The narrator of Martha Cooley's The Archivist, Matthias, wonders near the end of his story how to negotiate between a "writer's hunger for privacy" and "the reader's appetite -- voracious, insatiable -- for more words" (322). Michael, the narrator of Bernhard Schlink's The Reader, faces an issue that also has to do with texts and desires, "dignity and freedom" and "accrued or inherited responsibility," in his father's words (142-43). Each narrator bears some responsibility for the suicide of a loved one; each displays a certain emotional distance; each struggles with a personal response to the Holocaust. Each text also foregrounds issues of reading and writing in the characters' lives, connecting these issues to responsibility, guilt, and atonement. Last but not least, each text offers some grounds for distrusting the primary narrating voice.

Because of these similarities and because of one significant technical difference -- The Archivist includes a substantial section from a voice distinct from the narrator -- the two books constitute a fertile field for exploring the relationship among narrative structure, the ethos that is suggested by that structure and conveyed by a text's main voice or voices, and the moral responses of actual readers. Wayne C. Booth, in The Company We Keep: An Ethics of Fiction, connects ethos to technique, and the technical difference between The Archivist and The Reader could cause readers to deem the former, as Booth would say, "better company" -- it may lead us to "desire better desires." Finally, statistical analysis of a sample of actual readers' responses to these novels as posted on Amazon.com demonstrates that there may be more of a discrepancy than we would expect, based on the sort of anecdotal evidence cited by Booth, between what theory predicts and the behavior of ordinary readers.

This study is consistent with much other work done during the past two decades on the relationship between moral development and narrating. The psychological researchers Mark B. Tappan and Lyn Mikel Brown, for instance, demonstrate that narrating and moralizing are inseparable and that "authorship of one's own moral story" partakes of the telos of moral development (188, 190-91). Similarly, Paul C. Vitz makes the case that "One may view narratives as the laboratory of moral life" (718). Such philosophers as Colin McGinn, Alasdair MacIntyre, and Peter Levine provide more discursive bases for the need to ground moral understanding in "narrative thinking" rather than "prepositional thinking" (Bruner's terms; 11).

However, these researchers and theorists stop short of taking Booth's step: connecting literary technique to the ethos conveyed by a text and thence to its potential effect on readers. They all agree either explicitly or implicitly that some moral stances are preferable to others, and Levine and McGinn argue persuasively that humans have an innate moral sense. This is not to say that all humans agree on what is and isn't moral but that every culture has actions that it labels, in terms evoking disapprobation, what we call for example murder and betrayal. As McGinn says, "we know that stealing is wrong just by knowing what stealing is" (39). Because of this broad confluence of philosophical argument and research data, taking that next step is important in order to clarify the potential relationships between ethos and moral response and then to determine whether the potential is realized.

In The Company We Keep, Booth explores in detail the fact of ethical transactions between a text and its audiences. Booth coins the term "coduction" to denote "what we do whenever we say to the world (or prepare ourselves to say): 'Of the works of this general kind that I have experienced, comparing my experience with other more or less qualified observers, this one seems to me among the better (or weaker) ones, or the best (or worst). Here are my reasons'" (72-73; Booth's emphasis). Coduction differs from logical and scientific processes in that it can never be "apodeictic" and "can never be performed with confidence by one person alone" (73). Coduction is essential for any discussion of ethics and narrative: "In sharing descriptions of narrative experience we can always clarify and sometimes simultaneously refine our coductions of value. Thus we come to explain what we halfway discover in experience: we do have grounds for pursuing our natural temptation to debate about ethical qualities" (182).

For Booth, "the ethics of narrative is inherently a universal subject: in the beginning, and from then on, there was story, and it was largely in story that human beings were created and now continue to recreate themselves" (39). The study and discussion of this universal cannot escape a certain tautology because it must assume "from the start at least one of its major conclusions: some experiences with narrative are beneficial and some harmful" (40). This "formal circularity" doesn't negate the importance of the endeavor but does mean that ethical critics can't claim objective, scientific validity for their results (40). Keeping constantly in mind both the universality of narrative in human experience and the resulting circularity of method, Booth identifies a principle that helps explain the working of narrative: it can lead people "to desire better desires" (271; Booth's emphasis). He argues that someone who engages with a narrative in the context of an ethical system that is reflective, pluralistic, and subject to revision is also engaging more of the "whole self" than is someone who looks for "simple moral choices of the kinds that traditional codes are at least said to have provided" (271-72). Such a person is able "to form a second-order desire" and to make a "judgment about the ethos-I-would-prefer-to-have-and-will-therefore-cultivate" (271).

This line of reasoning leads Booth to pose what he terms "the key question in the ethics of narration": "Is the pattern of life that this would-be friend [the implied author of the text] offers one that friends might well pursue together? Or is this the offer of a sadist to a presumed masochist? Of a seducer or rapist to a victim? Of the exploiter to the exploited? Is this a friend, a lover...?" (222). Thus Booth values narrative works that do something impossible for "discursive philosophy," works that "'display to us the complexity, the indeterminacy, the sheer difficulty of moral choice'" (Martha Nussbaum, qtd. in Booth 288). Booth frequently insists that "no narrative will be good or bad for all readers in all circumstances" (489) but also insists that narrative as a form is universally powerful because individual narratives invite readers to share "patterns of life" by experiencing those patterns during the time of reading. And as suggested by Booth's examples of types of patterns, ethos may be the key element in a reader's reaction.

Booth's argument is much more subtle and detailed than can be conveyed in a brief space. I draw on him here because he explicitly connects narrative's ethical value to the recognition that even though we almost certainly have no "universal standards" for narrative or "universal supreme goods" -- aside from such tautologies as "'It is good to live life well' or 'It is always good to improve one's soul'" (56) -- nevertheless "we must take responsibility for what we are to become" (271), that is, for the desires that we desire. For this reason Booth (at least the ethos he displays in this book) is what theorists of moral development would term a "committed relativist": recognizing that moral choices are almost never simple and that there are no universally applicable rules, he still insists on nearly every page that choices matter, that we can become more or less admirable as humans depending on the desires we choose to desire, that "the company a reader keeps" when reading shapes the ethos of that reader. Theorists also typically rank this level quite high in comparison, for instance, to dualism (wanting to be able to make yes/no decisions).¹

Booth's discussion is especially useful for the connection he establishes between the technical elements of a text and the ethos conveyed by the text. Discussing a Yeats poem, including Yeats' revisions and some less successful versions Booth concocted, Booth says that "ethical appraisal" can't be separated "from what looks like judgment of sheer craft." He amplifies thus:

In discovering that we know something about the quality of this poem -- that at least some of its parts are well chosen for their task -- we already discover something about the ethos of the poet: he shows the integrity of a devoted craftsman. My substitute versions not only do not fit the poem; they imply a maker who is sloppy, vulgar, careless about cliché, inattentive to form -- in short, a bad craftsman. Thus, regardless of whether I think the whole poem a major achievement, I have already found a meeting point of craft and ethics: the poet has met a demanding ethical standard by honing at least some of the parts so fine that I can think of no possible improvement. (107-08)

Applying Booth's approach to The Archivist by Martha Cooley (published in 1998) and The Reader by Bernhard Schlink (published in German in 1995, English translation published in 1997),² one finds that because of its greater narrative complexity, The Archivist leans more toward the Milton side of the spectrum while The Reader is somewhat more in the direction of the limerick. To put the point another way, Cooley's novel asks more of a reader and thus exhibits a more desirable ethos than does Schlink's. The design of The Archivist is more likely to engage a reader's "whole self" because of its two voices; this pluralistic structure may appeal to readers either who occupy the relatively advanced moral stage of committed relativism or who are close enough to that stage to realize that it is worth desiring.

Obviously such a comparison cannot be quantified, because the two narratives differ in other ways as well, especially in length. There are no universally estimable narrative techniques; I can't suggest that a novel with five narrators is five times as likely as a novel with a single narrator to appeal to morally advanced readers. It is easy to imagine a multi-voiced novel displaying little if anything worth desiring to desire, just as there must be some ironic or otherwise ethically stimulating limericks. However, describing the kind of narrative experience offered by each of these novels follows Booth's lead -- suggesting "grounds for pursuing our natural temptation to debate about ethical qualities" (182) by relating ethos to technique.

According to the model of narrative transmission I developed in Rhetorical Narratology, a reader can adopt any of three basic positions with respect to a narrative and will shift among at least two of these many times during the reading. The reader can participate in the narrating audience (lose oneself temporarily in the narration), can play the specific role of the narratee (the internal addressee of the narrator), and can participate in the authorial audience (treat the text as a construction intended to move an audience to action, thought, or emotion). The authorial reader of course is not responding to the actual author but to a personalized intention that is inferred from the text and that is traditionally referred to as the "implied author." (Because this term has such wide currency and is used by Booth, it is preserved here, although some theorists deem it unnecessary. See for example Nünning.) Neither The Archivist nor The Reader offers readers an explicitly invoked narratee position (the narratee position being limited to such explicit invocations and thus diverging from Gerald Prince's seminal definition), so we shall concentrate on the narrating- and authorial-audience positions.

Readers easily move between these two positions during the reading experience. Playing the narrating-audience role results in the experience of losing oneself in the world of the work or in the experience of taking the narrating as actually happening. Reading of a tornado is unlikely to send me to the basement, but a vivid description by a terrified first-person narrator will at least mildly stimulate my emotions and allow me to imagine that an actual voice is telling me about the tornado. I won't necessarily share that voice's fear, but I recognize the fear as real within the world in which the voice is speaking. The questions of whether I'm "pretending" to react to the fear and whether the fear is "real" are complex, as shown by the exchange between Kendall Walton and Robert Newsom in the pages of the journal Narrative. Because our interest here concerns the pragmatics of response, it suffices to state that within the frame a reader applies when reading novels, the reader takes the voice as really afraid, as long as the reader is playing the narrating-audience role. To play the authorial-audience role is to resist being drawn in by the narrating; playing this role, the reader will look at the text as embodying an intention, like almost all human artifacts.

The degree and type of an actual reader's involvement with and placement within these possible positions can never be predicted with certainty. However, a shift in narrating voice, especially if unexpected, may cause a reader both to change the characteristics of the narrating-audience role the reader has adopted and to shift at least briefly out of that role and into the role of the authorial audience. The appearance of a second voice can remind the reader that the narrative is an artifact and has designs on the reader, and when the reader returns to the world within the narrative (that is, plays again the role of the narrating audience), the second voice can place different demands on and imply different expectations of the narrating audience. Such a shift may increase the reader's investment both emotionally, because the new narrating voice requires a different perspective, and intellectually, because the shift foregrounds the novel's discourse rather than its story. If this increase happens -- if the reader becomes aware of the narrative making greater demands -- then according to Booth the reader's estimation of the ethos behind the text (the implied author in Booth's terms) will probably rise. In addition, because a shift in narrating voice both makes the narrating situation more complex and introduces the possible need for discriminating between the voices, the shift may also appeal to those readers who recognize the desirability of the moral position of committed relativism. That is, readers may decide that the narrative is inviting them to weigh competing claims and to esteem one more highly, a task that is probably not amenable to the imposition of an absolute standard.

A shift in narrating voice is one means of creating tension in the novel's discourse. As explained in Rhetorical Narratology (drawing on the work of James Phelan), a reader's progression through a novel is influenced by instabilities in the story and tensions in the discourse. Instabilities are fairly easy to discuss, because they have to do with human dimensions within the story (conflicts between and within characters, for example). Discourse tensions are necessarily more abstract, but in general they have to do with the story's "how" rather than its "what"; they foreground the novel's synthetic nature and may also contribute thematically and mimetically. Because they necessarily distance a reader, if only momentarily, from the role of narrating audience, they can make the reading experience more complex. (Discourse tensions may also frustrate readers whose main goal for reading is to exist within the world of the novel -- to remain in the role of narrating audience.) That is, discourse tensions can lead a reader to ask a question that contributes directly to the reader's experience of the implied author's ethos: what does the author expect me to do with this tension? Both a narrator's perspective and a narrator's reason for narrating can establish discourse tension; this will usually happen because the reader senses a discrepancy in values between what the implied author seems to expect of the authorial audience and what the narrator seems to expect of the narrating audience.

Michael Berg, the narrator of The Reader, chronicles his relationship with Hanna Schmitz, a woman in her thirties who seduces the 15-year-old Michael. Hanna remains a mystery to him in many ways through their half-year relationship; the mystery is enhanced when she disappears. She reappears a few years later as one of the defendants in a war-crimes trial being observed by Michael's law class, ultimately receiving a life sentence because her co-defendants portray her as their de facto leader and author of the report that covered up the crime for which they’re now being tried: allowing a group of prisoners to burn to death in a church. Michael realizes that the sentence is unfair because, as he deduces, she could not then and still cannot read and so would not have been the person in charge as alleged by her co-defendants; he suddenly understands why an important element of their relationship consisted of his reading aloud to her. (Hanna's illiteracy in fact is offered as one contributing cause of her decision to join the SS.) He does nothing with this knowledge; however, eight years later he locates where she is imprisoned and begins to record and send her cassette tapes, at first of classics and then of a broader range of material, including his own writings. As he later learns, these tapes stimulated her to learn to read and write; she actually begins sending him letters, although he never replies with anything similarly personal. After serving 18 years of her sentence she is granted clemency. On the day of her release, when he is to pick her up and help her become reestablished in society, she hangs herself. The narration takes place 10 years after her suicide, which places it approximately in 1993, or two years before the first publication date of The Reader.

A retrospective narrator, Michael in the final chapter spells out his role as the writer of this story, a tactic that can draw an authorial reader’s attention to this narrator’s ethos as well as to the ethos of the implied author:

Soon after her death, I decided to write the story of me and Hanna. Since then I've done it many times in my head, each time a little differently, each time with new images, and new strands of action and thought. Thus there are many different stories in addition to the one I have written. The guarantee that the written one is the right one lies in the fact that I wrote it and not the other versions. The written version wanted to be written, the many others did not. (216-17)

Michael's privileging of the final version of this story is consistent with his practice in sending tapes to Hanna in prison. When he begins sending tapes of his own work, he sends only the final version, never a working draft, nor does he ever personalize the packages:

When I began writing myself, I read these pieces aloud to her as well. I waited until I had dictated my handwritten text, and revised the typewritten version, and had the feeling that now it was finished. When I read it aloud, I could tell if the feeling was right or not.... Hanna became the court before which once again I concentrated all my energies, all my creativity, all my critical imagination....

I never made a personal remark on the tapes, never asked after Hanna, never told her anything about myself.... When the text was finished, I waited a moment, closed the book, and pressed the Stop button. (185-86)

The narrator's perspective, considered as a discourse element, also contributes to the less than estimable ethos. Michael Berg is the only person telling Hanna's story. We may be inclined to think fairly well of him because he admits his own limitations and was seduced by Hanna (if we believe his version of their affair), but we only hear of Hanna what he chooses we should hear: he selects the dialogue and letters, he selects the pieces of the trial transcript; he determines which version of the story is "true." "Atonement" seems important to Michael (e.g., 212). By writing this story, he may be atoning for loving and then betraying a criminal; by reading the story, the narrating audience may be participating in his absolution. This participation carries a cost, because the actual reader is forced to share in the revelation of Hanna's secret, her illiteracy. To atone for such complicitous participation, the reader may try to imagine Hanna's story and may note that Michael draws no ethical conclusions from this lifetime of experiences. The reader can thus rise above Michael’s level and try to give Hanna the voice that he denies her. This particular exercise may not occur to all or even most readers, but it's a possibility embedded within the discourse tension between the narrating audience's face-value acceptance of Michael's story and the authorial audience's curiosity about the entirety of Hanna's story.

These technical points -- the thematizing of the narrator's authorial role and the filtering of Hanna's story entirely through the narrator -- make more problematic the relationship between the narrator's ethos and that of the actual author. On the one hand, The Reader offers readers a set of complex moral issues invoked by Hanna's trial, by Michael's conflicts of conscience regarding both her complicity in the death of the prisoners she was guarding and his silence about her secret, by his refusal to admit their relationship to his friends and family, and by the whole concept of "war crimes." The novel also suggests that these issues can't be assessed by absolute standards. Michael's father, a professor of philosophy, offers a principle for resolving Michael's conflict over whether to reveal Hanna's secret during the trial: "If one knows what is good for another person who in turn is blind to it, then one must try to open his eyes. One has to leave him the last word, but one must talk to him, to him and not to someone else behind his back" (143). This principle is a good example of committed relativism: "one must" make the effort, although the final decision belongs to the other person. That Michael can't bring himself to apply the principle suggests that his moral development is still somewhat limited at this point in his life. He replies to his father, "And what if you can't talk to him?" (143). The story identifies no barrier between Hanna and Michael aside from the latter's unwillingness to approach his former lover. Thus a reader who recognizes the value of committed relativism may esteem this novel because it dramatizes the plight of someone who faces extremely difficult questions but seems to lack the moral grounding to answer them.

On the other hand, readers may wonder if the author realized just how flawed, morally, the narrator is. This question came up repeatedly in a reading group that was discussing this novel: the group members, twenty-two undergraduate and graduate students and faculty, tended to want to know Schlink's position with respect to his narrator. They wondered if the "would-be friend," the creator of this text, was offering them "a pattern of life" worth pursuing or if like his created narrator he possessed at least some of the qualities of an "exploiter." Most members of the group concluded that they couldn't be sure, because there were no signals marking a distance between author and narrator. Booth relates discourse tension explicitly to the ethics of reading: "The combination of these two responses, sharing the fixed norms of the implied author and only pretending to share certain others, creates a complex tension in every fictional transaction" (145). The problem is that those "fixed norms" cannot be identified. Michael isn't practicing coduction when he uses an imaginary Hanna as his "court." Does Schlink have Michael make that point in order to damn himself? Michael "never asked after Hanna"; should he have done so?

It's important to recognize that for the type of ethical criticism advocated by Booth, there is no such thing as "intentional fallacy." Coduction is a natural process, part of "our natural temptation to debate about ethical qualities" and dependant on the fact that people tell stories for a reason; even though we never can be sure about an actual author's intention, we naturally try to infer an ethos and an intention from a text. We may decide that we desire not to be like Michael Berg or that we desire the desire to be more committed than he is able to be. To this extent reading the novel may have a salutory moral effect, but there remains the question that some readers may want answered: is Schlink, like Michael, exploiting Hanna's story? If there are no technical elements that can reasonably be agreed to distinguish the implied author from the narrator, and if as a result the reader comes to suspect that Schlink is presenting autobiography as fiction, then Schlink will be placed in the category of undesirable companions. This outcome (reading the story as fictionalized autobiography) is at least somewhat likely for the reader who calculates the story's chronology and realizes how closely the ostensible date of Michael's narration matches the actual date of Schlink's composition.

The narrator of The Archivist, Matthias Lane, seems as distant and passionless as Michael Berg. Like Michael, Matthias is narrating retrospectively from a time very close to the present. His story is really two stories: that of his marriage to Judith, who committed suicide in 1965 after being institutionalized for six years, and that of his relationship with Roberta Spires, a much younger poet who wants to read the letters sent by T.S. Eliot to Emily Hale that are under his care and are to remain sequestered until 2020. Cooley creates complex and interesting parallels among the characters: Matthias seems troubled by Judith's physical attraction to and need for him in the same way that Eliot was troubled by that of his wife, Vivienne; Judith learned as a child that the people she thought were her parents were not, and Roberta learned that her parents were born Jewish; Judith too was a poet.

These complexities are reinforced by the novel's structure. After a one-page opening set in what seems to be the narrating present (although only re-readers will pick up the plot hints dropped here), Matthias tells the two stories in chronological order, alternating between them. Then, roughly one third of the way through the novel and with no warning, the reader encounters a new part, labeled "Two," and is plunged into a journal kept by Judith:

Hayden House

April 20

Light came first, then words: in the beginning the luminous mirror. First light, then language; first the en-sof, the unknowable, then the letters of the Torah all jumbled, from which somehow we were supposed to construct a world--somehow!--though how did He expect us to do this once Adam the dust-man had broken the sacred vessels and scattered the light like motes of his own pale dust everywhere, leaving what the Kabbalists call shards. (109)

How does Judith's journal reach the reader? For mimetic plausibility, it should come from Matthias, who might introduce it something like this: "I opened the package and sat down to read." He does receive it from Dr. Clay, who wrote that he was legally required to return all of Judith's effects to her next of kin and had only read the last page, where typically the suicide note would be found (244). And Matthias does refer to reading the journal. But this explanation of the transmission violates mimetic logic, because the journal is introduced almost 150 pages before he refers to it. The other conventional explanation, that the journal is part of the novel's world and the reader is being given access to it by an implied author, is also flawed. Such a transmission would constitute a privileged communication between implied author and authorial readers, with the implication that the document is being truthfully and fully transmitted. However, some details don't fit this implication. Matthias says that he read the journal twice in a period of twelve hours (245); even taking into account that a person reading such a document might read it quite slowly, that seems a long time for a typed text of 106 pages. (I read it at what felt like a leisurely pace in ninety minutes.) Another problem is that for a journal covering more than five years, 62 dated entries seems thin: this is an average of less than one entry a month. So there is no entirely satisfactory way to explain the presence of the journal in the novel.

This discourse issue -- how the journal reaches the reader -- pertains to the reader's potential awareness of being involved in something that was not to be made public. If the reader takes the implied author as the source, the reader probably senses that no violation occurred, because actual reader and implied author stand outside the world of the novel. But if Matthias is the source, then the reader may take the journal as an undesired communication about Judith, undesired because even though it presents interesting information it was to have been burned. Thus discourse tension contributes to one of the novel's overt themes, the conflict between the "rights" of the writer and the reader.

The possibility of the reader feeling an unasked-for complicity is similar to what may happen in the reading of The Reader, when we learn about Hanna the secret that she accepted a prison term of life in order to protect, but here the presence of Judith's voice compensates by doubling the perspective. The Archivist dramatizes "the sheer difficulty of moral choice" not just by posing issues that can't be resolved with recourse to traditional codes but also by allowing for an engagement with more of the reader's "whole self." The "pattern of life" offered by Cooley seems to include the principle that "all voices deserve to be heard," a principle that fits well with committed relativism. Matthias himself may strike readers as no less controlling than Michael. The novel ends with his burning of Eliot's letters (but not the accompanying drafts of poems), because he has decided that they were never meant to be read, and yet he has read them, in violation of his professional code as an archivist. (He also violated that code by giving Roberta photocopies of the poems.) Reading the papers and then concluding that they should never be made public could be admirable, but setting himself up as the judge of who else is worthy of them can strike readers as a mark of either hubris or confusion -- it does not indicate an ethos worth desiring. When Cooley places the reader in this same situation, however, she makes possible the experiencing of this "pattern of life" and thus may make more difficult the judging of Matthias by a simple code, for example that of consistency (he should allow either everyone or no one to read the papers). Also, Matthias may finally be a more estimable person than Michael Berg; whereas the latter kept rewriting the story of him and Hanna until he arrived at one that felt right, Matthias notes, of his "shared life" with Judith, that it "is not, after all, some story whose end I can keep rewriting. It was brought to a close" (101).

Judith's voice in the novel, even though isolated between the sections of Matthias' narrative and so in a sense disempowered, can also complicate a reader's sense of the relationship between her and Matthias as well as the questions of righteousness and atonement, thus contributing in this way as well to the novel's possible appeal to readers who recognize the desirability of committed relativism. She writes:

I want to be inscribed in the Book of Life as one of the righteous. But first I must be judged. No -- first I must judge myself. And atone.... I have to write myself in.... It doesn't just happen. (176)

What these differences between The Archivist and The Reader boil down to is that the former offers readers the opportunity to do more technical work with the implied author, in the sense of adapting to the two major shifts of perspective (into and out of Judith's journal). As Booth says, "craft and ethics" support each other. This isn't to say that The Reader fails or that Schlink is a poor craftsman but that considering the particular element of perspective, The Archivist may position readers in a more complex place, ethically, while also giving them the sense that the novel's implied author is better company by setting a higher standard of authorial responsibility. I phrase these points cautiously because one reader might decide that Cooley gave the matter of perspective no thought (a decision that could be supported by the lack of mimetic justification for Judith's journal) and that Schlink, by not creating ironic distance between his narrator and himself, forces readers to engage in the complex ethical task of assessing his stance toward his narrator. I think Booth would consider such a line of argument because of his commitment to pluralism. However, his "key question in the ethics of narration" -- "Is the pattern of life" offered by the implied author "one that friends might well pursue together?" -- would finally lead him to reject these alternate readings of the two novels' narrative perspectives. He would, I suggest, regard them as inconsistent with his image of "friends" sharing an inquiry. Booth might also suggest that these alternatives don't effectively lead readers to "desire better desires," because both assume a hierarchy rather than an equality in the transaction between reader and implied author, with the reading of The Archivist granting the reader superior knowledge about the author's business and the reading of The Reader granting the author a kind of distant superiority -- let readers figure it out if they can.

On the Amazon website, a potential buyer can read snippets from published reviews of a given novel (including reviews by the site itself) as well as reviews by individual readers.³ In addition to reading the reviews, the potential buyer can indicate whether or not a given review was helpful, can write a review, and can see a list of other titles that were purchased by people who purchased the book in question. The site also provides a sales ranking for the book (in January 2002, The Archivist was number 42,363, with The Reader coming in at number 5654) and may note cities in which a book is popular. An overall readers' rating is also given: by this measure, both received three and one-half stars, with five stars being the highest possible rating.

The personal reviews tend to be fairly short and to conform to the conventions of standard edited English. Most are probably written not by professional readers (academics and published reviewers) but by lay readers. They lack the relative sophistication and polish of papers written in college literature courses, but they don't simply recapitulate the plot or describe the characters. (The lack of summary may be due to this basic information being provided by the snippets from published reviews.) While there is no way of knowing how many individual reviewers actually read the published reviews, it seems likely that individuals write reviews with an awareness of this larger context, in which evaluative responses are the norm. This context probably suggests, to individual reviewers, parameters for their responses: they are writing to and for other readers, are participating in a dialogue, and will even be able to determine whether their reviews were deemed helpful by other readers. (Every individual review is accompanied by a yes/no "button" with which visitors to the site can assess the helpfulness of the review.) This dialogic context makes the reviews useful and valid as data: each review is an unsolicited response generated by an interested individual for an audience whose interest the individual can count on.

My preliminary study of these reviews led me to the impression that reviewers of The Archivist valued that novel's relatively complex structure and were somewhat more likely than reviewers of The Reader to value moral complexity. Reviewers of The Reader seemed somewhat more likely to express opinions that suggest a dualistic (black or white) moral stance and also seemed somewhat more likely to prefer to remain within the narrating-audience position.

To test the validity of these impressions and thus to test my hypothesis that a more complex narrative structure implies a more mature authorial ethos and invites a more mature moral response, I decided to have well-trained readers score a sampling of reviews of each novel. My scorers were five advanced undergraduate and graduate English majors, all with experience in holistic grading. I compiled five unique instruments, one for each scorer, by randomly selecting twenty reviews of each novel (from a total of 38 on The Archivist and 484 on The Reader) and then combining these 40 reviews in five different random orders. Each review was scored according to the following items, the first four using a five-point Likert scale and the fifth using a five-point "strongly positive to strongly negative" scale:

A. The reviewer's response is based on moral principles applied to the novel.
B. The reviewer's response is based on logical analysis applied to the novel.
C. The reviewer's response is based on elements of the novel's technique.
D. The reviewer's response is based on the plausibility of the story and the characters.
E. Rate the review's overall response to the novel.

Statistical analysis of the data (t-test for dependent means applied to the array of mean scores for each item and each novel and also applied to the differences between novels, item by item) partially confirmed these predictions. The most dramatic (and least expected) result was that readers of The Reader were much more likely to base their responses on moral principles (item A) than were readers of The Archivist; this result was significant to the p < 0.01 level. On reflection, I have determined that readers of The Reader may have reacted negatively to the novel because its subject matter included the seduction of an under-age boy by a mature woman, a reaction reflected in readers' strong preference (significant to the p < 0.01 level) for The Archivist (item E). Tending to support the theory is the third highly significant result (p < 0.01): readers of The Archivist were much more likely than readers of The Reader to base their responses on elements of the novel's technique (item C).

Obviously, much more work needs to be done on the relationships between technique and moral response. We humanists would like to believe, with Booth, that a writer who attempts Paradise Lost will be more esteemed by more readers than will a writer who attempts a collection of limericks, and that readers will grow more from traveling with such a writer. My study does suggest that within the population of general readers who are likely to post reviews to Amazon.com, technical expertise is valued, but my study also suggests that readers who may tend to respond on the basis of somewhat dualistic moral principles (seduction of a minor by an adult is always bad) may also be less aware of a novel's technical sophistication. Those readers may indeed have been responding to the ethos projected by the narrator of The Reader and implied by his creator. It may be the case that readers take an estimable ethos somewhat for granted: as long as the narrator is more or less worth traveling with, the reader feels free to attend to other elements of the narrating situation, such as plot and technique, but if either the narrator or the implied author is perceived as morally flawed, then the reader's response will be based more on moral principles and less on other considerations. This supposition is in line with Booth's assertion (echoed by all other ethical theorists I have consulted) that we are fundamentally moral creatures. At the very least, my study indicates that we teachers would do well to be careful how we challenge students, not assuming that Paradise Lost will always have more of an impact than the collection of limericks because it asks its readers to engage in more sophisticated technical work -- that technical invitation, to be effective, probably needs to be coupled with a relatively estimable ethos.

¹ The theorists and researchers on which I base this statement include Perry, Rich and DeVitis, Gilligan, Gibbs, and Feagin. There is disagreement about where to locate moral-development stages, how much these stages depend on factors such as gender and class, how best to identify an individual's stage, and even whether development proceeds step-wise or in a more recursive fashion. Nevertheless, the most often cited researchers rank more highly the ability to handle morally complex issues and the recognition that moral judgments often need to be tempered with sensitivity to local conditions.

² Issues of translation are not relevant, except insofar as I avoid certain interpretive moves that the title The Reader seems to invite. The English title allows for an ambiguity not permitted by the German title, Der Vorleser, but considerations of who is really "the reader" in/of this novel do not come up in the responses I study.

³ Thanks to an anonymous reviewer for the Rocky Mountain Review who provided an extremely thoughtful reading of an earlier version of the manuscript and suggested an empirical assessment of theory. That portion of the research was generously supported by a Faculty Research Development Grant from The University of Texas of the Permian Basin.

Booth, Wayne. The Company We Keep: An Ethics of Fiction. Berkeley: The University of California Press, 1988.

Bruner, Jerome. Actual Minds, Possible Worlds. Cambridge: Harvard University Press, 1986.

Cooley, Martha. The Archivist. Boston: Little, Brown, 1998.

Feagin, Susan L. Reading with Feeling: The Aesthetics of Appreciation. Ithaca: Cornell University Press, 1996.

Gibbs, John C. "Moral-Cognitive Development and the Motivation of Moral Behavior." The Role of Values in Psychology and Human Development. Ed. William M. Kurtines, Margarita Azmitia, and Jacob L. Gewirtz. NY: John Wiley, 1992. 222-238.

Gilligan, Carol. In a Different Voice: Psychological Theory and Women's Development. Cambridge: Harvard University Press, 1982.

Kearns, Michael. Rhetorical Narratology. Lincoln: University of Nebraska Press, 1999.

Levine, Peter. Living Without Philosophy: On Narrative, Rhetoric, and Morality. Albany: State University of New York Press, 1998.

MacIntyre, Alasdair C. After Virtue: A Study in Moral Theory. Notre Dame, IN: University of Notre Dame Press, 1981.

McGinn, Colin. Ethics and Evil in Fiction. Oxford: Oxford University Press, 1997.

Newsom, Robert. "Fear of Fictions" and "Doing Duality Once More." Narrative 2 (1994): 140-151 and 159-160.

Nünning, Ansgar. "Unreliable, Compared to What: Towards a Cognitive Theory of Unreliable Narration: Prolegomena and Hypotheses." Grenzüberschreitungen: Narratologie im Kontext [Transcending Boundaries: Narratology in Context]. Ed. Walter Grünzweig and Andreas Solbach. Tübingen: Gunter Narr Verlag, 1999. 53-73.

Perry, William. Forms of Intellectual and Ethical Development in the College Years: A Scheme. NY: Holt, Rinehart and Winston, 1968.

Prince, Gerald. "Introduction to the Study of the Narratee." Poetique #14 (1973). Rpt. in Reader-Response Criticism: From Formalism to Post-Structuralism. Ed. Jane P. Tompkins. Baltimore: Johns Hopkins University Press, 1980. 7-25.

Rich, John Martin and Joseph L. DeVitis. Theories of Moral Development. Springfield: Charles C. Thomas, 1985.

Schlink, Bernhard. The Reader. Trans. Carol Brown Janeway. NY: Vintage, 1997. Der Vorleser. Zurich: Diogenes, 1995.

Tappan, Mark B. and Lyn Mikel Brown. "Stories Told and Lessons Learned." Harvard Educational Review 59 (1989): 182-204.

Vitz, Paul C. "The Use of Stories in Moral Development." American Psychologist 45 (1990): 709-720.

Walton, Kendall. "Duality Without Paradox: A Response to Robert Newsom." Narrative 2 (1994): 152-158.