior, points out Bentley, would be an abuse of this principle.(66) But without presuming a hermeneutic standard, even the difference between a lectio impossibilior and a difficilior lectio is impossible to measure; the grounding in a social and historical (and hermeneutic) particularity is what makes a reading resist the appropriation -- or misappropriation -- of a later reader in the first place. In the end, a historical priority, not an ontological one, is all that can be claimed for the choice; the "harder reading" remains a reading among other readings and the text, a collage of different readings elbowing one another for historical advantage.

Elsewhere, Erasmus turns to a consensus of knowledgeable readers for authorization. To support using the controversial sermo rather than the Vulgate's verbum in the Latin translation of the 1519 New Testament, Erasmus cites a variety of patristic sources and scholastic authors, as well as the Glossa ordinaria, the daily choir, the daily office, and the traditional usage of the schools.(67) For Erasmus, consensus ideally seems to be a rhetorical principle possessing a hermeneutic authority based on the essential harmony of a truth variously expressed but unified in the one Christ.(68) Nonetheless, consensus patrum is never an absolute principle, if only because Erasmus contends that one must sometimes disagree with the Fathers. "And if there are anywhere palpable lapses, instead of concealing them, let us dissent from them -- not furiously railing at human error -- but diminishing and cleansing as far as is possible."(69) Moreover, disagreement obviously exists among the Fathers themselves. "I grant that the Church has the authority to interpret Scripture, but the teachers of the Church, no matter how famous they may be, hesitate over many passages of Scripture, cannot agree on many of them and actually interpret some of them falsely."(70)

At first, the appeal to consensus seems to run counter to that of the philological techniques that rely on the resistances between readings. The former ostensibly builds on agreement; the latter take advantage of disagreement. But in fact, both procedures are premised on differences and the adjudication of these differences -- the common denominator is praxis. Both categories of appeal acknowledge a universe of discourse defined by language in action, language whose reference is always shared by (and distributed among) a world of various speakers and various speakings. Moreover, any methodology based on praxis remains vulnerable to further praxis. Erasmus might have aimed for a consensus that would render any further interpretive action unnecessary, but the ongoing controversies surrounding his own work demonstrate how difficult a definitive consensus may be to achieve. Antagonist Jacobus Stunica's (Zuniga) bitter Annotationes in defensionem translationis Novi Testamenti (1520) attacked many of Erasmus's conclusions. Englishman Edward Lee even wrote a resentful volume of annotations on the annotations of Erasmus.(71) In the end, one could argue that instead of securing the boundaries of meaning in the New Testament, Erasmus's Annotations inadvertently made those boundaries the focus of further contention by generating a new flood of linguistic response and interpretation.

An appeal to consensus is inherently very different from an appeal to the precise authority of a single, monologic Word. No voice caught in history is ever wholly identical to another; no contextualized speaking ever precisely equivalent to a previous speaking. Consensus requires that all participating voices be temporarily decontextualized, that differences be suppressed within the magic circle of agreement. Consensus is at best an approximate authority arising out of a negotiation that aligns voices within this circle. It can be challenged at any time by invoking other voices, by drawing on different testimony from the same witnesses, or by re-asking the question in such a way as to excavate some differences and suppress others. Evolving out of negotiation, consensus, unlike syllogistic reasoning, is vulnerable to further negotiation.

Erasmus, however, does not always seem troubled with this potential for further readings or additional discussion. In his long annotation on Romans 1:4, for instance, he comments that "this passage can be read in a variety of ways. I shall set forth all the ways simply, as is my purpose in this work; the right to judge and the power to choose will belong to the reader."(72) Then beginning with the "ancient commentators," he discusses at length three possible readings of what the Vulgate renders as "ex resurrectione mortuorum Iesu Christi" ("from the resurrection of the dead of Jesus Christ"). Erasmus is clearly impressed with Valla's rendering "of Jesus Christ" in the ablative case instead of the genitive and with the "cure" that that choice seems to effect for a potential confusion in Paul's text. In fact, he has followed Valla in the Latin that prefaces his annotation, translating the passage "ex eo quod resurrexit e mortuis Iesus Christus" ("in that he rose from the dead, Jesus Christ"). But in the text of his New Testament translation, Erasmus shows his usual flexibility translating the phrase in the nominative case. He makes this case in the annotation as well, albeit quite briefly.

Thus Erasmus's annotation is really a lengthy dialogue on these matters -- an effect quite opposite to that created in the paraphrases. Erasmus presents the arguments for and against different readings and explains how each reading has a particular impact on the theological cruxes of the verse, specifically on whether the original Greek refers to the resurrection of the saints as well as that of Christ, the "total resurrection," or simply the resurrection of Christ himself.(73) He does find some arguments to be stronger than others, yet he remains willing to lay out each argument carefully and fairly. Seeking to make his own case, he is content at the same time to leave the issues open to further discussion, acting as "an informer, leaving judgment to the reader."(74)

But surely Erasmus has done far more than just produce a cleaner and prospectively more readable text. In his Annotations Erasmus not only corrects errors which have accreted to Scripture; he also makes an effort to account for those errors, analyzing their possible origins and perpetuation. Rummel has collated a list of Erasmian explanations. She describes how words have been transferred from the margin into the text; how one word has been substituted for another; how mistakes in pronunciation result in a change of tense; visual mistakes lead to one letter being misread for another; and errors in memory adversely affect biblical lists.(75) What the Annotations add up to, I propose, is really a history of the hands that have copied the text physically from one manuscript into another manuscript into another, in other words, a history of the dialogic process of transmission. In philology and hermeneutics, both text and reader are irremediably caught in history. And although history and historical distance may seem impediments to any interpretive understanding of a text or event, they finally prove themselves, as Gadamer explains so well, to be essential prerequisites to that understanding.(76)

A truly originary and inspired text would seem to have its source in that which is first written, or even in that which exists before anything is written down at all. This kind of text must be at least prior to the historical contingencies of independent human authorship. Indeed, although Erasmus is willing to admit the presence of small errors in Scripture -- hence the need for philology and hermeneutics -- fundamental error is simply unintelligible to him.(77) The text of the Scripture may be caught up in the contingencies and indeterminacies of praxis and dialogic exchange, but its metaphysical authority remains uncompromised if fundamentally unreachable.(78) The auctoritas of Scripture resides in the perfect authority of its divine authorship that a philological project of restoration ideally enhances.(79) However, instead of effacing human authorship, Erasmus's methodology emphasizes it and dwells on the materiality of the text. The idea of an original text signifying an indisputable Word seems increasingly unlikely as it becomes clear that there is always something more (another text) or someone more (another interpreter) standing between the present text and its origin. Implicitly, the origin itself recedes behind the progressive opacity of history and the imperfection of human transmission -- the very process of recovery ultimately points to the improbability of a full recovery and instead hints at a kind of infinite regress of material texts.

Writing to his friend Johannes Sixtinus in 1499, Erasmus retells a version of the story of Cain that he had recounted at an Oxford dinner attended by John Colet, Richard Charnock, and various fellows of Charnock's college of St. Mary. He had hoped by artfully playing "the part of a poet" to resolve a dispute that had become inappropriately heated. In order to certify the authority of his narrative, Erasmus assures his listeners that he will appeal to the "very oldest authorities."(80) This particular variation of the Cain story has been taken from "an extremely old book," he claims, so old that "its title and author's name had been effaced by time and eaten away by worms" (ibid., lines 45-46). Here at last is an almost-first text, a text whose authority seems to be guaranteed by an anonymity that preserves that authority from the materiality of authorship. In fact, this text is so close to being the ideal text that it almost doesn't exist at all.

But after this witty evocation of a truly authoritative text, Erasmus still makes only a half-hearted claim for its truth, asserting that it is "an account, either true, or, if not, at least a very plausible approximation to the truth" (ibid., lines 50-51). Are these values of equal weight? Is the perfect text the text that offers a truly authentic signification of divine truth, the text that has disappeared altogether? Here at least Erasmus seems relatively unperturbed by the prospect that the humanist strategy of ad fontes may never yield a first text where meaning is precisely and transparently signified.

In the Ratio verae theologiae Erasmus describes himself as one of the many-headed statues of Mercury found at Roman crossroads directing the believer along the proper way or via.(81) The via that Erasmus has in mind may be akin to the spiritual viae of the Modern Devotionalists; it may have been, as Boyle suggests, appropriated from scholastics like Peter of Spain whose famous Summulae logicales (1246) formulates the late medieval view of the dialectical method, calling it "the road to the principle of all methods."(82) Or perhaps this via should be understood as a rhetorical ordering of discourse suitable for evoking the narrative of Christ's life, the order of salvation history, or the moral progress of Christian life.(83) Whatever the case, the Erasmian via does not seem to lead to any constant or resolute hermeneutic principle or any final revelation of meaning, simply to further reading, further language study, and more linguistic action. And although the many-headed god stands at what are largely felicitous crossings, these are not places without the potential for linguistic peril: Mercury, or Hermes, is the god of interpretation -- the deity that carries messages back and forth among the other gods -- but he is also the champion of commerce and the patron of thieves. Travelers along the Erasmian way find themselves challenged by a protean word that continues to unfold as it continues to teach, delight, and move.

(1) Erasmus, 1988, 32. Further references to this translation will appear in the text.

(2) William Woodward, 241, noted that although Erasmus seemed to have a philosophy based on Plato's Cratylus, it was "but very imperfectly worked out." Much more recently, Marjorie Boyle, 51-53, has argued that there is evidence for both Erasmus, view of language as communicating "prior thought" and the "modern thesis that language is the act in which thought is thought." See also Bataillon, 12, Cave, 19 and Waswo, 217.

(3) Ecclesiastes, in Erasmus, 1969- (hereinafter ASD), 5, 5:248, 128-250, 132 (Erasmus, 1703-06 [hereinafter LB] 5, 1047A-B) "Neque enim est abbreuiata menus Domini. Nec absurdum est hoc quoque voluisse Spiritum Sanctum, vt Scriptura nonnunquam varios gignat sensus, pro cuiusque affectu. Sicuti manna cuique sapiebat quod volebat. Nec haec est Scripturarum incertitudo, sed foecunditas."

(4) Erasmus, 1933, 180 (LB 5:76E-77A): "Quod datur videre, pronus exosculare quod non datur, tamen opertum quicquid est adore simplici fide proculque venerare absit impia curiositas."

(5) See Baldwin, 1:76-103, 130-34, passim; 2:152, 176-96; Simon, 102-23, passim; McConica, 1983; Todd; and Grafton end Jardine, 122-60

(6) Erasmus, 1978, 302. Further references to this translation will appear in the text.

(7) A few alternative Latin translations and annotations of the Vulgate existed, but none, save two, of equal ambition. Erasmus's discovery of Lorenzo Valla's Admotationes in Novum Testamentum (Paris: Bade, 1505) in 1504 at the Abbey of Parc near Louvain was a compelling precedent for his own Annotations, but he claims in a letter to Dorp that his borrowings from the Valla are minor and that he differs from the Italian scholar on several theological points (letter 337 [1515], in Erasmus, 1976, lines 876-86). See Rummel, 48; and Bentley, 1983, 140-69. On the six volume Complutensian Polyglot, its scope, and date of publication, see Lyell, esp. 24-52; and Bentley, 1983, 70-111.

(8) Erasmus follows Augustine in several of these recommendations. See On Christian Doctrine, 3:11-16.

(9) See also Hoffmann, 1991; Rummel, 1986; Bentley, 1983, Rabil, Tracy- and McConica, 1969.

(10) Erasmus, 1933, 193 (LB 5:83F-84A): "At praecipuus theologorum scopus est sapienter enarrare divines litteras, de fide, non de frivolis quaestionibus rationem reddere, de pietate graviter atque efficaciter disserere, lacrimus excutere, ad caelestia inflammare animos."

(11) But this does not mean that Erasmus abandons all notions of formal linguistic representation. Waswo, 222 and 229-30, argues that Erasmus tried for an "affective semantics" that entailed both a "new way of apprehending meaning -- that is interpreting a text not by extracting from it kernels of abstraction but by using all of its concrete detail to move the will" and "a new kind of meaning apprehended: it becomes our emotional experience of the text." Although he is right on target in recognizing that Erasmus is not very interested in "kernels of abstraction," the distinction between emotional experience and intellectual apprehension rests on a post-humanist assumption that representation either wholly succeeds or entirely fails and that it can be separated from the rhetorical function of language. This distinction simply repeats the dichotomy between rhetoric and meaning, displacing it inwardly as the difference between feeling and knowing.

(12) Waswo, 218-19.

(13) Hoffmann makes the notions of accommodation and mediation -- between the spirit and the letter, God and mankind, res and verbum -- the keynotes of both his essay (1991) and his recent book on Erasmus's rhetoric and theology (1994). He also argues that the theological function of accommodation is transformation. I agree with both propositions, but would add that if a hermeneutic closure were actually achieved for Scripture, it would theoretically put an end to this kind of transformation. Thus that "perfect unity between history and mystery that has not yet come about" may be the end but not the means of our transformation (1991, 5).

(14) On the hieroglyphic nature of proverbs, see "Festina lente," II.i.1 in Erasmus, 1982, 171-90. LB 2:397D-407D.

(15) Erasmus, 1982,4-7, introduction.

(16) Erasmus, 1933, 211 (LB 5:92E): "Neque vero confundit hanc harmoniam Christi varietal; immo sicut e diversis vocibus apte compositis concentus suavissimus redditur, ita Christi varietal pleniorem efficit concentum."

(17) Hoffmann, 1994, 14-47. Hoffmann also makes the point that amplification must be balanced rhetorically and theologically by concision.

(18) Ecclesiastes ASD 5, 5:154, 22-155, 27 (LB 5:1008B): "Null persuadet efficacius nulla rem euidentius ponit ob oculos, nulla potentius moues affectus, nulla plus adfert dignitatis, venustatis aut incunditatis, aut etiam copiae; de qua nunc dicendum est, si prius admonuerimus exempla magnam habere vim et ad persuadendum et ad inflammandos animos aemulatione virtutis."

(19) ASD 5, 5:64, 238-52 (LB 5:975D-E): "Augmentum addit et emphasis, quod fit verbis significantibus, quae plus tradunt cogitation) quam si res simplicibus verbis exprimeretur. Qua quidem in re re gnat metaphora et quae hinc constant schemata, parabola, allegoria, abusus, imago . . . Verum de his et supra diximus nonnihil, et post incidet dicendi locus, quandoquidem haec schemata sere primas tenent in omni virtute dictionis, siue copiam spectes, siue euidentiam, siue probabilitatem, siue vehementiam, siue splendorem et amplitudinem, siue iucunditatem et gratiam."

(20) "Silenis Alcibiadis," III.iii.1 in Erasmus, 1992, 275. ASD 2,5:158-90,648; LB, 770 C-782C.

(21) Hoffmann, 1994, 76.

(22) Ibid., 109-11.

(23) Erasmus's ardent exclamation in the Paraclesis -- that scriptural writings "bring you the living image of His holy mind, and the speaking, healing, dying, rising Christ himself, and thus render Him so fully present that you would see less than if you gazed upon Him with your very eyes" (Erasmus, 1987, 108) -- seems to contradict the notion that the Word has a primarily rhetorical rather than representational efficacy. Perhaps so, though Erasmus shows more passionate optimism in the Paraclesis (often cited, for instance, to prove his otherwise dubious support for a vernacular translation of Scripture) than almost anywhere else. However, see Hoffmann's comments on the rhetorical force of fabula, an allegorical figure through which facts are rendered more intensely by being dramatically narrativized (Hoffmann, 1994, 82 and 250, n. 96).

(24) Erasmus, 1933, 211 (LB 5:92E): "Sic omnia factus est omnibus, ut nusquam tamen sui dissimilis." Erasmus then illustrates this variability through a series of New Testament examples that end by calling Christ a kind of Proteus: "Adeo cum nostro Christo nihil sit simplicius, tamen arcano quodam consilio Proteum quedam reprasentat varietate vitae atque doctinae" (Holborn ea., 214).

(25) Ibid., 259-60 (LB 5:117B): "Nam tropis et allegories ac similibus seu parabolic fere opertus est et obliquus, nonnumquam usque ad aenigmatis obscuritatem, sive id Christo visum est, quo prophetarum sermonem, cui Iudaeorum aures assueverant, referret, sive hac difficultate segnitem nostram exercere voluit, ut postea gratior esset fructus non sine negotio quaesitus, sive hoc consilio sue mysteria profanis et impiis operta celataque esse voluit, at sic, ut interim piis scrutatoribus non intercluderetur assequendi spes, sive genus hoc dictionis potissimum placuit, quod ut ad persuadendum cum primis efficax est, ita doctis pariter et indoctis expositum et familiare maximeque secundum naturum ...."

(26) Ecclesiastes, ASD 5, 5:250, 143-57 (LB 5:1047C-E): "Postremo sicut habet plus calorie solis radius speculo aut aenea pelui exceptus, ita vehementius afficiunt animos nostros quae per allegoriam traduntur quam quae simpliciter narrantur. Quod idem usu venit in picturis. Altius insidunt animis nostril, quae de Christo referuntur, quod sue morte nos liberarit a tyrannide Satanae, ab ignorantia veri, a seruitute peccati, si adhibeatur allegoria de phase. Quid quod etiam memoriam adiuuant. Nam qui profitentur artificium memoriae, per imagines quasdam infigunt animis quod nolunt per obliuionem excidere . . . Denique perspicacissimus ille Gregorius Nazianzenus libro Theologiae quinto demonstrat in coelesti philosophia docendi rationem esse commodissimam, si non statim aperiantur summa, sed per gradus quosdam auditores deducantur ad perfectam cognitionem."

(27) See above note 25.

(28) Erasmus, 1933, 274: "Proinde lectionis cursam salebris quibusdam lamis, voragininibus ac similibus obicibus intersecat, quaedam admiscens, quae vel fieri non potuerint aut possint vel ...." This passage was added in 1523 and is not in the text of the Leiden edition.

(29) Saussure, 65-70; Peirce, 99.

(30) Traditionally the consensus has been that Erasmus held pretty conventional views on the subject, regarding words as separate and subordinate to matter, though closely related. Marjorie Boyle has made the argument, however, that the res of Erasmus's res et verbum need not be thought of as necessarily nonverbal ideas or thoughts. Boyle observes that in the Apologia de "in principio erat sermo" Erasmus reverses Augustine's assertion that primordial speech is thought, maintaining instead that thinking is a form of talking to oneself. Along the same lines, Terence Cave, 19, points out that imitation as it is described in the De copia means following other writers and texts rather than miming reality or ideas, with the result that the hierarchy between res and verbum is already "disturbed if not inverted" -- the res is "already there, embedded in language." Cave also takes up the idea of linguistic practice, though his development of idea is considerably more Derridean than mine.

(31) In the Ecclesiastes the relation between word and thought is conceived of in somewhat more organic terms, words and figures functioning as the flesh and skin for the body, the clothing of the bones and the nerves. ASD V-4 279:705-280:715 (LB 5:861-62): "Inuentio quae res suppeditat, tametsi reuera complectitur et eloquutionem et ordinem, hoc est in oratione quod ossa in corpore animantis, quae nisi solida sit, caetera omnia collabuntur. Dispositio sine ordo, hoc est in oratione quod nerui in corpore animantis, parses orationis apte inter se connecters . . . Praeterea eloquutio quae verba et figures ad rem appositas suggerit, hoc est in oratione quod caro et cutis in corpore, decentur conuestiens ossa et neruos."

(32) Hoffmann, 1994, 90.

(33) Bakhtin's concept of the dialogic nature of all discourse is a fundamental part of my idea of praxis, but I do not agree with all that he has to say about the dialogic. For instance, the quite precise distinctions he draws between poetic, dramatic, and novelistic discourse in his essay "Discourse in the Novel" seem unclear to me. For more on the dialogic, see also Bakhtin, 1986(1) and 1986.(2)

(34) In the Ecclesiastes, Erasmus again explains that the same principles can be applied to both amplificatio and concisio. ASD 5, 5:61, 184-85 (LB5:974C): "Singula faciunt ad augmentum, et quae diminuuntur, et quae augentur."

(35) The phrase "a Protean variety of shapes" actually applies to another demonstration of copia where Erasmus transforms the sentence "Your letter pleased me mightily" into over 145 different "shapes" (Erasmus, 1978, 348-54).

(36) In another demonstration of the relation between compression and copiousness, Erasmus shows how a very succinct passage from the Aeneid -- "the plains where Troy once stood?" -- can be unfolded into a more ample Virgilian evocation of the same moment. As Betty I. Knott's annotations show, Erasmus takes the initial phrase from Aeneid 3.11, whereas the elaboration that follows is a conflation of non-sequential lines from the second book of that work (De copia 298, f's. 15 and 17). She does not comment, however, on how this example seems to be more of a variation than an elaboration, or on the assumption implicit in this rather odd exercise -- that rearranging the lines in a text does not impinge on its integrity as representation. My own feeling is simply that Erasmus's view of method infers a rather loose idea of representation.

(37) Erasmus to Grimani ([151/1 letter 710) in Erasmus, 1984, 2-3.

(38) Chomarat, 1981(2), shows in some detail how Erasmus applies the devices described in De copia to "develop" the basic text of Scripture.

(39) LB 9:1116B: "I am si nihil adderem ei, quod in Sacris Litteris habetur expressum, non essem Paraphrastes, nec Explanator. Mihi satis est quad quae adjicio non dis crep ant a Scripturarum sensibus . Et t amen hoc erat tuae libertatis , indicare loca quae damnas."

(40) Sider, 20. For "in terms quite plain and clear," see Erasmus, 1984, 14: "immo omnibus crassius et explanatius loquentem" (LB 7:777-78).

(41) Erasmus, 1991, 38. Further references to this translation will appear in the text.

(42) Erasmus to Archduke Ferdinand ([1523] letter 1333) in ibid., 3.

(43) Erasmus to Luis Nunez Coronel [April 1522], in Erasmus, 1981, lines 41-43. "For a paraphrase is not a translation but something looser, a kind of continuous commentary in which the writer and his author retain separate roles."

(44) "Erasmus, 1985, 455-56.

(45) Cicero, I.xxv.25: "Sic igitur, inquit Crassus, sentio naturam primum, atque ingenium ad dicenum vim afferre maximam; neque vero istis, de quibus paulo ante dixit Antonius scriptoribus artis, rationem dicendi et viam, sed naturam deluisse."

(46) Quintilian, XII.i.3: "Neque enim tantum iddico, eum, qui sit orator, virum bonum esse oportere, sed ne futuram quidem oratorem nisi virum bonum."

(47) Erasmus, 1933, 178 (LB 5:76A): "non tantum purum ab omnibus, quod fieri potest, vitiorum inquinamentis, verumetiam ab omni cupiditatum tumultu tranquillum ac requietum, quo expressius in nobus, velut in amne placido aut speculo levi et exterso, reluceat aeternae illius veritatis imago."

(48) See note 44.

(49) Bateman is citing Payne, Rabil, and Smith, 12.

(50) Payne, Rabil, and Smith, xvii; LB 9:879A.

(51) Chomarat, 1981, 30.

(52) See footnotes to Erasmus, 1991, 249-58.

(53) Onians, 916.

(54) Trinkaus, 13, comments that Erasmus follows humanist practice by understanding usage "as the index of what meaning can and cannot be derived from language" rather than the equivalence of theological truth. I go further, and I think Erasmus did too. Although Erasmus may have begun thinking of usage this way, he found finally that it is an "index" that can never be perfected since it inevitably fails to fully index its own meaning.

(55) Bakhtin, 1986(2), 106.

(56) See Bentley, 1976; idem, 1978; and idem, 1983, 112-93, 194-212. See also Schwarz, 92-166; and Bouyer.

(57) Erasmus to Leo X, letter 384 (1516), in Erasmus, 1974, lines 44-55.

(58) Bentley, 1976, 52, states that Erasmus had "no idea of reconstructing an archetype or a so-called 'neutral', uncorrupted text of the New Testament, yet he had enough confidence in his erudition and his critical faculties to suppose that he could produce a text of the New Testament that was closer to the original than that underlying the Vulgate translation." I would argue that the logical end of Erasmus's methodology goes beyond producing a good copy -- even if he realizes that a perfect copy cannot be achieved in practice.

(59) LB 6:704D: "atque ita primum scriptum erat in exemplar), quae viderim, omnium emendatissimo, Bibliothecae Paulinae: quod tamen ita fun' erasum, ut evidentissimum rasurae vestigium testetur adbuc germanam scripturam."

(60) LB 6:143E: "In Constantiensi codice rasura testabatur depravationem."

(61) See Erasmus. 1993. especially Harry Vredeveld's introduction to the volume. xii-xlix..

(62) Translation in Rummel, 1986, 132. LB 6:756E: "Porro codices eius generis nihil aliud sunt, quam amussis alba in albo lapide."

(63) Bentley, 1978, notes that Augustine argued for the priority of a harder reading at least once in his New Testament commentaries.

(64) Translation in Rummel, 1986, 117. LB 6:742D: "Et quoties Veteres satentur lectionem esse diversam, semper mihi suspeaior esse soles ea, quae prima specie videtur absurdior; ut consentaneum sit; Lectorem vel parum eruditum, vel parum attentuni, offensum absurditatis imagine mutasse scripturam." Rummel comments that Erasmus's own remark has to be subjected to the process of the "harder reading" here. See Bentley, 1978, 318-20, for a comprehensive analysis of this particular difficulty.

(65) Bentley, 1978, 313-16.

(66) Ibid., 310.

(67) See McConica, 1969, 93; also Boyle, 15-17. These authorities include Cyprian, Tertullian, Augustine, Hilary, Ambrose, Laaantius, Claudian, Prudentius, Aquinas, Cardinal Hugo, Bede, Remegius, Hugh of St. Cher, Nicolas of Lyra, Anselm of Laon, and Anselm of Canterbury.

(68) Hoffmann, 1994, 164. Hoffmann, 181-82, suggests also that the process of colIatio, which Erasmus recommends as a means of coping with murky scriptural passages by comparing one passage with another, is an approach to arriving at a kind of consensus within the text. In an ecclesiastical sense, Erasmus embraced the idea of the true body of Christendom as a consensus of believers not necessarily defined by absolute conformity to Institutional doctrines. "By church I mean the consensus of Christian people throughout the world" he maintains (letter 1729, lines 25-27, quoted and trans. in Tracy, 225). Tracy argues that Erasmus increasingly relied on the idea of consensus as the pressures of the Reformation forced him to separate himself more firmly from Protestant opinions. Bainton, 194-95, however, maintains that Erasmus appealed to the consensus "only on theological matters which he considered insoluble or inconsequential" and was willing to deviate from the generally perceived consensus on moral matters to which he gave priority.

(69) Erasmus, 1933, 205 (LB 5:90B): "et sicubi manifestius lapsi sunt, quam ut dissimulari possit, reverenter ab illis dissentiamus, non insectantes conviciis humanos lapsus, sed quod licet attenuantes atque purgantes."

(70) Translation in Rabil, 120. From the Annotations on Romans 5:12, LB 6:589B "Fatemur Ecclesiam habere auctoritatem interpretandi Scripturas, sed Ecclesiae Doctores, quam libet celebres, in multis Scripturae locis haesitant, multa varie, nonnulla etiam perperam sunt mterpretati." Erasmus reveals the same pattern of making a general assertion which he then deconstructs with qualifications when he discusses the Ciceronian standard for Latin eloquence. In the De copia (Erasmus, 1978, 312) he is willing to admit that Latin achieved the greatest eloquence at the time of Cicero but then immediately starts to rescind that statement by pointing out that there was considerable variation and inconsisteng among contemporaneous Latin authors.

(71) Annotationes in Annotationes Erasmi (Paris: Gourmont, 1520). On Edward Lee, see Coogan, 1986 and 1992; and Rummel, 1989, 1:95-120.

(72) Erasmus, 1984, 19-20.

(73) Ibid., 19-23, and passim.

(74) Ibid., 24.

(75) Rummel, 1986, 111-13. See 2 Corinthians 8, note 9, and Matthew 7, note 1, for words that slip from the margin into the text; 2 Thessalonians 2, note 25, for a word substitution; Matthew 3, note 2, Matthew 4, note 5, and Romans 3, note 4, for places where the perfect tense has been substituted for the future tense, Acts 17, note 25, for where Erasmus hypothesizes that someone misread a Greek letter and turned "elements" (stoichoi) into "Stoics" (stoikoi); and Romans 1, note 66, and Romans 8, note 49, for Erasmus's comments on the forgetfulness of scribes.

(76) Gadamer, esp. 295, 340, and 379-97.

(77) "If we think that the authority of all Holy Writ was to collapse directly the slightest error was ever found in it, and if it is certainly more than plausible that in all texts now used by the Catholic church none is so perfect as not to have been blemished with some error either by chance or intention -- if one accepts the above premise one must either deny all mistakes or say that all belief in divine Scripture must collapse" argues Erasmus in his Annotations. Quoted and trans. in Rummel 1986,

(78) For an interesting contemporary discussion of how this can work in a secular context, see Davidson.

(79) Hoffmann, 1994, 86.

(80) Erasmus to Johannes Sixtinus (November 1499) letter 116, in Erasmus, 1974, line 42. Further references to this translation will appear in the text.

(81) Erasmus, 1933, 177-78 (LB 5:75D): "aut certe Mercuriales illas statuses polykephalous, quae quondam in compitis pond solitae suo nonnunquam indicio viatorem eo provehunt, quo nunquam ipsae sint perventurae." See Post, 317-20, on the the of the Modern Devotionalists.

(82) Boyle, 66.

(83) Hoffmann, 1994, 155-56, has pointed out that this pattern of a via possessing a beginning, middle, and end is always counterbalanced in Erasmus's thought by a topoi motif characterized by profound moral and hermeneutic dualism. Perhaps it is this double aspect, this placing of the spatial along the route of the temporal, that gives the Erasmian via its plentiful, digressive, and unfolding character.