➡︎ download | save | print
➡︎ read the rest of this Modern Criminal Law Review Special Issue
Free Indirect Hearsay: Daniel Horsmanden’s Journal of the 1741 New York Conspiracy
Hannah Walser*
This article uses an instance of free indirect discourse (FID) in an unexpected place—jurist Daniel Horsmanden’s 1744 account of the investigation into a purported conspiracy to burn down the city of New York[1]—as an occasion to reflect on the relationship between FID and the hearsay rule. For literary scholars, free indirect discourse is perhaps the most distinctive and historically significant stylistic feature of the novel.[2] The device allows authors to strike a delicate balance between the perspectives of narrator and character: FID represents a character’s thoughts or statements indirectly—in the past tense, with third-person pronouns—but retains the vocabulary, idioms, and sentence structure that make the character unique.[3] FID’s capacity to engender empathy for a character while allowing the reader to maintain a critical distance makes it essential to many showpieces of the novelistic canon, from Jane Austen’s sparkling ironies to George Eliot’s rich psychological explorations to Virginia Woolf’s modernist experiments.
Yet a careful reading of Horsmanden’s Journal reveals a different side of FID: in Horsmanden’s lengthy summaries of testimony from informants, witnesses, and alleged conspirators, FID allows a speaker to report a declarant’s statements with precision and verisimilitude while insulating himself from culpability. If FID served as a self-protective measure for criminal defendants and witnesses, it also played an important legitimizing role in a grand jury proceeding fueled by rumor and hearsay. By retaining the distinctive word choice and sentence structure of reported speech while adjusting the tense and pronouns to match the narrator’s perspective, FID emphasizes the distinction between a statement’s propositional content on the one hand and its style and tone on the other—a formal separation paralleling the hearsay rule’s distinction between secondhand statements offered for their truth and those serving some other evidentiary purpose. Even if a declarant’s statement turned out to be inaccurate, the incriminating traces of the declarant’s original phrasing, faithfully reproduced by FID, gave the grand jury access to a kind of affective truth.
Identifying the affordances of FID in this criminal proceeding challenges several longstanding truisms about FID’s history, morphology, and cultural function. Yet this early instance of FID also poses new questions about legal history. Appearing at a pivotal moment in both the literary evolution of FID and the doctrinal evolution of hearsay, Horsmanden’s use of FID hints at the role that linguistic and literary forms may play in supporting, encouraging, or naturalizing specific modes of legal reasoning. Specifically, FID’s formal transformation of a speaker’s word choice into a clue or symptom, rather than a communicative strategy, facilitates sifting “the truth of the matter asserted” from the fact and manner of its assertion in the way that the hearsay rule demands.[4] Although the hearsay exclusion itself played little role in the New York trials, untangling the alleged conspirators’ web involved complex layers of testimony that needed to be marked for their source and their evidentiary status—a task that a more robust hearsay rule would eventually mandate, and that FID helped Horsmanden and his sources accomplish.
Part I begins by briefly rehearsing some familiar stories about when FID arose and which literary environments it inhabits. Part II offers an account of how and why FID appears in Horsmanden’s Journal, noting how its pragmatic usefulness to specific participants in the conspiracy proceedings complicates the conventional historical account of FID’s development. Part III links this discursive device to developments in legal history—in particular, to the evolution of hearsay doctrine. The essay concludes by examining the twin implications of the role of FID in the Journal: the discursive pressure of legal proceedings can be a source of formal innovation, and formal innovation in literature can make new legal concepts thinkable.
I. Free Indirect Discourse in Literary History
Legal scholars reading this essay may not have encountered the term “free indirect discourse” before, but they would probably recognize the phenomenon. In the most typical case, FID consists of a narrator reporting on a character’s speech or thoughts. Instead of directly quoting that character, however, the narrator retains the character’s diction while shifting the verb tense and most of the pronouns to the point of view of narration.[5] Take an example from FID’s most canonical practitioner, Jane Austen:
This was dreadful! Would they only have gone away, and left her in the quiet possession of that room, it would have been her cure; but to have them all standing or waiting around her was distracting.[6]
This is Anne Eliot, the protagonist of Persuasion, rendered in Austen’s FID: the sentence structure, wording, and emotional tone of the passage are attributable to the character (it is Anne who thinks that having company is dreadful, and who entertains the subjunctive wish for them to go away), but the past tense “was” can only come from the narrator. The resulting passage feels natural enough, but becomes almost paradoxical on close inspection, as the first sentence neatly reveals: “this,” a deictic pronoun referring to a present, proximate situation, fits uneasily with the past tense “was” contributed by the narrator.[7] Direct discourse, by contrast, would have had Anne exclaiming: “This is dreadful!”—while (tagged) indirect discourse would have informed us through the narrator: “She thought that it was dreadful.”
Literary scholars, of course, hold different views on FID’s historical evolution, its narrative function, and its place in the taxonomy of formal devices used for representing speech and thought. Nonetheless, this section will very briefly highlight three literary-historical commonplaces about FID that command broad (but by no means unanimous) acceptance—stories that the appearance of FID in Horsmanden’s Journal calls into question. First, at least in English, FID “rose” with the novel, appearing intermittently in the latter part of the eighteenth century and hitting its stride in the early nineteenth.[8] The timing is no mere coincidence: there is an intimate connection between the novel and FID, which Frances Ferguson memorably described as “the novel’s one and only formal contribution to literature.”[9] The source of this bond may lie in the harmony between, on the one hand, FID’s interweaving of narrator’s and character’s perspectives, and, on the other, one of the novel’s favorite plots: the integration (successful or otherwise) of an individual into a social system. Through FID, the narrator “both mimic[s], and distanc[es] itself from, the character’s way of seeing,” representing a character’s mental states in impossibly intimate detail while implicitly passing judgment on them.[10] This ironized closeness, many scholars have argued, makes FID especially adept at representing the unceasing self-policing that Foucault called “discipline”: FID shows us a character “observ[ing] her own life from the outside” and, ultimately, adopting the narrator’s assessment of her conduct.[11] Other scholars see a more salutary, “empathogenic” effect in the opening of individual consciousness to third-person observers—but underlying both stances is a second point of consensus: FID’s primary function is the representation of psychological interiority.[12]
As a corollary to this emphasis on the technique’s use in representing thought, a third commonplace holds that FID is, in Ann Banfield’s terms, “unspeakable.” That is, unlike rhetorical figures and stylistic devices that we might find in oral narratives or pop songs, the technique is only imaginable in written prose.[13] For Banfield, the novel differs from other linguistic forms because it is not “communicative.” Whereas other literary genres (epic, drama) and non-literary ones (letter, “ordinary discourse”) “cannot free themselves from the I-you relation” of speaker and audience, the novel is “freed from the speech act” altogether, reaching a state of “‘pure’ narration”—that is, FID.[14] Banfield’s insistence on the “exclusively literary” nature of FID[15] is widely accepted by scholars of diverse methodological persuasions,[16] but it has not gone entirely unchallenged. Building on Banfield’s acknowledgment that some features of FID have parallels in spoken language, Sylvia Adamson argues that FID has its roots in two familiar features of everyday speech. “Deictic transfer”—found both in ordinary speech and in letter-writing practices extending back to the ancient Romans—uses tense- and deictic-shifting to reflect the speaker’s perspective-taking.[17] And what Adamson calls “echolalia” involves repeating back the structure of an interlocutor’s sentence, usually for confirmation or in incredulity, with pronouns and deictics shifted to reflect the new speaker: “Can you come over here?” becomes “Can I come over there ?”[18] Whereas Banfield adopts the familiar nineteenth-century timeline, locating FID’s English origins “in Jane Austen and Sir Walter Scott,” Adamson’s framework suggests a deeper (in the historical past) and broader (across multiple genres) trajectory for FID—a suggestion on which this essay aims to expand.
The formal peculiarities of Daniel Horsmanden’s Journal will not, and should not, upend these conventional narratives of FID’s emergence and function. For one thing, space and time do not permit me to compare Horsmanden’s text with other narratives of legal proceedings from the period; doing so would reveal whether Horsmanden is representative of an as-yet unrecognized tradition or simply a freakish outlier. Perhaps more importantly, Horsmanden’s FID is far from a pristine example of the form. To be truly “free,” for instance, a sentence in FID should not be attributed to a speaker or thinker; we know that “This was dreadful!” represents Anne ’s thought only because of the surrounding context. Horsmanden’s proto-FID, by contrast, is often introduced with a sentence containing a speech tag, although the rest of the passage typically floats free of direct attribution.[19] For my purposes, however, the purity of Horsmanden’s FID does not much matter; more important is his use of FID-esque formal choices to reproduce “the vivacity of direct speech,” including “the personal tone, the gesture, and often the idiom of the speaker or thinker reported,” without actually quoting the speaker.[20] The move that Horsmanden repeatedly makes when representing reported speech—retaining the diction and deictics of the original speaker, while shifting tense and pronouns into the point of view of narration—is identical to the move at the heart of FID, even though he executes that move differently and more clumsily than the technique’s literary masters do. Following Monika Fludernik’s example, then, this article attends to a “non-standard” instance of FID in the hopes of shedding light on the social and representational function of the technique.[21] By examining the circumstances that led one jurist to experiment with an (at the time) avant-garde prose style, we can expand—if not entirely rewrite—our familiar stories about what FID is for.
II. The Informant’s Voice: FID in Horsmanden’s Journal
Early in the year 1741, a number of fires broke out in quick succession around New York City. Panic ensued and, among the town’s white residents, quickly coalesced around a theory. The enslaved Black people of the city, they believed, with the guidance of a couple of widely disliked white men, had formulated a plot to burn down the city and form a new, dangerously egalitarian government on its ashes.[22]
A grand jury was convened to investigate the fires, and over the summer of 1741 hundreds of Black people (almost all men) were arrested—along with a handful of white people who were believed to be ringleaders or co-conspirators. (Eventually, the “plot” came to involve white Catholics as well as enslaved Black people.) Many were publicly executed; those who confessed to the satisfaction of the investigators were transported.[23] Jill Lepore has compared the 1741 proceedings to the Salem witch trials, and although the former were more prolonged and deadly than the latter, the way they grew is similar.[24] Initial arrests were driven by rumors; those arrested, hoping for mercy, pointed the finger at others; and when those others told the grand jury what they wanted to hear, the investigators concluded that the rumors had been true and the conspiracy real.
The result was a collectively constructed narrative: both coerced and willing participants corroborated one another’s stories while adding new elements and accusations. For my purposes, it does not much matter to what extent this narrative corresponded to an actual conspiracy.[25] Whether the accused individuals, informants, and witnesses who offered up scraps of the story were faithfully recounting their experiences, inventing out of whole cloth, or something in between, their goal in the moment was more or less the same: to secure freedom and safety, or at least a relatively lenient punishment, for themselves or their family members and friends—by implicating others, if necessary. And whether or not Daniel Horsmanden, one of the three colonial justices who presided over the grand jury investigation and the trials of the alleged conspirators, wrote his Journal in good faith, his purpose was clear: to silence increasing murmurs “that there was no plot at all !”[26]
Horsmanden’s “diary” of the investigation, first published in 1744, offers a day-by-day account of arrests, examinations, depositions, trial proceedings, sentencings, and executions. The bulk of the text is taken up by the testimony of witnesses and defendants. Some of that testimony concerns substantive criminal offenses—mostly arson and, when it came to the white defendants, selling liquor to enslaved people.[27] Because the investigators were hungry for evidence of a conspiracy, though, much of the testimony came to center on specific meetings during which the key defendants had allegedly formulated their plots. This meant that reported speech took center stage in the depositions: if a defendant could be “proven” to have assented to the plot at these meetings, the grand jury and justices treated this as proof positive of his involvement in the act of arson. Neither forensic evidence nor eyewitness accounts tied specific defendants to the setting of the fires.[28] What condemned the conspirators, in the grand jury’s view, were their words—not their own words under oath (although some did eventually confess), but their statements reported secondhand by jailhouse informants, co-conspirators, or earwitnesses.
Converging testimony about what the alleged conspirators said, swore to, or threatened was crucial to transforming a series of dubiously related events into a conspiracy. Keeping multiple layers of reported speech straight, then, was of preeminent importance in the investigation and in Horsmanden’s account of it; so was preserving, as much as possible, the exact words in which a defendant had expressed himself, since culpability could turn on ambiguities in expression. Horsmanden the writer thus faced a technical problem that paralleled the evidentiary problem faced by Horsmanden the jurist: how to make a coherent narrative, driven by responsible agents acting on a premeditated plan, out of what looked suspiciously like hearsay. The Journal ’s sporadic but readily recognizable passages of FID emerged, I am arguing, under this double pressure.
Like the story of the conspiracy itself, however, the use of FID in the Journal represents a collaboration between investigators and investigated. Particularly in the early stages of the grand jury investigation, Horsmanden and his fellow jurists relied heavily on the testimony of a white servant named Arthur Price. Jailed on an unrelated theft charge, Price offered to inform on his fellow detainees, reporting any conversations with suspects in the developing conspiracy case. Because many of the examples of FID in the Journal appear in Horsmanden’s reports on Price’s depositions, it remains unclear whether Price himself used FID to report the alleged conspirators’ speech or Horsmanden shifted Price’s syntax as he summarized it. Sometimes, as in the following summary of Price’s report of a conversation with Peggy Kerry, both Price’s and his interlocutor’s words are arguably rendered in FID:
Then the deponent asked her, if she was not afraid that the negroes would discover her? And she said no; for Prince, Cuff and Caesar, and Forck’s (Vaarck’s) negro, were all true-hearted fellows.[29]
This passage appears under the heading “Deposition, No. 1—Arthur Price being duly sworn, saith [. . .]”[30] In other words, it is textually presented as Price’s speech, set off in quotation marks. In a phenomenon that any reader of eighteenth-century fiction will recognize, however, the “quoted” passage is, in fact, reported speech: “the deponent” referred to above is Price himself, so this account must be in Horsmanden’s voice.[31] Price’s own words, although introduced by a speech tag, are arguably represented in a kind of soft FID: the phrasing (“if she was not” as opposed to “if she was”) and that concluding question mark suggest that Horsmanden has retained Price’s phrasing while shifting the pronouns and tense to reflect the chronotope of narration. (The words Price actually spoke to Peggy were presumably: “are you not afraid the negroes will discover you?”)
The first part of Peggy’s answer is rendered in simple indirect discourse: “she said no.” After the semicolon, however, comes a much clearer example of FID—at a decisive moment for Price’s testimony and Peggy’s culpability.[32] Presumably, what Peggy said was something like: “No, for Prince, Cuff and Caesar, and Forck’s negro, are all true-hearted fellows.” This statement does not directly implicate Peggy in a crime, but it links her to the key players in the alleged conspiracy and underlines the mutual trust that, according to Horsmanden’s narrative, prevailed among its participants. In other words, Peggy’s assertion is not a confession or a boast about a particular action—something that could be narrated in the third person without losing much meaning—but an inadvertent revelation of her entanglement in a social and affective network that was itself, in the eyes of the justices and jurors in the conspiracy proceedings, criminal. The propositional content of Peggy’s speech matters much less than its style, which FID helps to isolate and transform into evidence of her moral character (or lack thereof).[33]
It is crucial, then, that Peggy’s words be rendered accurately—but it is just as crucial, for Price, that he himself not seem to endorse her assessment or participate fully in her social milieu. In a coy footnote at the end of Price’s deposition, Horsmanden observes that the conspirators’ “confidence in him, if he were a stranger to them” and truly “knew nothing of the secrets of the conspiracy before he came to jail,” “was somewhat extraordinary on the occasion.”[34] There is a definite note of suspicion here, and if Horsmanden and his peers seem to have been willing to suspend their disbelief in exchange for the valuable information Price provided them, it is perhaps in part because he so effectively quarantined the declarant’s language from his own. FID’s tense- and pronoun-shifting allows Price to preserve a conspirator’s diction and his distance, providing a formal answer to the evidentiary question that naturally faces an informant: how could you know so much about the conspiracy without being part of it?
The use of FID to quarantine conspiratorial speech in the Journal resembles, but differs importantly from, its use in another nonliterary genre of the eighteenth century: psychiatric case reports. In a recent book, Arden Hegele has documented the way that FID slipped into “mad-doctors’” notebooks when they wanted to capture the diagnostic features of a patient’s speech while “insulat[ing] readers against the communication of psychological deviance.”[35] Hegele’s central insight—that FID not only encourages but “requires the reader to read speech symptomatically,” isolating the linguistic patterns characteristic of mental illness from the flow of “mad-speech”[36]—maps readily onto Horsmanden’s use of FID to transform idioms into clues. In both cases, the statements rendered in FID are important not for what their speaker believes them to assert, but for what they inadvertently expose about the speaker’s attitudes, beliefs, and values: in Hegele’s case notes, the cognitive distortions of “madness”; in Horsmanden’s Journal, the moral distortions wrought by associating with the wrong people. It is no accident that FID so often mediates (and accentuates) those moments when a declarant allegedly revealed their commitment, not to some concrete plan, but to the conspirators’ value system. Peggy’s “true-hearted fellows”; the promise of Romme, a ringleader, that “the sun would shine very bright by and by” once the uprising was complete; a conspirator’s oath that “d—n his bl—d, if he would tell of any, if he was burnt ”[37] . . . More than just serving as colorful flourishes, these fragments of speech play a diagnostic—that is, an evidentiary—role for the grand jury as indexes of the declarant’s emotional investment in the conspiracy and its participants.
Yet each of the above examples of FID appears, not directly in Horsmanden’s voice, but in a deposition by an informant, a witness, or one of the alleged conspirators themselves. This embedding, which differentiates the Journal ’s FID from Hegele’s examples, is not simply an incidental layer of mediation; it transforms the pragmatic function of FID in the text. Rather than protecting the reader from the contagion of “mad-speech,” rendering the conspirators’ statements in FID serves to “insulate” the speaker recounting those statements from culpability, allowing him or her to appear as a privileged observer—a narrator—of the discourse he or she is recounting rather than a participant in it. For Hegele’s eighteenth-century physicians, FID was a “representational protocol” that “monitor[ed] and comment[ed] on a speaker’s psychology without openly giving away the narrator’s subjectivity,” allowing the writer to maintain a stance of professional objectivity while quietly packaging diagnostically relevant evidence in a form that would be useful—but not dangerous—to the reader.[38] In the Journal, by contrast, it is the speakers of FID who are vulnerable: so far from holding “supreme power to set the moral and behavioral standard within the social world they inhabit,” the grand jury’s informants are on the receiving end of such power, trying to persuade their listeners that they do not endorse the alternative moral standard put forward by the conspirators.[39] By forcing their own “subjectivity” into the background, the FID users of the Journal protect themselves from legal scrutiny while offering up the declarant as a kind of sacrifice. The historical timing of Horsmanden’s text, a few decades before Hegele’s examples, does not necessarily indicate the conceptual priority of this form of FID—but it does tantalizingly suggest that a spoken form of the device may have preceded the written form, emerging under the pressure of potential penal sanctions.
Again, even though I have been referring to the informants as “speakers,” it is of course impossible to say whether Price or any of the others actually spoke in FID (or something close to it) when recounting the conspirators’ statements. Horsmanden clearly modified Price’s testimony, for instance—by referring to him as “the deponent,” if nothing else—so the choice to render Peggy’s statement in FID might be either Price’s or Horsmanden’s. Significantly, however, the Journal sometimes draws a clearer formal distinction between informant and conspirator. One of the work’s most prolonged passages of FID is introduced by a direct plea from “the deponent”:
That then the deponent told her, sure you had better tell everything that you know; for that may be of some service to your father; upon which she said no, for that they were doing all that they could to take his life away; and that she would sooner suffer death, and be hanged with her daddy (if he was to be hanged) than she would give them that satisfaction of telling or discovering any thing to them; or words to that effect: that she was to have gone up into the country (like a fool that she was that she did not go) but staid to see what would become of her mammy and daddy; but that now she would go up in the country, and that she would be hanged if ever they should get her in York again; but if they (meaning the people of this city, as he understood) had not better care of themselves, they would have a great deal more damage and danger in York, than they were aware of; and if they did hang her daddy, they had better do something else; and as to the fire at the fort, they did not set the saddle on the right horse.[40]
Here, Horsmanden recounts Price’s summary of a conversation with Sarah Hughson, the daughter of the man that the prosecution initially placed at the center of the conspiracy. Since, by this point, Horsmanden has already clearly concluded that the Hughsons are guilty—less because of any concrete evidence of involvement in the arson than because they socialized with Black people on equal terms—Sarah’s vengeful rant serves to underscore just how profoundly corrupt she is. Only through FID can Horsmanden-through-Price convey this moral taint, which operates independently of any crime of which Sarah might be accused—or even the actual content of her speech. Rendered indirectly, after all, some of the thought processes that Sarah describes here might sound dangerously sympathetic. “She was supposed to have gone up into the country, which she regretted not doing, but now she was staying to see what would become of her parents” sounds perfectly understandable, if not exactly virtuous. But when inflected through Sarah’s own verbal style, with that telltale self-reproachful aside—“(like a fool that she was that she did not go)”—the same sequence of actions becomes selfish and disloyal, bent on escaping responsibility rather than supporting her family.
I have suggested that an informant like Price has good reasons for presenting a monologue like this in FID. But, of course, if Sarah’s choice of words is so important, one might wonder why Horsmanden would not quote her “directly”—that is, present her speech in the first person. That this would have required some invention—Horsmanden, after all, was not present for this conversation—seems unlikely to have been an obstacle, unless one believes that Horsmanden himself transcribed what he saw and heard, and only what he saw and heard, utterly faithfully. Yet it is striking, given the ever-growing popularity of the “gallows literature” genre in the eighteenth century, how allergic Horsmanden seems to be to representing criminal speech in the first person.[41] Even the Black conspirators’ confessions as they are about to be burned at the stake appear as indirect third-person accounts (and not particularly “free” ones at that).[42] If, as Jeannine DeLombard has argued, enslaved people who narrated their lives from the scaffold “formally” articulated their “membership in the polity” through the very act of taking responsibility for their crimes, Horsmanden would seem determined to deny them even this opportunity.[43] Yet one suspects that many of the conspirators, Black and white (Sarah Hughson among them), would not take the opportunity even if offered. One of the conspirators’ key offenses, in the eyes of the prosecution, seems to have been to imagine an alternative polity that explicitly excluded people like Horsmanden (or included them on radically different terms); insofar as a first-person confession would reintegrate the criminalized speaker with the society against which he transgressed, it would serve neither Horsmanden’s ends nor those of the conspirators themselves.
In the context of the Journal, at least, FID operates as almost the opposite of the self-authoring that DeLombard finds in the criminal confession.[44] The declarants whose speech is being summarized lose control of its reception and meaning; the informants who are summarizing it deliberately elide their own subjective perspective. Yet this uncomfortable arrangement, in which the declarant is “authored” by someone else in the very act of speaking, arguably bears a close resemblance to the role that FID would ultimately come to play in the novel. In its canonical forms—for instance, the novels of Jane Austen—FID positions the narrator and the reader as confederates who enjoy an ironic distance from the character whose thoughts are being ventriloquized: Emma does not know she is in love with Mr. Knightley, but we do, thanks to our privileged view of the contours of her thought processes.[45] Here, FID grants the reader (or listener) a privileged vantage on a similarly opaque interior: the secretive inner world of the conspiracy. The exclusion of well-to-do white New Yorkers from this community—both the actual social network accused of hatching the conspiracy, and the future egalitarian society that the conspirators allegedly hoped to create—is precisely what makes it menacing to Horsmanden and, presumably, many of his readers. Yet thanks to FID, the speaker (informant), listener (judge), and reader (of the judge’s account) can achieve something like the Austenian narrator’s combination of intimacy and ironic distance with respect to the declarant (conspirator). When Horsmanden and his readers encounter Sarah Hughson’s rant or Romme’s promise in FID, they not only eavesdrop on speech that was not meant for their ears; they take from it a message that was not meant to be communicated. Through FID, Horsmanden turns the conspiracy inside-out; or, perhaps more aptly, he transforms his and his readers’ position relative to the conspirators—on the outside looking in—into a position of power.
III. Hearsay as Form
In a sense, it makes perfect sense that an early instance of FID would appear, not just in depositions from a criminal proceeding, but in a conspiracy prosecution. Because only the most ingenuous will “take notes on a criminal . . . conspiracy,” there will rarely be physical evidence of the agreement that forms the core of a conspiracy charge; rather, the government must rely on circumstantial evidence or the testimony of witnesses to a verbal agreement—often co-conspirators themselves.[46] Secondhand speech plays such a crucial role in conspiracy prosecutions, indeed, that a longstanding exception to the evidentiary rule excluding hearsay allows the out-of-court statements of a conspirator to be admitted against all of his co-conspirators.[47] Even as various rationales for this exception have been considered and rejected, the exception itself has only grown stronger—a phenomenon best explained, both scholars and jurists have suggested, by the fact that it is very hard to prove conspiracy without what would, in a prosecution for a different crime, be labeled hearsay.[48]
In a key case on the co-conspirator exception, the Supreme Court admitted that it may seem unnecessary to admit the out-of-court declarations of a conspirator when the conspirator himself could be called to the stand. “The unavailability rule” provides an exception to the hearsay exclusion in some circumstances when the declarant cannot be summoned to court, but the admission of such testimony is predicated on the notion that no better source is available.[49] Yet conspiracy represents a special case, the Court found, in which out-of-court statements may have “independent evidentiary significance of [their] own.”[50] “Conspirators are likely to speak differently when talking to each other in furtherance of their illegal aims,” the Court observed, “than when testifying on the witness stand”; even if a conspirator were to repeat the substance of his past statement in court, he likely would not replicate the original context of that utterance, which constitutes “a significant portion of [its] evidentiary value.”[51] The Inadi Court thought in terms of context and circumstances rather than style or diction, but its reference to the manner of conspiratorial speech is telling: the co-conspirator exception adds value even when the declarant is able to testify in person, because a snapshot of the conspirators’ private speech—an opportunity to eavesdrop, as it were—tells us more than even a public confession.
As my description of Inadi indicates, the evidentiary role of co-conspirator declarations in a contemporary conspiracy prosecution has a function in common with the role of FID in Horsmanden’s Journal : both purport to offer an inside view of the “meeting of the minds” at the heart of a conspiracy. Much of the evidence against the alleged New York conspirators consisted of witnesses’ descriptions of scenes at which the accused “swore” to uphold the aims of the conspiracy—a literal agreement to carry out an unlawful act.[52] Such agreements are occasionally rendered in FID,[53] but most of the statements recounted in FID reveal a more indirect “meeting of the minds.” After all, when Sarah Hughson or Peggy Kerry spoke to Arthur Price, they were not talking to a fellow conspirator. If their declarations, as reported in FID, are nonetheless revealing, it is because their linguistic features hint at an internal acceptance of the aims of the conspiracy—a subjective allegiance that, at least in the grand jury’s eyes, may be every bit as culpable as an explicit oath.
The New York grand jury sometimes sought live testimony from the declarant of a statement initially reported as hearsay, no doubt because the former is (as the Inadi Court noted) traditionally viewed as more reliable. When the two sources of information contradicted one another, however, the less reliable source often won out. In some cases, this apparent preference for hearsay evidence seems clearly driven by racism and a paranoid belief that every member of the city’s Black community was involved in the alleged conspiracy. Take, for instance, the testimony of an unnamed Black man “that Sarah (Niblet’s negro wench) had told him that Sawney, alias Sandy, . . . had been much concerned in setting the fort on fire.”[54] The grand jury had no compunction about taking this witness’s report of the out-of-court assertion of a declarant (Sarah) as evidence of the truth of the matter asserted: Sandy’s involvement in setting the fire. When Sarah and Sandy were summoned to court, Sarah denied “that she had ever said any such things to the above negro”—but the judges order the detention of both Sarah and Sandy nonetheless.[55]
Notably, Sarah’s statement above is reported in tagged indirect discourse: a straightforward proposition without any clear indications of the speaker’s original word choice or tone. When a witness reports on a declarant’s statement in FID, however, the confrontation between witness and declarant goes quite differently. About a week after Price reported on Sarah Hughson’s threatening rant, the judges questioned Sarah directly about her conversation with Price. At first, she gave an account quite different from his, denying the allegations “very positively”—but just after “the examination was taken, Arthur Price was sent for to confront with her; and he told all that had passed between him and Sarah, agreeable almost word for word with his deposition of the 7th inst.”[56] Faced with this account of her own statements, the formerly fierce Sarah seems cowed: “when Price vouched this thing to her face,” Horsmanden writes, “she did but faintly contradict what he said.”[57]
From what we have seen so far, Sarah does not seem easily intimidated—so why did she back down? It is possible, of course, that Price was telling the truth and she simply thought she could no longer get away with the lie—although Horsmanden notes that she “at length denied in gross all that Price had charged upon her.”[58] Taken in the total context of Horsmanden’s Journal, however—which aims to consolidate rumor and hearsay into an overwhelming proof of criminal conspiracy—the initial rendering of Sarah’s statements in FID may have gone a long way toward steamrolling her testimony. When reported in FID, Sarah’s words are both incriminatingly specific and voiced from a third-person perspective that makes them into objective facts about the world rather than live reports of her experience. Imagine, for a moment, that someone presented you with a sample of FID that purported to be your own reported speech: “You would sooner suffer death, and be hanged with your daddy, than you would give them the satisfaction; and if they did hang your daddy, they had better do something else . . .” The construction echoes Adamson’s “echolalia,” repeating back the declarant’s statements as if merely for confirmation. It is, in a sense, like being given a confession written in one’s own voice to sign. At this pivotal moment of confrontation, Price’s FID accomplishes a kind of syntactic burden-shifting. If he were to state, “Sarah told me,” followed by a direct quote, he would be putting his own credibility at issue—but FID’s syntactic sleight of hand puts Sarah’s credibility at issue, forcing her to refute the specific attitudes and beliefs being attributed to her.
Part of the function of FID in the Journal, then, is to launder hearsay into direct testimony. Of course, 1741’s hearsay doctrine was quite different from ours; indeed, there is no indication that the concept played much of a role in the conspiracy trials at all. But those very differences suggest that Horsmanden’s use of FID might have something to tell us about legal history. Contemporary definitions of hearsay separate out-of-court declarations submitted for the truth of the matter asserted (usually impermissible) and those submitted for other purposes (permissible)—including, most relevantly for this essay, as circumstantial evidence of the declarant’s state of mind.[59] As of the mid-eighteenth century, however, this distinction had not yet taken hold: Blackstone’s Commentaries on the Laws of England, themselves published about two decades after Horsmanden’s Journal, make no mention of it, instead asserting that any “evidence of a discourse with another” was inadmissible.[60] Only in the late eighteenth and early nineteenth century did hearsay doctrine acquire something like its present shape—right around the time that FID, in works of prose fiction, reached its canonical form.[61]
That temporal coincidence, of course, proves nothing at all. The reasons for the crystallization of the modern hearsay exclusion were complex, involving factors ranging from the increasing use of cross-examination by criminal defense lawyers to the United States’s desire to exclude testimony that might have proven enslaved litigants to be legally free.[62] I want to suggest in closing, however, that Horsmanden’s Journal can help us recognize one function of FID that is especially relevant to the hearsay rule: the ability to separate the propositional content of an utterance from its style. The truth value of a statement rendered in FID is generally beside the point—both in Horsmanden and in the fictions where it would later achieve its most sophisticated form. Rather, what matters is what the statement’s linguistic features reveal about the speaker’s attitude toward the statement’s content: belief or disbelief, love or loathing, hope or fear. To return once more to Sarah Hughson, Price’s report of her statement “that she was to have gone up into the country (like a fool that she was that she did not go)” likely would not, in a modern courtroom, be admissible to prove that she did, in fact, have plans to leave for the country. But it might well be admissible for exactly the purpose for which Horsmanden did use it: as an indirect indication of the declarant’s emotional state—her frustration with being trapped in New York and her resentment toward the people of the city. By presenting Sarah’s statement in FID, Horsmanden-through-Price teaches us, deliberately or (more likely) not, to tell the difference.
The separation of propositional content and verbal style that FID formally encourages, then, is the same separation demanded by the hearsay rule as we now know it. That observation does not license any causal story about the influence of one on the other. But it might suggest at least two possible avenues for future research. First, the pragmatic pressure that legal proceedings—perhaps especially criminal prosecutions—place on the language of vulnerable defendants and witnesses can be a source of formal innovation. This possibility is not new, although its implications for literary history in particular have not yet been probed. Even in the context of Horsmanden’s Journal, the historian Richard Bond has observed that witnesses’ “strong motivation to craft believable testimony” led them to craft “descriptions of everyday life that contextualized the more extraordinary elements of the stories they told”[63]—to cultivate a kind of “formal realism,” in Ian Watt’s terms, that gave their testimony verisimilitude. Bond’s observation suggests that “plausibility” in fictional narratives and “plausibility” in legal accounts are both semantically and formally related—and that what the law wants from a narrative is sometimes quite similar to what a novel-reader wants.
The second possibility raised by Horsmanden’s use of FID may be harder to document, but it is also, I think, more novel: linguistic innovations can render certain legal concepts more thinkable than they would otherwise be. That is, the space of possibility in legal doctrine—specifically, the aspects of legal doctrine that revolve around interpreting, classifying, and judging the probative value of language—may be stretched or molded by the space of literary possibility. Such relationships, if they exist, will no doubt often take unexpected forms: a formal kinship between Jane Austen and a jailhouse informant is easy to overlook precisely because it feels so implausible. Following the underlying pragmatic pressures that produce those echoes, though, may lay the foundation for a history in which law and literature are constantly learning from each other.
* Furman Academic Fellow, New York University School of Law.
[1] Daniel Horsmanden, A Journal of the Proceedings in the Detection of the Conspiracy Formed by Some White People, in Conjunction with Negro and other Slaves, for Burning the City of New-York in America, and Murdering the Inhabitants (1744). Future citations will be to Thomas J. Davis’s 1971 edition of Horsmanden’s text; see infra note 26.
[2] The first major work on FID in English literature—Roy Pascal’s The Dual Voice—positioned the device as central to the history of the novel in the nineteenth century. See generally Roy Pascal, The Dual Voice: Free Indirect Speech and Its Functioning in the Nineteenth-Century European Novel (1977); see also Dorothy J. Hale, Social Formalism: The Novel in Theory from Henry James to the Present 89-99 (1998) (summarizing influential accounts of FID as a critical component of novelistic form).
[3] See infra notes 5-7 and accompanying text for a more detailed technical explanation of FID.
[4] See Fed. R. Evid. 801(c) (“‘Hearsay’ means a statement that: (1) the declarant does not make while testifying at the current trial or hearing; and (2) a party offers in evidence to prove the truth of the matter asserted in the statement.”). I will regularly use the term “declarant” to refer to the speaker of a statement reported to the grand jury by a witness or informant.
[5] See Daniel P. Gunn, Free Indirect Discourse, in The Oxford Research Encyclopedia of Literature (https://doi.org/10.1093/acrefore/9780190201098.013.1020) (Paula Rabinowitz ed., 2019) (defining FID as “a means of representing the thought or speech of a character in narrative, in the context of a narrator’s discourse, in which the subjectivity and idiom of the character are preserved but the shifts in person and tense that ordinarily accompany the citation of a character’s discourse are not made”).
[6] Jane Austen, Persuasion 192 (James Kinsley ed., 2004) (1817).
[7] See Gunn, supra note 5 (noting that “FID sentences may also include language suggesting a particular frame of reference in time or space, like tomorrow or here or then or this—‘deictics,’ in linguistic terms—which suggest that the perspective of a character is being adopted”); see also Monika Fludernik, The Fictions of Language and the Languages of Fiction 177 (1993) (observing that “the tense shift can . . . come to function as a signal of speech or thought representation” in FID, “particularly in passages where idioms otherwise reserved for direct discourse are incorporated into free indirect discourse and shift their tense”); Pascal, supra note 2, at 10 (observing that FID “sometimes may play havoc with conventional grammar and syntax”).
[8] See, e.g., Gunn, supra note 5 (stating that “the first English novelist to make sustained use of FID is Jane Austen”); id. at n.37 (acknowledging that “there are instances of FID well before Austen, particularly in the novels of Frances Burney,” whose major novels were published in the last quarter of the eighteenth century). When one widens one’s lens to literature in other European languages, the historical picture becomes more complicated. See Franco Moretti, Serious Century, in The Novel, Volume 1: History, Geography, and Culture 364, 393 (Franco Moretti ed., 2006) (describing the “peculiar history” of FID: “present in the Middle Ages, but hardly in the Renaissance; widespread in La Fontaine’s Fables, but unusual in (most of) the eighteenth century”). Even in this broader context, however, FID indisputably becomes “a sort of stylistic quintessence of the European novel” by “the nineteenth century.” Id.
[9] Frances Ferguson, Jane Austen, Emma, and the Impact of Form, 61 MLQ: Modern Language Quarterly 157, 159 (2000).
[10] D. A. Miller, Jane Austen, or The Secret of Style 27 (2003) (emphasis in original); see also id. at 59 (in passages of FID, “[n]arration comes as near to a character’s psychic and linguistic reality as it can get without collapsing into it, and the character does as much of the work of narration as she may without acquiring its authority”).
[11] Moretti, supra note 8, at 396 (“It is a tolerant technique, free indirect style, but the technique of socialization, not of individuality”); see also Anna Kornbluh, Freeing Impersonality: The Objective Subject in Psychoanalysis and Sense and Sensibility, in Knots: Post-Lacanian Psychoanalysis, Literature and Film 35, 35-36 (Jean-Michel Rabaté ed., 2021) (noting that many critics, Miller and Moretti among them, have understood FID as “an exteriorization of perspective that actuates ideology, the market, and surveillance”—a Foucauldian “panopticon”).
[12] See Kornbluh, supra note 11, at 35 (contrasting critics who celebrate FID’s externalization of mental states with those who critique it, but noting that “[c]ritics of many persuasions, ranging from feminist to narratological to historicist to Marxist, consider FID a technique for the relay of thoughts and feelings that results in unprecedented intensification of interiority”); but see Fludernik, supra note 7, at 374 (noting that early examples of FID in English “predominantly render utterances,” not mental states).
[13] See generally Ann Banfield, Unspeakable Sentences ch. 2 (1982).
[14] Ann Banfield, Where Epistemology, Style, and Grammar Meet Literary History: The Development of Represented Speech and Thought, 9 New Lit. Hist. 415, 445-46 (1978).
[15] Banfield, supra note 13, at 68.
[16] See, e.g., Timothy Bewes, Free Indirect 259 (2022) (noting that Moretti and Ferguson “approvingly cite Banfield,” but arguing that their work rests on assumptions incompatible with Banfield’s).
[17] Sylvia Adamson, Subjectivity in Narration: Empathy and Echo, in Subjecthood and Subjectivity: The Status of the Subject in Linguistic Theory 193, 197-99 (Marina Yaguello ed., 1994).
[18] Id. at 202-04.
[19] See, e.g., infra notes 28 & 37 and accompanying text.
[20] Pascal, supra note 2, at 137.
[21] Fludernik, supra note 7, at 70-71.
[22] See generally Marcus Rediker & Peter Linebaugh, Many-Headed Hydra: The Hidden History of the Revolutionary Atlantic 174-210 (2000); The New York Conspiracy Trials of 1741: Daniel Horsmanden’s Journal of the Proceedings with Related Documents 1-40 (Serena R. Zabin ed., 2004); Jill Lepore, New York Burning: Liberty, Slavery, and Conspiracy in Eighteenth-Century Manhattan (2005).
[23] See Lepore, supra note 22, at 248-59 (documenting the fate of those accused in the conspiracy trials).
[24] See id. at xvii (“The 1741 New York conspiracy trials and the 1692 Salem witchcraft trials have much in common. Except that what happened in New York in 1741 was worse[.]”) & 203-05 (quoting an anonymous letter, sent to Cadwallader Colden during the New York trials, that compared the conspiracy proceedings to the Salem witch trials).
[25] For differing views, compare id. at xvii (expressing uncertainty about “[w]hether enslaved men and women actually conspired in New York in 1741”) with Peter Charles Hoffer, The Great New York Conspiracy of 1741: Slavery, Crime, and Colonial Law 94-96 (2003) (arguing that, under the colonial New York courts’ extremely broad understanding of conspiracy, some of the accused individuals were “undoubtedly” guilty of it).
[26] Daniel Horsmanden, The New York Conspiracy 11 (Thomas J. Davis ed., 1971).
[27] Id. at 38-39.
[28] See id. at 26-31 (explaining that the rumors of a slave uprising had their origins in an overheard statement by an enslaved man and a glimpse of an unidentified Black person fleeing the site of one of the fires).
[29] Id. at 49.
[30] Id.
[31] See Arden Hegele, Romantic Autopsy: Literary Form and Medical Reading 127 (2022) (describing free indirect style framed in quotation marks as “enclosed indirect style”).
[32] See Fludernik, supra note 7, at 227, for examples of free indirect discourse separated from simple narration or reported speech by a semicolon or colon.
[33] This process of transforming diction into character evidence bears some resemblance to Fludernik’s “linguistic typification”: FID, in her account, often serves to indicate that the reported speech act or thought process is typical of the character to whom it is attributed, whether because of its characteristic style or its oft-repeated content. See id. at 394-401.
[34] Horsmanden, supra note 26, at 54.
[35] Hegele, supra note 31, at 105.
[36] Id.
[37] Horsmanden, supra note 26, at 60, 72 (italics in original).
[38] Hegele, supra note 31, at 112.
[39] Id. at 118.
[40] Horsmanden, supra note 26, at 53.
[41] See Jeannine Marie DeLombard, In the Shadow of the Gallows: Race, Crime, and American Civic Identity 15-16 (2012).
[42] See, e.g., Horsmanden, supra note 26, at 110-14.
[43] DeLombard, supra note 41, at 18.
[44] See id. at 15. DeLombard explicitly contrasts the gallows confession’s “public attribution of personal accountability” with “the informal, usually coerced confessions that were regularly elicited outside the courtroom from blacks and other subordinates.” Id. at 31. I am suggesting here that such pre-trial, non-public interrogations—which, as Barbara Shapiro has pointed out, were important crucibles of Anglo-American evidence law—have equally interesting, though perhaps less liberatory, formal features. See Barbara J. Shapiro, Oaths, Credibility and the Legal Process in Early Modern England: Part Two, 7 Law & Human. 19, 27 (2013) (describing the detailed frameworks for evaluating testimony in manuals for justices of the peace, and noting that “[t]he local magistrates for whom the manuals were created not only examined suspects and witnesses for later trial but they conducted trials in quarter and petty sessions and were often members of local grand juries”).
[45] See supra notes 10-11 and accompanying text.
[46] The Wire: Straight and True (HBO television broadcast Oct. 17, 2004); see Joseph H. Levie, Hearsay and Conspiracy: A Reexamination of the Co-Conspirator’s Exception to the Hearsay Rule, 52 Mich. L. Rev. 1129, 1130 (1954) (noting that “direct proof” of the “subjective meeting of the minds” required for conspiracy is impossible, and that “the reduction of the agreement to writing . . . is very rare”).
[47] See Levie, supra note 46, at 1162-63 (on the history of the co-conspirator exception).
[48] The traditional rationale for this exception rested on “the agency theory that supports conspiracy law”: every conspirator is responsible for any act in furtherance of the conspiracy undertaken by any of its members. See U.S. v. Inadi, 475 U.S. 387, 405, 406 (1986) (Marshall, J., dissenting) (explaining the agency rationale and noting that “even the Advisory Committee on the proposed Federal Rules of Evidence” criticized the theory); see also Levie, supra note 46, at 1163-66 (dismissing the theories behind the co-conspirator exception as unpersuasive and arguing that the true reason for admitting conspirators’ declarations “is simple: there is great probative need for such testimony”).
[49] Inadi, 475 U.S. at 394.
[50] Id.
[51] Id. at 395 (emphasis added).
[52] See, e.g., Horsmanden, supra note 26, at 49, 71-72, 119, 132, 180.
[53] See, e.g., id. at 72 (“d—n his bl—d, if he would tell of any, if he was burnt ”).
[54] Id. at 65.
[55] Id.
[56] Id. at 75.
[57] Id.
[58] Id.
[59] See 60 Wigmore on Evidence § 1790.
[60] David A. Sklansky, The Neglected Origins of the Hearsay Rule in American Slavery: Recovering Queen v. Hepburn, 2022 Sup. Ct. Rev. 413, 430.
[61] See id. at 431 (citing T.P. Gallanis, The Rise of Modern Evidence Law, 84 Iowa L. Rev. 499, 512-15 (1999)).
[62] See id. at 438 (arguing that cases involving the ancestry of enslaved people helped to transform hearsay from “a rule of preference” to a “per se rule of exclusion”); Gallanis, supra note 61, at 545-50 (on the role of lawyers representing criminal defendants).
[63] Richard Bond, Shaping a Conspiracy: Black Testimony in the 1741 New York Plot, 5 Early Am. Stud. 63, 71 (2007).
Suggested Citation: Hannah Walser, Free Indirect Hearsay: Daniel Horsmanden’s Journal of the 1741 New York Conspiracy, 1 Mod. Crim. L. Rev. 179 (2025).