Robert Appelbaum • Pleasant and Useful? A Tale from the Middle Ages

About the blog: The Instrumental Narratives blog aims to popularize the insights and methods of narrative scholarship and features analyses of instrumental storytelling by high-profile narrative scholars. The analyzed cases deal with uses or abuses of the narrative form, storytelling practices or narrative sense-making in many areas of life: politics, journalism, business, identity work, artistic or literary sphere, activism, and forms of social participation. The blog texts evaluate possible societal risks or benefits of contemporary storytelling, for example through cases from the author’s own national, linguistic, or cultural sphere.

In the Middle Ages and the Renaissance, it was taken for granted that the purpose of telling a story was to provide a moral example. That meant showing something that readers ought to emulate, or else something that they ought to avoid. There was no idea that fiction, in other words, was anything other than instrumental. That fiction ought to entertain was also taken for granted, though. The ancient authority Horace said that art ought to be dulce et utile, pleasant and useful. So it would be a surprise to most storytellers that there is anything to worry about if fiction was meant to be useful, and the very idea of “instrumentality,” as in Adorno and Horkheimer’s famous injunction against “instrumental reason,” would have seemed odd.

But that didn’t mean that instrumentality in fiction was straightforward. The example of the perennially popular Aesop’s Fables apart, early European fiction writers knew that though a story was one thing, the interpretation of a story was another. Moreover, although it was pretty important to know, along with the grasshopper, that it was best to live like an ant, and always prepare for the future, the stories that attracted the most interest could be morally ambiguous. In fact, it was common for storytellers to say that a story meant one thing when it actually meant another, or else to put out a story, salacious though it may be, and pretend that it expressed a strait-laced moral truth which it couldn’t possibly be expressing.

* * *

Here is an example from a medieval French fabliau (a comic short story) by one Hues Piaucele, collected (and translated) in what is called the British Library Manuscript.[1] It is about a loving couple named John and Yfame, who had recently fallen on hard times. Knowing this, three separate monks from the local monastery offered substantial amounts of money to Yfame if she would have sex with him. She indignantly told John about the monks and John came up with a plan for revenge, as also a plan for getting a hold of the monks’ money. He told Yfame to accept their offers, and have each of them come to their home at different times one evening the next week, saying that her husband would be gone from the house that night.

In came the first monk – the narrator loves adding the monk was old and fat. He put his money on the table and went after Yfame in the middle of the main room of the house, bringing the two of them down to floor. John, who had been hiding in the loft, came down and whacked the monk on the head with a club, instantly killing him. John took the corpse outside, dumping it beside a tree, where he thought he would later bury him, and collected the cash. In came the second monk, putting down his money and trying to take Yfame on her bed. Again, John whacked and killed him, and dragged the corpse out of the horse for burial. In came the third monk, and the same thing.

The narrator doesn’t spare his readers or listeners the gory details. On the first occasion, for example, we are told that John

who threw himself upon them,
Very hard with the club:
He hit his head so hard
That the blood and brains flowed out.
The man fell dead speechless.

But now it was getting late and John had the hard work ahead of him of burying the three men. And he was exhausted, though delighted at all the money he has suddenly accumulated from the monks’ down payments on sex they would never enjoy. So he asked his wife to go fetch her nephew Estormi, a simpleminded young man much addicted to gambling. She found him in a tavern, losing at cards, and offered to pay his debts if he came right away and helped her in a private affair.

When Estormi arrived at John and Yfame’s home, John told him that the body of a demonic monk had come back from the dead, and his body needed to be buried as soon as possible to prevent him from coming back to life altogether and cause who knows what kind of harm. So Estormi buried him. But when he came back to the house, John showed him the second monk, pretending that it was the same monk, again arisen from death. “There he is again.” Go bury him, John said, lest he return another time. Estormi did what he was told. But then, coming back to the house, he was shown the third monk, the same monk again, it is alleged, arisen one more time. With much effort, worn out by the previous two burials, Estormi went to work again, and found himself successful at ridding the world a third time of the ghostly priest. But as he was finishing the job, he saw a fourth priest walking past him down the street. “Look,” Estormi said, “this priest is getting away from me! By God’s ass, he’s going back! What is this, Sir Priest?”

Estormi took his shovel and attacked the fourth priest. As he reports the incident later on,

“And I gave him one with the pick
So hard that I made
His brains flow out on the street.
Then I took him, and I went back
Down there by the back door.
And I threw him down;
I stuck him into a mud pit.”

When the husband heard this he was flabbergasted. And he said aloud, though in a low voice,

“In faith, now things are going worse,
Because this man [the fourth  monk] hadn’t done anything wrong here.
Someone is paying the penalty
Who has not deserved death for it.
Very unjustly did the priest
Whom Estormi killed lose his life.
The devil has a great talent
For tricking and trapping people.”

And that is just about the end of the story. It has been modelled in typical folktale fashion, following what is often called “the rule of three,” and it has besides exploited what might also be called “the supplement of the fourth.” The rule of three portends an economical balance: readers will probably be most aware of “Goldilocks and the Three Bears,” where the third alternative portends a golden mean which solves a narrative problem. The device is common in all sorts of tales. There are three little pigs. In the classic Greek myth, The Judgement of Paris, the title character has to choose between three different goddesses as to which one is the loveliest. When a magic being is freed from a bottle or other impediment, his rescuer is entitled to three wishes. But there are cases when a folktale cannot end on the resolution of the rule of three, and so brings in a fourth phenomenon, which upsets the balance altogether. I call it the supplement of the fourth.

And in any case, after the fourth monk is killed, and John bewails the injustice of it, this supplement of the fourth overturning the story’s balance, there is still more to come, because after John finishes speaking, the narrator comes in to tell us the moral of the story:

Through these priests,
I would like to teach you
That it is folly to covet
Or woo somebody else’s wife.
The reason for this is very clear.
Do you think that because of any poverty
A decent woman would forget her duties?
No! She would rather let her throat
Be cut with a sharp razor
Than ever to do for money
A thing that would bring shame to her lord.

Note how different are the responses of John and the narrator, John ending with alarm at the evil circulating in the world, the narrator ending with a caution against trying to seduce another man’s wife by paying her money. Meanwhile four monks lie dead.

* * *

For John and the narrator alike the story is an example of something, a higher moral truth which can come either in the form of a maxim (John) or in the form of a caution (the narrator). It is precisely because of its exemplarity that the story is understood to be “instrumental,” which is to say “useful” as well as pleasant. Watch out for the devil! Don’t try to bribe a married woman! But the author of the story, who uncharacteristically identifies himself at the end of the tale, seems disposed rather to point out that morals of this kind are arbitrary and beside the point. For it should be clear to any discerning reader that the story does not prove that the devil is everywhere, or that all wives would resist a bribe for sex, or that bribing a woman for sex is bound to get the perpetrator in trouble. In other words, whatever John or the narrator may think, the story does not mean what they say it means. Both John and the narrator seem to be committing the fallacy of the excluded middle, where a particular becomes a universal without a middle term to tie the universal to the particular. To put it another way, both John and the narrator assume an either/or state of affairs, although the story itself puts forward any number of alternatives. For example, instead of blaming the devil, John might have blamed himself for entrusting serious work to a fool. Instead of cautioning men to stay away from other men’s wives, the narrator might have said that the would-be adulterer should proceed by steps and make sure of a woman’s affection before he tries to seduce her with gifts.

The question may then arise as to the author’s intention. Why did he tell this story? One obvious answer is that he told it in order to mock exemplary fables. Stories like this, he implies, seem to have a moral but really don’t. There is nothing instrumental about them. Or he may be anticipating the Renaissance tradition which begins with Boccaccio’s Decameron, where time and again it is shown that different people may interpret the same story differently, given their own prejudices, needs and struggles. Or again, he may be anticipating the naturalistic tradition of storytelling, which becomes especially prominent in the nineteenth century, according to which the purpose of fiction is to illuminate the (ugly) truth of how the world really works. The story of John and Yfame casts light on a world where poverty can lead people to extremes, where religious people may ride roughshod over their vows to austerity, continence and charity, where violence and vengeance may seem to satisfy our inner needs coupled with our sense of justice … and so forth.

I was myself raised in this last tradition. I was taught, in high school, that Pride and Prejudice was about the vexing rise of materialism in early nineteenth-century Britain, that Macbeth was about the dangers of ambition, and that “Bartleby the Scrivener” was about the power of irrationality in the face of modern-day bureaucratic capitalism. Or at least, that is what I remember having been taught. But what if the stories are not about these things? What if they have no “use,” no propositional and ethical uptake? Or what if, as in the case of the fabliau, the instrumentality of fiction indefinitely recedes into the abyss of excluded middles?

I don’t have a ready answer to the question, but I hope I have at least pointed out that the problem of the instrumentality of fiction is nothing new; it was already a vital concern among storytellers in the thirteenth century.



[1] The French Fabliau B.N. MS. 837, ed. and trans. Raymond Eichmann and John Duval, Two Volumes, Vol. I (Abingdon: Routledge, 2018). Kindle Edition, position 951-1472.




Robert Appelbaum, Professor Emeritus of English Literature, Uppsala University, and Senior Professor in Arts and Communications at Malmö University, is the author, most recently, of The Aesthetics of Violence: Art, Fiction, Drama and Film (2017). He is currently working on a book entitled The Renaissance Discovery of Violence, from Boccaccio to Shakespeare.


James Phelan • Assessing Nonfictional Narratives in Contest: The Rhetoric of Devin Nunes

About the blog: The Instrumental Narratives blog aims to popularize the insights and methods of narrative scholarship and features analyses of instrumental storytelling by high-profile narrative scholars. The analyzed cases deal with uses or abuses of the narrative form, storytelling practices or narrative sense-making in many areas of life: politics, journalism, business, identity work, artistic or literary sphere, activism, and forms of social participation. The blog texts evaluate possible societal risks or benefits of contemporary storytelling, for example through cases from the author’s own national, linguistic, or cultural sphere.

The impeachment inquiry into U.S. President Donald J. Trump provides so much material for those interested in “instrumental narratives and the limits of storytelling” that I can imagine some scholars devoting their whole careers to analyzing it. Too late for me to do that, but not too late for me to zero in on the eight-minute speech Devin Nunes delivered as the Intelligence Committee of the House of Representatives (HIC) ended its hearings on November 21, 2019:

Devin Nunes
Photo: Chip Somodevilla/Getty Images


I was initially dumbfounded by Nunes’s speech.  He says almost nothing about the testimony of the many witnesses who appeared before the HIC over the previous two weeks. He never directly engages with the specific question the HIC was deliberating: did President Trump abuse his power by withholding aid to Ukraine until that country’s s recently elected President, Volodomyr Zelensky, announced an investigation into Trump’s political rival Joe Biden? Epic fail! My negative judgment only deepened when I compared Nunes’s performance to that of his Democratic counterpart, Adam Schiff, who spoke immediately after him: Schiff begins by directly addressing the day’s witnesses and continues by making the case that (a) the collective evidence brought forth at the hearings supports an affirmative answer to the question about Trump’s behavior; (b) the defenses offered by his Republican colleagues are unpersuasive; and (c) the appropriate ethical response to Trump’s unethical behavior is to hold him accountable for it. In this contest of narratives—and narratives in the service of arguments—Schiff seemed to me the clear winner. (See Phelan 2008 for more on the contest of narratives.)

 * * *

As I thought more about Nunes’s speech as a rhetorical performance, however, I began to question whether I had judged it too hastily and missed some key features of its rhetorical logic. In what follows, I will attempt to shed some light on the relationship between the standard rhetorical situation of nonfictional narratives in contest and the decidedly nonstandard way Nunes concluded his participation in the HIC contest. Understanding that relationship can then, I hope, illuminate some larger issues involved in assessing the efficacy of narratives in contest.

I find it helpful to start from the default rhetorical definition of narrative: somebody telling somebody else on some occasion and for some purpose(s) that something happened.  Viewing a contest of narratives from the vantage point of this definition means attending to the ways in which each of the somebodies-who-tell takes some shared common ground—the something-that-happened—and shapes that material in ways that they believe the somebodies-who-listen will find more persuasive than the story told by their competitors. Not surprisingly, the competing tellers will sometimes dispute the something-that-happened because that common ground is so important. The tellers must be responsible to it. It puts constraints on the ways they can plausibly and effectively shape it. Ignoring, distorting, or contradicting what happened weakens any teller’s narrative. Furthermore, when tellers fail to acknowledge any common ground, there can be no genuine contest between their narratives.

* * *

The inquiry played out as an integral part of a larger unfolding contest of narratives about President Trump’s behavior and about what, if anything, Congress should do about it. In September 2019, the Speaker of the House, Nancy Pelosi, had declared that Trump’s actions, as described in a report by an anonymous whistleblower about a phone call involving Trump and Zelensky on July 25th, had made it incumbent on the House to conduct an official impeachment inquiry. The House must act, Pelosi said, because it now had credible evidence that the President had violated his oath of office by making U.S. aid to Ukraine contingent on Zelensky’s agreement to investigate Biden. On the other side, Trump denied any wrongdoing, characterizing the phone call as “perfect,” and many Republicans came to his defense.

During the hearings, the contest of narratives followed a predictable pattern: Democrats attempted to use their questioning of the witnesses to confirm and further develop the whistleblower’s story about Trump’s actions, while the Republicans attempted to use their questioning either to discredit that story or to contend that neither the story nor the testimony indicated that Trump had abused his power. Nunes and other Republicans such as Jim Jordan tried to expand the relevant something-that-happened to include evidence of past corruption in Ukraine and Biden’s possible connections to it through his son Hunter, who was employed by a Ukrainian energy company.  By attempting to expand the common ground, the Republicans tried to advance an alternative narrative about Trump’s behavior: he wasn’t abusing his power but expressing legitimate concerns about Ukraine’s history of corruption.

* * *

As I watched Nunes’s closing speech, I expected him to use the occasion to offer a narrative defending Trump’s behavior that would be as persuasive as he could possibly make it for his multiple audiences: the members of the HIC, the House, and the Senate, as well as those watching the hearings on television, and the members of the media who would represent the speech in various ways for their viewers, readers, and listeners. But the most important member of Nunes’s audience was Trump himself. Trump’s tweeting throughout the hearings continually reminded everyone that he was not only watching closely but also trying to influence the interpretation of the testimony. In appealing to all his audiences, Nunes needed, I thought, to contest the Democrats’ Narrative about the Abuse of Power. Such contestation might proceed by either establishing a Narrative about the Appropriate Use of Power or by arguing that the Narrative about the Abuse of Power is inadequate, or by doing both.

Nunes, however, shows no interest in any of these strategies. Instead, he sets forth a different narrative, and he begins with its abstract: “I have stressed in these hearings that the whistle-blower complaint was a pretext for Donald Trump’s political opponents to do what they’ve been trying to do since he was elected, oust the President from office.” Nunes then sketches this Narrative of Ouster by drawing a time-line upon which he places a wide range of events involving different, often unspecified people, expressing opposition to Trump. The turning point in Nunes’s narrative is that, after the Mueller Report fails to find clear evidence of collusion between Russia and the Trump campaign in the 2016 election, Trump’s opponents need a new pretext for ousting him. They find it in the whistleblower report, which gives rise to what Nunes calls the “Ukraine hoax.” Nunes then draws on his Narrative of Ouster in order to label the impeachment inquiry a “show trial” in which the “verdict was decided even before the trial began.” He rounds off this use of narrative in the service of argument by characterizing the inquiry as a waste of time, by attacking its process as an example of the “tyranny of the majority,” and labeling it a “farce” and a “travesty.” His final move is to wish his colleagues in the House Judiciary Committee well as they seek to defend the idea that the “American people’s vote actually means something.”

Nunes marshals the Narrative of Ouster in the service of an argument from motives. Since the Democrats’ Narrative about Abuse of Power is driven not by their professed concern for the rule of law but by their unrelenting effort to get rid of Trump, their narrative is no more credible than the ones underlying their previous attempts to oust him. The problem, in short, is not Trump but the Democrats.  Although Nunes does have some valid points (the Democrats have shown that they are committed to the position that Trump’s behavior warrants impeachment), his general strategy has several problems. An argument from motives is a version of an ad hominem/feminam argument, and it suffers from the same logical fallacy: it substitutes an assessment of one’s opponent for an assessment of their argument (or, in this case, their narrative). Some additional problems arise from the particular way Nunes constructs his narrative. The two most significant are that (a) Nunes ignores key recalcitrant evidence in the common ground of the relevant historical record, namely, Nancy Pelosi’s strong stance against impeachment prior to the whistleblower’s report; and (b) in tracing his time-line, he treats the disparate actions by a range of agents as parts of a larger collective effort.

I will come back to these secondary problems, but now I want to focus on an extraordinary consequence of Nunes’s performance. By attacking the Democrats’ motives instead of engaging with their Narrative of Abuse of Power, Nunes effectively opts out of the contest of narratives. Not surprisingly, then, if we assess his performance within the frame of the unfolding contest, we’ll find it deficient. Let’s set aside my spontaneous overflow of powerful judgment (“epic fail!”) and more soberly consider the effectiveness of his speech for one segment of his audience. How might it help Republican voters genuinely concerned about the Democrats’ shaping of Trump’s phone call into the Narrative of Abuse of Power? Nunes’s performance wouldn’t help at all, because it fails to provide what such voters need: an alternative narrative about that crucial event. By opting out of the contest of narratives, Nunes also opts out of trying to persuade such listeners. Amazing. But that observation raises the next questions: what is he opting into, and how might that move illuminate his purposes?

* * *

Nunes’s purpose is less about persuasion and more about consolidating Republican opposition to impeachment. In other words, when he opts out of the contest of narratives, he opts in to circling the wagons. Nunes, to be sure, would be happy to hear that some listeners found his speech persuasive, but his main goal is to fashion a negative characterization of the Democrats that those already on his side can endorse. The speech is far less committed to logical reasoning than to political solidarity. For that reason, the logical problems of the speech are beside the point. What ultimately matters is whether Nunes’s fellow Republicans in Congress and in the electorate find the Narrative of Ouster a useful mechanism for attacking the impeachment inquiry itself.

These points about the relations among strategy (argument from motives), consequence (opting out of the contest), and purpose (consolidating opposition to the inquiry itself) become even clearer when we reflect on how Nunes’s performance implicitly appeals to his most important audience, President Trump himself. Throughout the speech, Nunes reinforces and extends allegations Trump has made throughout his time in office (the Mueller investigation was a “witch hunt”), and he incorporates more recent talking points (impeachment is about overturning the results of the election). More generally, in constructing the Narrative of Ouster, Nunes spends his eight minutes deploying two of Trump’s main strategies in his numerous tweets of 280 characters or less. (1) Deny the existence of any common ground upon which to conduct a contest of narratives (the inquiry is a “hoax”). (2) Attack those on the other side committed to the contest (“Do-Nothing Democrats,” “Crazy Nancy,” “Shifty Schiff”). Interestingly, a third Trumpian rhetorical strategy is conspicuous by its absence: the explicit assertion of innocence (the phone call was “perfect.”). If Nunes were to deploy that strategy, he would run the risk of falling back into the contest of narratives with its acknowledgment of common ground in Trump’s phone call. Far better to stay out of that contest and let Trump himself continue to proclaim his innocence.

In this account, then, Nunes opts out of the contest of narratives because he cares far less about Trump’s behavior toward Ukraine than about keeping Trump in office. Judged on its own terms, the performance, rather than being a failure, is quite effective:  impugning the motives of the Democrats fits his purpose of reinforcing Republican opposition to impeachment. Furthermore, subsequent events have indicated that Nunes achieved his purpose. Other Republicans have found the speech a useful jumping off point for their own attacks on the Democrats and the impeachment inquiry. When the inquiry moved to the House Judiciary Committee in December 2019, Republican members often invoked the Narrative of Ouster, and they repeated Nunes’s concluding allegation that the impeachment was an effort to undo the results of the 2016 election. Ultimately, every Republican member of the House voted against the articles of impeachment. As of this writing, the Republicans remain united in their support for Trump, and his acquittal by the Senate appears inevitable.

* * *

If this analysis passes the Horseshoes Test (that is, is close enough to count for something), then it suggests a few larger points about the efficacy of nonfictional narrative and about judging that efficacy.

  1. The efficacy of any nonfictional narrative is to some degree context-dependent, and contexts are frequently shot through with power differentials. Nunes has the luxury of opting out of the contest of narratives because the Republican majority in the Senate gives the Party the ultimate power about impeachment. Within its own terms, his Narrative of Ouster doesn’t need to be logically sound and scrupulous about adhering to the historical record. It just needs to be plausible enough to reinforce the beliefs of those already on his side.
  2. The efficacy of nonfictional narrative is not wholly context-dependent. Ignoring the common ground in a contest inevitably weakens the force of one’s narrative. The context and its associated power differentials do not eliminate the problems in Nunes’s Narrative: its failure to address the central question of the impeachment inquiry, and its significant logical flaws.

Thus, in assessing efficacy,

  1. We should make what I’ll call a step one judgment in relation to the terms the teller sets for the narrative. When I judged Nunes’s performance by terms he was no longer concerned with, I misjudged it. When I viewed it in relation to its own terms, I got a clearer sense of its appeal and its effectiveness.
  2. We should make a step two judgment by assessing the terms the teller sets for the narrative. In Nunes’s case, the U.S. Constitution provides terms within which to assess his. The Constitution calls for an inquiry into the President’s actual behavior. Seen in this light, Nunes’s performance is an ethically deficient act of evasion in the guise of a bold attack. Its logical and ethical problems bleed into each other. It’s hard to imagine that history will be kind to his rhetorical behavior.


Phelan, James. “Narratives in Contest; Or, Another Twist in the Narrative Turn.”  PMLA 123 (2008): 166–75.

head shot.jpg

James Phelan is Distinguished University Professor of English and the Director of Project Narrative at Ohio State University.  He has devoted his research to developing a viable account of narrative as rhetoric. He has written about style in Worlds from Words; about character and narrative progression in Reading People, Reading Plots; about voice, character narration, ethics, and audiences in Narrative as Rhetoric; about the rhetoric and ethics of character narration in Living to Tell about It; and about narrative judgments and progression in Experiencing Fiction.  He has taken up the relationship between literary history and rhetorical analysis in Reading the American Novel, 1920-2010, and he has further extended the conception and consequences of his rhetorical approach in Somebody Telling Somebody Else. In February 2020, he and Matthew Clark will publish Debating Rhetorical Narratology: On the Synthetic, Mimetic, and Thematic Aspects of Narrative.  Since 1992, Phelan has been editor of Narrative, the journal of the International Society for the Study of Narrative.  Since 1993, he has been a co-editor of the Theory and Interpretation of Narrative Series at Ohio State University Press.


Peter Lamarque • Narrative and Emotion: On Not Getting Too Carried Away

About the blog: The Instrumental Narratives blog aims to popularize the insights and methods of narrative scholarship and features analyses of instrumental storytelling by high-profile narrative scholars. The analyzed cases deal with uses or abuses of the narrative form, storytelling practices or narrative sense-making in many areas of life: politics, journalism, business, identity work, artistic or literary sphere, activism, and forms of social participation. The blog texts evaluate possible societal risks or benefits of contemporary storytelling, for example through cases from the author’s own national, linguistic, or cultural sphere.


One of the pleasures, undoubtedly, of a good novel or short story is experiencing the emotions it arouses.  Sometimes these are positive, making us cheerful, upbeat, hopeful, or amused, but sometimes we are disturbed, shocked, anxious, or deflated. Still, as David Hume recognised, even these latter can afford a kind of pleasure.

My question is this: how important are such emotions among instrumental values of narrative? My answer: not as important as is sometimes claimed.


I will focus on a couple of examples: two emotion-inducing narratives.  In both cases the narrative passages are from well-known novels, both appearing towards the end of the novels.

First, Mrs Bennet’s effusive reaction to news of her daughter Elizabeth’s engagement to Mr Darcy, in Jane Austen’s Pride and Prejudice. Most readers would agree that her happiness, for all its self-serving, materialistic basis, is pleasingly infectious, and amusing:

“Good gracious! Lord bless me! only think! dear me! Mr. Darcy! Who would have thought it! And is it really true? Oh! my sweetest Lizzy! how rich and how great you will be! What pin-money, what jewels, what carriages you will have! Jane’s is nothing to it—nothing at all. I am so pleased—so happy. Such a charming man!—so handsome! so tall!—Oh, my dear Lizzy! pray apologise for my having disliked him so much before. I hope he will overlook it. Dear, dear Lizzy. A house in town! Every thing that is charming! Three daughters married! Ten thousand a year! Oh, Lord! What will become of me. I shall go distracted.” (Ch. 59)

In striking contrast, second, here is the dour and dispiriting end (the final two paragraphs) of Thomas Hardy’s novel Tess of the d’Urbervilles:

Upon the cornice of the tower a tall staff was fixed. Their eyes were riveted on it. A few minutes after the hour had struck something moved slowly up the staff, and extended itself upon the breeze. It was a black flag.

“Justice” was done, and the President of the Immortals, in Æschylean phrase, had ended his sport with Tess. And the d’Urberville knights and dames slept on in their tombs unknowing. The two speechless gazers bent themselves down to the earth, as if in prayer, and remained thus a long time, absolutely motionless: the flag continued to wave silently. As soon as they had strength, they arose, joined hands again, and went on.

The black flag on the prison tower signals that Tess is dead: she has been hanged. The two “speechless gazers” are Tess’s sister Liza-Lu and Angel Clare, Tess’s husband.

The first passage offers a reader delight and amusement, the second a kind of weary, cathartic sadness.

What role do such emotions play in our appreciation of literary narratives? Answer: It depends on what kind of evaluation is at stake and what kind of reader we have in mind.


I will frame the discussion by introducing two distinctions and then addressing, and challenging, two claims about the significance of emotions for narrative. I will end with some polemical claims of my own.

The first distinction is between two kinds of readers.

I will label them, in a rough and ready way, the Ordinary Reader (OR) and the Professional Literary Critic (PLC)

These are roles rather than exclusive classes, so one and the same person could be an OR on one occasion and a PLC on a different occasion, even regarding a single text.

The role categories are not precisely defined.  Roughly, but precise enough for our purposes, an OR is a reader who gives primary focus to character, plot, incident and (fictional) world, someone prepared to get caught up in a story, imagining the events as if real, and taking pleasure from emotional engagement with the fictional goings-on.

A PLC in contrast, perhaps a literary scholar, academic, teacher, or reviewer, is a reader with a different kind of interest, an immersion in detail certainly, but someone who treats the text as a literary and linguistic artefact, keen to understand how it works, what tradition it belongs in (or repudiates), what its aims are, what themes it explores, and how successful it is at an artistic or aesthetic level.

The contrast between these kinds of readers relates to what I have referred to, in The Opacity of Narrative (2014), as “transparent” and “opaque” modes of reading.

The second distinction is between what I will call felt emotion and expressed emotion.

A felt emotion is an actual emotional experience of a reader in response to a narrative or narrative fragment.

A reader of the first passage is likely to share some of Mrs Bennet’s joy, delighted that the trial of love has ended happily, perhaps smiling empathetically at Mrs B’s enthusiasm, albeit perhaps more impressed by the happy outcome than the promise of new-found wealth for Lizzy.

The second passage will not elicit such positive feelings. There is no feel-good experience at this ending. Only depression and a kind of heavy despair. Readers are likely to be emotionally drained to encounter the end of Tess’s tragic life in this lonely and macabre setting.

An expressed emotion is not an actual felt emotion in a reader but a quality of the narrative text itself.

The Mrs B passage expresses joy, a kind of giddy excitement, and an overwhelming delight. These expressed emotions are emergent properties of the writing, supervening on features of the text: the broken sentences, the short staccato phrases, the vocabulary, the tone, the exclamation marks, the implied breathlessness. Also expressed, with authorial irony, is a view (at the heart of Mrs B’s philosophy) about where the true value of marital happiness lies. Delight about the riches and wealth to come is itself an emotion expressed, both by the character and in the text’s sentences.

The passage about Tess is equally expressive but in a completely contrasting manner. Here the expressed emotions are not attributed, at least directly, to any character. The observers are “speechless”. If the joy in the passage about Mrs B emerges from the bright, exuberant exclamations, the solemn gloominess and hopelessness expressed in the Tess passage shows itself in the plainness of the prose, as well as in the external viewpoint. There is little sign of rhetorical flourish or witticism and only an implied emotion in the characters.

But the narrative point of view is unmistakable. The inverted commas round “Justice” mocks the patent injustice manifested and the bitterly sarcastic phrase “President of the Immortals” to refer to a supreme, but implacably cruel, god (seemingly based on Zeus as portrayed in Aeschylus’ tragedy of Prometheus) emphasises the crushing of Tess by forces (fate, the gods) quite beyond her control. The term “sport” reminds us of Gloucester’s anguished cry in King Lear: “As flies to wanton boys are we to th’ gods. / They kill us for their sport”. The final words in the novel “they arose, joined hands again, and went on”, echo the end of Paradise Lost when Adam and Eve leave the Garden of Eden, “They, hand in hand, with wandering steps and slow, / Through Eden took their solitary way”. Such allusions again reinforce the expression of gloom and despair.


Before we seek out some connections between these distinctions, let me introduce two claims sometimes made about emotional responses to fictional narrative, both concerning the instrumental values they might contribute to.

These claims concern a supposed didactic value of such emotions, and a clarificatory value.

First, it is sometimes claimed that the emotions aroused by literary narratives perform a didactic function: they can teach us about emotional response itself, its forms and its appropriateness, and if all goes well can make us more emotionally sensitive, in particular more empathetic.

Jenefer Robinson makes such a claim: “it is only through an emotional experience of a novel that one can genuinely learn from it … the emotional process of engaging with characters and situations in a novel is part of a ‘sentimental education’, an education by the emotions” (Deeper than Reason, 2005, p. 156).

Then, second, it is sometimes claimed that emotions aroused by a literary narrative can also serve to clarify the narrative itself, thus aiding our appreciation of a work. Here, again, is Jenefer Robinson: “our emotional responses are a vital part of understanding a narrative text…:  if I laugh and cry, shiver, tense, and relax in all the appropriate places, then I can be said to have understood the story” (Deeper than Reason, pp. 122–123, italics in original).

The didactic and clarificatory roles of emotion are distinct because one could occur without the other.

I suggest we should be cautious about both claims.

Take the didactic claim: that the arousing of emotions in literary narratives has the capacity to increase emotional sensitivity in readers, in particular empathy. Clearly this is a claim about felt, not expressed, emotions: it involves the actual emotional states of readers. It is also an empirical claim. Do readers who respond with (appropriate) emotions to works of fiction exhibit significantly high degrees of sensitivity or empathy (against a control group)? There is no conclusive evidence one way or the other. How could there be? The terms are far too vague.

What kinds of readers are at issue? If they are those who habitually adopt the PLC role then the field of enquiry is narrowed but anecdotally there is no strong evidence that literary academics or teachers are more empathetic than academics in other areas or members of the public at large. However, ORs are simply too diffuse a group, sociologically, psychologically, or in terms of age, education, interests, or disposition, to promise any kind of coherent empirical result.

Yet it is ORs that seem more suitable experimental subjects. Why? Because felt emotions are likely to play a more prominent role in their responses than in the responses of PLCs. A PLC reader will attend primarily to expressed rather than felt emotions: i.e. emotions grounded in narrative itself. That is not to say that those in a PLC role do not feel emotional responses to the works they study but given their focus of interest they will seek out empathy, for example, not in their own felt reactions but as a quality of the writing they set out to explore.

Does feeling joyful with Mrs B and depressed at Tess’s death tend to make us more empathetic? Surely the explanation is the other way round. Only a reader already possessed of some degree of empathy will respond in these ways. Empathy is a condition for the responses not a consequence of them. There is no reason to believe that empathy is increased by exposure to such passages. So the didactic claim, in this context, is unsupported.

What about the clarificatory claim: that the arousing of emotions in literary narratives can help us understand and appreciate the narratives better?

This is also partially an empirical claim but it is also normative, to the effect that a condition for a proper understanding is having the appropriate emotions. Once again, it is felt, not expressed, emotions that are required. And that suggests that the focus is on ORs more than PLCs.

However, in both our examples a test for understanding the passages is, arguably, the ability to grasp the expressed emotions: to see in the narrative the emotions of excitement and depression. It might be that some readers (ORs) will recognize the expressed emotion on the strength of experiencing the felt emotion. But for an experienced PLC reader such a prompter is not needed. The expressive qualities of the writing are plain to see. In fact it is virtually inconceivable that any (serious) reader should wonder, for example, if it is joy or sadness expressed in the lines about Tess and only able to confirm the right answer by appealing to an emotion felt. Some degree of understanding must be a condition for an emotional response, even though misunderstanding can lead to inappropriate responses. So the normative claim is deeply suspect.

This suggests that there is no indispensable role for felt emotions in understanding narrative. What a proper understanding requires is the recognition of expressed emotions. This is part of the training of a PLC and it is rooted in a sensitivity to language as much as a sensitivity to feelings. Needless to say, someone with no emotional capacity is unlikely to have a deep grasp of expressiveness in language.


In conclusion, our two passages offer little support for either the didactic view of emotional response or the clarificatory view. Admittedly, much more could be said about both views. What the passages do suggest is that the felt emotional responses––joy or depression––in ORs arise when readers become attached to characters, with the pleasures that this attachment entails. Only because the lives of Lizzy Bennet and Tess in some sense matter to us do we react as we do. But these empathetic feelings do not in themselves either explain or justify a heightened literary value in the works. In fact, such responses are more characteristic of, and often more pronounced in, many forms of fiction we rarely consider in terms of literary value, romances or TV soap operas, for instance.

PLC readers, in contrast, will not give special weight to felt emotions, concentrating on the expressive qualities, among other kinds, exhibited in the novels. Thus for an informed opinion of literary and aesthetic value the PLC standpoint is likely to be more securely grounded than that resting on felt emotions alone.

It is one thing to feel pity and fear watching a tragedy, another to discern the pitiable and fearful in the tragedy itself. We do not do justice to narrative if we overemphasise the former at the expense of the latter.



Austen, Jane. Pride and Prejudice. 1813.

Hardy, Thomas. Tess of the d’Urbervilles. 1891.

Lamarque, Peter. The Opacity of Narrative. London: Rowman and Littlefield. 2014.

Robinson, Jenefer. Deeper than Reason: Emotion and its Role in Literature, Music, and Art. Oxford: Oxford University Press. 2005.



Peter LamarqPL image (standing).jpgue is Professor of Philosophy at the University of York and was Editor of the British Journal of Aesthetics (1994–2008). He works principally in aesthetics and the philosophy of literature. His books include Truth, Fiction, and Literature (Clarendon Press, 1994, with Stein Haugom Olsen); Fictional Points of View (Cornell UP, 1996); The Philosophy of Literature (Blackwell, 2008); Work and Object: Explorations in the Metaphysics of Art (Oxford UP, 2010); The Opacity of Narrative (Rowman & Littlefield International, 2014); and The Uselessness of Art: Essays in the Philosophy of Art and Literature (Sussex Academic Press, 2019). A revised and enlarged second edition of his co-edited volume (with Stein Haugom Olsen) Aesthetics and the Philosophy of Art: The Analytic Tradition: An Anthology (Wiley-Blackwell) was published in 2019. His book Work and Object was awarded the American Society for Aesthetics Outstanding Monograph Prize (2011) and the Premio Internazionale d’Estetika (2018) from the Italian Society of Aesthetics.


Marie-Laure Ryan • “Sharpiegate” as a Network of Stories

About the blog: The Instrumental Narratives blog aims to popularize the insights and methods of narrative scholarship and features analyses of instrumental storytelling by high-profile narrative scholars. The analyzed cases deal with uses or abuses of the narrative form, storytelling practices or narrative sense-making in many areas of life: politics, journalism, business, identity work, artistic or literary sphere, activism, and forms of social participation. The blog texts evaluate possible societal risks or benefits of contemporary storytelling, for example through cases from the author’s own national, linguistic, or cultural sphere.

It was my original intent to open this blog with some narratological considerations about how the current storytelling movement favors “experientiality” and “what it is like” -based conceptions of narrative (Fludernik; Herman) over plot-based conceptions that emphasize causality, agency, problem-solving and the explanatory value of narrative. But then I came across an example that caught my imagination as it was unfolding live in the media, and that contradicts the above statement: it is all about plotting (by the characters), and the representation of emotions does not come into play, though of course one can argue that the behavior of the characters was to some extent emotionally motivated. This narrative is known as Sharpiegate (following the post-Watergate usage of “gate” as a suffix meaning “political scandal”), and it unfolded from September 1 to September 11, 2019 as what Françoise Revaz has called a “feuilleton médiatique” — a media serial. Here are the installments (a listing of the sources can be found in the Wikipedia entry):


 (Photo: Bill O’Leary/The Washington Post)

  1. Trump’s tweet

September 1, as hurricane Dorian is approaching the U.S., but before it hits the Bahamas, President Trump (who had cancelled a trip to Poland to monitor the hurricane) issues the following tweet: “In addition to Florida – South Carolina, North Carolina, Georgia, and Alabama, will most likely be hit (much) harder than anticipated.”

  1. The Birmingham Weather disclaimer

September 1, 20 minutes after Trump’s tweet (1): The Birmingham Weather forecast tweets a disclaimer: “Alabama will NOT see any impacts from #Dorian. We repeat, no impacts from Hurricane #Dorian will be felt across Alabama. The system will remain too far east.”

  1. Trump’s corrective tweeting

September 2 to 5: At least partly in response to the Birmingham Weather disclaimer (2), Trump in multiple tweets insists that he was right—Alabama WAS in the possible path of the storm. On September 4, during an Oval Office briefing on the status of Hurricane Dorian, Trump produces a doctored National Hurricane Service map [from August 29] to prove his point. On the map, the cone of possible paths of Dorian is extended with a black sharpie pen line to include a small part of Alabama. When asked about the sharpie doctoring, Trump says he does not know anything about it. The next day, Trump tweets: “Alabama was going to be hit or grazed, and then Hurricane Dorian took a different path (up along the East Coast). The Fake News knows this very well. That’s why they’re the Fake News!”

  1. NOAA unsigned statement

September 6: Officials at NOAA (National Oceanic and Atmospheric Administration, which is part of the Department of Commerce) release an unsigned statement about the Alabama matter. The statement referred to the Birmingham Weather disclaimer (2) as “inconsistent with probabilities from the best forecast products available at the time.”

  1. NOAA protest of earlier statement

September 6, afternoon: Scientists at NOAA protest NOAA unsigned statement (4). Neil Jacobs, the acting head of NOAA, sends an all-staff email showing support of the Birmingham weather forecasters against the unsigned statement (4). On Monday, September 9, Louis Ucellini, director of the National Weather Service, says in a public speech: “Let me be clear: The Birmingham office did this to stop public panic, to ensure public safety – the same goal as all the National Weather Service offices were working toward at that time.”

  1. NYT NOAA firings report

September 9: the New York Times reports that Secretary of Commerce Wilbur Ross threatened to fire managers at NOAA over the Birmingham Weather disclaimer (2).

  1. NOAA firings denial

September 9: The Commerce Department denies the NYT NOAA firings report (6).

  1. The Post report of Trump’s involvement

September 11: The Washington Post reports that Trump personally pressed an aide to ask NOAA to produce the NOAA unsigned statement (4).

  1. Trump’s denial of involvement

September 11: Trump tells reporters (referring to The Post report of Trump’s involvement (8)): “No, I never did that. I never did that. It’s a hoax by the media. That’s just fake news.”

  1. House Committee investigation

September 11: Members of the House Committee on Science, Space and Technology, which has jurisdiction over NOAA, launch an investigation into the Commerce Department’s interactions with NOAA regarding Dorian. They are specifically interested in the two directives from NOAA to the National Weather Service and the issuance of the September 6 statement.

To consider this list a narrative, one must extend the conception of narrative beyond James Phelan’s definition: “Somebody telling somebody else that something happened.” Sharpiegate differs from this definition because it is not told retrospectively, from the perspective of the end, but rather develops in real time, which means that as the installments reach the reader, the action is ongoing and has not yet reached a conclusion. Every new day can potentially bring a new twist in the plot, and different people will follow the story for a different amount of time. Another difference from Phelan’s formula is that there is no narrator in the outline presented above, just a collection of events that I picked from the stream of news that continually flows out of the media. It is my interpretive act that turns this list of data into a coherent story; the story, therefore, is a mental construct. If one conceives narrative as an artifact deliberately created to convey a story, one could argue that in this case there is strictly speaking no narrative at all, there is just a story suggested by events that happen in the world. In standard narrative communication, stories are constructed by readers out of data presented by the discourse of a narrator, but here there is no encompassing, top-down narrative consciousness that organizes all the documents for the reader to interpret. The organization of the data into distinct entries is strictly mine; I could have bundled information differently (for instance by distinguishing the Trump tweets of September 2 and 3 from the sharpie incident of September 4), or made a different selection (omitting (10) altogether, since it will be several weeks before it produces a decision, if at all). But my selection is not arbitrary; if the various entries suggest a coherent story, it is because each of them refers and reacts to a previous entry, thus forming a neat causal chain of moves and countermoves:

2 (The Birmingham Weather disclaimer) corrects 1 (Trump’s tweet);

3 (Trump’s corrective tweeting) and 4 (NOAA unsigned statement) correct 2;

6 (NYT NOAA firings report) and 8 (The Post report of Trump’s involvement) correct 4;

7 (NOAA firings denial) corrects 6; and 9 (Trump’s denial of involvement) corrects 8.

10 (House Committee investigation) proposes to review the whole sequence, in order to decide whether the denied facts or the denials are true.

While the story I extract out of the data I have selected is an interpretation, the data itself consists of objectively existing, verifiable documents: it is an undeniable fact that Trump tweeted (1) and (3), that he displayed a map that had been altered with a sharpie pen to prove his point, that the officials at NOAA discredited the report of the weather forecasters of Birmingham, etc. These facts concern communicative acts that actually occurred. But the content of communicative acts does not necessarily correspond to facts: the proposition “John said that p” can be true even if p is false. It is indeed the truth of its component speech acts that constitutes the topic, the aboutness of the story. In one possible storification of the data, the proposition of Trump’s tweet (1), that Alabama would be hit by Dorian, was plausible at the time it was released, and Sharpiegate was not really a scandal, as the suffix gate suggests, but an instance of presidential harassment by the “fake news industry,” of which the New York Times and the Washington Post are prominent representatives.

In this interpretation (let’s call it scenario A):

Trump’s tweet (1) is true (or rather, possible at the time it was communicated),

The Birmingham Weather disclaimer (2) is false,

Trump’s corrective tweeting (3) is true;

NOAA unsigned statement (4) is justified,

NOAA protest of earlier statement (5) is not,

NYT NOAA firings report (6) is fake news,

NOAA firings denial (7) is true,

The Post report of Trump’s involvement (8) is fake news,

Trump’s denial of involvement (9) is true.

(These values concern the embedded content of the communicative acts.) In the other interpretation (scenario B, which reflects more closely Trump’s reputation as a pathological liar and is by far the most dominant in the mediascape):

The content of Trump’s tweet (1) is false,

The Birmingham Weather disclaimer (2) is true,

Trump’s corrective tweeting (3) is false,

NOAA unsigned statement (4) is a cover-up,

NOAA protest of earlier statement (5) is a justified reaction to (4),

NYT NOAA firings report (6) is true,

NOAA firings denial (7) is a lie,

The Post report of Trump’s involvement (8) is true,

Trump’s denial of involvement (9) is a lie.

The moral of this version turns out to be something like errare humanum est, perseverare diabolicum: by trying to cover up an insignificant mistake, Trump digs himself deeper and deeper into trouble and exposes the corruption of his administration and the flaws of his character. In so doing he acts like a child who tells big lies to hide a small misbehavior and gets into serious trouble, while punishment could have been easily avoided by admitting the mistake right away.

The storification of the data involves not only the assignment of truth value to the content of its communicative acts, it also requires the attribution of goals and plans to the agents. I have called these goals and plans “virtual embedded narratives” because they consist of causally linked sequence of events and actions (hence narratives) imagined by the characters as future possible developments that may or may not be actualized (hence, virtual). The actions of characters aim at either their realization or their prevention. The intent of Trump in mentioning Alabama is murky (honest mistake? misunderstanding of how hurricanes develop? slip of the tongue? political motivation?) but all these beliefs are compatible with scenario B, while scenario A is compatible with only one intent: Trump’s justified belief that hurricane Dorian threatens Alabama. From the Birmingham Weather disclaimer (2) on, the embedded narratives of motivations become much more specific, and diverge very sharply according to the two scenarios. It may seem clear that the Birmingham weather people wanted to prevent a panic and needless evacuation, but fanatic partisans of scenario A could argue that (2) was part of vast conspiracy to humiliate the President. Scenarios A and B thus consist of the same facts (those listed in my summary, which carefully avoids interpretation), but they differ on the level of subjective motivations and valuation of the communicated contents.

Is Sharpiegate an instrumental narrative produced to serve a specific goal? Originally not—it’s just a sequence of events that took place without top-down design. However, once it happened and reached an end—an end rather than a closure: it just ceased to inspire new moves by its protagonists—Sharpiegate became narrative material that can be told from a retrospective point of view, and that can serve a political agenda. A case in point is the retelling of the story according to scenario B by Rachel Maddow of MSNBC, a left-leaning TV channel passionately dedicated to the defeat of Trump in the 2020 election. By contrast, Sean Hannity of the Trump-supporting Fox news and a personal friend of the President made no mention of the story, thereby demonstrating that disnarration can be as biased and as instrumental as narration.

When a story captures the public’s imagination, it continues to grow, by incorporating new materials, new details, or by inspiring other stories. Sharpiegate is no exception. It was expanded by the Washington Post into an “oral history” titled “This is not just a stupid story, it’s a big story” that includes the reactions and opinions of many peripheral characters, such as NOAA officials, members of the Birmingham weather bureau, hurricane victims and even a Trump supporter. From this expanded version we learn that the Birmingham weather forecasters claim that they did not respond to Trump’s tweet, of which they had no knowledge, but to signs of panic in the Alabama population. Sharpiegate inspired countless opinion pieces and letters to the editor that range from attacks on the media for blowing up a trivial incident to comparison of Trump’s distortion of facts with the policies of the state in George Orwell’s 1984. It became a favorite target of comedians and political cartoonists, who created a profusion of images altered with sharpie lines showing, among other things, that the wall is finished, that Trump is taller than Obama, or that the curve of climate change is declining. But just as popular stories generate new stories, they are also shaped by other, larger stories. If Sharpiegate matters, it is because it is read in the context of Trump’s attitude toward science, of his handling of previous hurricanes, of his war on the media, of his refusal to admit any mistakes, and more generally, in the context of the ongoing drama of his reelection.

Try to google Sharpiegate: there is so much material that you will only be able to scratch the surface. But just as hurricanes flatten everything in their path, Sharpiegate will soon be swept away by the next wave of the Sea of Stories (to borrow Salman Rushdie’s expression), sharing the fate of so many other scandals that seem huge only as long as they ride the crest of the current media wave.


Fludernik, Monika. Towards a ‘natural’ narratology. London: Routledge, 1996.

Herman, David. Basic elements of narrative. Malden, MA: Wiley-Blackwell, 2009.

“Hurricane Dorian: Alabama controversy.” Wikipedia.

Phelan, James. Narrative as rhetoric: techniques, audiences, ethics, ideology. Columbus: Ohio State University Press, 1996.

Revaz, Françoise. Introduction à la narratologie: action et narration. Brussels: de Boeck et Duculot, 2009.

Ryan, Marie-Laure. Possible worlds, artificial intelligence and narrative theory. Bloomington: Indiana University Press, 1991.

“This is not just a stupid story, it’s a big story: an oral history of Sharpiegate.” Washington Post.


A native of Geneva, Switzerland, Marie-Laure Ryan is an independent scholar based in Colorado, working currently in the areas of narrative theory, media theory, and representations of space. She is the author of Possible Worlds, Artificial Intelligence and Narrative Theory (1991), Narrative as Virtual Reality: Immersion and Interactivity in Literature and Electronic Media (2001, 2nd edition 2015), Avatars of Story (2006) and the co-author of Narrating Space/Spatializing Narrative (2016), as well as the editor of several books. Her latest (co-edited with Alice Bell) is Possible Worlds Theory and Contemporary Narratology (Nebraska, 2019). She has been scholar in residence at the University of Colorado, Boulder, and Johannes Gutenberg Fellow at the University of Mainz, Germany. In 2017 she received the lifetime achievement award from the International Society for the Study of Narrative. Her web site is at