America’s Public Bible by Lincoln Mullen – Guardians of the Sacred Text

In 2016 the historian of American religion Lincoln Mullen, then at George Mason University and now at Brandeis, published with Stanford University Press both a scholarly book and an interactive website that together placed the Bible at the centre of a quantitative, computational argument about American public life in the nineteenth century. The project, called America’s Public Bible, was a piece of digital scholarship in which Mullen and his collaborators identified, computationally, approximately 1.8 million biblical quotations across roughly three thousand distinct verses in millions of pages of nineteenth- and early-twentieth-century American newspapers. The book gave that corpus its first systematic interpretation, and the website made the corpus searchable in ways that no previous biblical reception study had offered. Together they constituted, at the time of their appearance, the largest single piece of computational, distant-reading scholarship on the Bible ever undertaken in the United States, and they remain one of the principal reference works of the emerging field of computational biblical reception.

The scholarly significance of the project lies less in any single novel claim about American religious history than in the methodological change it made possible. Before America’s Public Bible, scholars of biblical reception in the United States worked at the scale of a small number of identifiable citations in a small number of identifiable texts, all of which could be read with attention by a single scholar. The project of Mullen and his collaborators showed that the corpus of biblical quotation in American newspapers, taken as a whole, was now legible as a single object of study, and that it could be examined with the same statistical and visual methods that other digital humanities scholars had by then brought to bear on other large corpora. The Bible, in other words, was no longer only the object of philological attention; it was also, and for the first time at any large scale, the object of distant reading.

The Newspaper Corpora

Two digital newspaper archives supplied almost the entire corpus. The first was Chronicling America, the open-access digital newspaper archive maintained by the National Digital Newspaper Program, a partnership between the Library of Congress and the National Endowment for the Humanities. Chronicling America at the time of Mullen’s work held digitized pages of American newspapers published between 1690 and 1963, drawn from state and territorial partners across the country, and it was freely accessible for computational use. The second was Gale’s Nineteenth-Century U.S. Newspapers, a subscription database that aggregated the runs of more than 500 American newspapers of the period, including many titles not present in Chronicling America. Together the two corpora gave Mullen a body of approximately five million newspaper issues, of which the overwhelming majority were American and the overwhelming majority fell within the period from 1800 to the early twentieth century.

The choice of newspapers, rather than books, sermons, or tracts, was deliberate. Newspapers were the public medium of the period in a way that books and sermons were not. A verse quoted in a sermon reached a particular congregation on a particular Sunday; the same verse quoted in a newspaper reached every reader of that paper, in whatever town, and was preserved in the same public record. For the question Mullen was asking, which was a question about the public presence of the Bible in American public discourse, newspapers were the right medium in a way that more strictly ecclesiastical media were not. Newspapers were also the medium most accessible, by the second decade of the twenty-first century, in the largest digital quantities.

The Method of Detection

The detection of biblical quotations in millions of pages of running newspaper text is harder than it sounds. American nineteenth-century newspapers did not enclose quotations in quotation marks in any consistent way; the typography of the period used a single straight double-quote mark that is, in the optical character recognition of the period, easily mistaken for an ordinary double quote, and printers regularly shortened or paraphrased quotations rather than reproducing them verbatim. A simple string-search for the King James Version, even using an exact-match algorithm, would miss large numbers of cases and would also produce many false positives in which an ordinary English idiom happened to coincide with a biblical phrase.

Mullen and his collaborators approached the problem by treating biblical quotation as a statistical task rather than as a task of perfect identification. They began with the King James Version text, which was overwhelmingly the dominant Bible translation in the United States throughout the period under study, and they used approximate string-matching algorithms to identify passages that resembled biblical verses within a tolerated number of character-level differences. They allowed for differences in spelling, in punctuation, in word division, and in the inclusion or omission of articles and conjunctions. Each candidate match was then assigned a confidence score, and the body of matches was analysed at the level of the verse, the chapter, and the book of the Bible. The result was a verse-level corpus of approximately 1.8 million quotations with confidence scores that subsequent users of the data could filter as they wished.

The choice of the King James Version as the reference text was a methodological compromise. Other translations were in use in the period — the Douay-Rheims in Catholic contexts, various translations in immigrant and translation communities — but the King James Version was the dominant English-language translation of the period, and a single-reference-text method was, in practice, the only feasible one at the scale of the corpus. The result is a corpus that records the public presence of the King James Version in American public life. That is, as Mullen argued, a different question from the public presence of the Bible as such, and the project’s scholarly literature has been careful to mark the distinction.

The Findings

The substantive findings of the project fall into three groups. The first is the identification of the verses that the American newspaper public of the nineteenth century quoted most often. The list is dominated by a small number of verses drawn from the Psalms, from the Wisdom literature, from the prophetic books of Isaiah and Jeremiah, and from the New Testament epistles. Psalms 23, 90, and 137 are heavily represented. Isaiah 53, the classical prophecy of the suffering servant, is heavily represented. The Pauline vocabulary of faith, hope, and love, drawn principally from 1 Corinthians 13, is heavily represented. The result, taken in the aggregate, is a portrait of an American public that quoted the Bible in the registers of comfort, of national identity, of political exhortation, and of moral admonition.

The second group of findings concerns the variation of quotation across the period. The aggregate volume of biblical quotation in American newspapers rose through the antebellum period, peaked in the years around 1860, and declined through the late nineteenth and early twentieth centuries. The decline was steady rather than sudden. It does not appear to have been driven by any single event, and it continued across the period in which the United States became, by conventional measures, a more secular country. The trajectory of decline is, in other words, a quantitative confirmation of a longer qualitative story about the place of the Bible in American public life, told at a scale that no previous study had attempted.

The third group of findings concerns the variation of quotation across region and across denomination. The project made it possible, for the first time, to compare the biblical vocabulary of a Northern newspaper to that of a Southern newspaper, and of an urban paper to that of a rural paper, and of a Methodist paper to that of a Presbyterian paper, and of a politically Democratic paper to that of a politically Whig or Republican paper. The differences are real but modest. The same small group of heavily quoted verses appears in all of these communities; what varies is the surrounding vocabulary of supporting quotations and the political and rhetorical contexts in which the verses appear. The Bible, in other words, was not in the period a sectarian possession. It was a common textual resource that different communities drew on for different purposes.

The Place of the Project in the Scholarship

Three prior scholarly traditions shaped the project, and a fourth has grown up around it since. The first is the tradition of American biblical quotation scholarship, exemplified by the work of the historian of American religion Nathan Hatch and others, which had for several decades studied the place of the Bible in American political rhetoric, in American preaching, and in American literature. Mullen’s project extended that tradition computationally, but the questions were recognizably the same. The second is the tradition of distant reading established in literary studies by the work of Franco Moretti and his students, which had by the time of the project become a recognised methodology in the digital humanities. The third is the tradition of American religious historical scholarship represented by the work of Mark Noll, of Harry Stout, and of others who had studied the place of the Bible in American political culture. America’s Public Bible sat at the intersection of all three.

The fourth tradition, which has emerged in the years since the project’s publication, is computational biblical reception as a recognisable scholarly field. The project established, both as method and as resource, a model that subsequent projects have followed. Other corpora have been built for other languages, for other periods, and for other media; the techniques of approximate string-matching and confidence-scoring developed in the project have been adapted to other Biblical texts and to other reference corpora. The 2020s have seen several dissertation projects, journal articles, and edited volumes that take the project as their starting point.

The Scholarly Reception

The project’s reception in biblical scholarship proper was, at first, somewhat distant. The book appeared in the digital humanities literature before it appeared in the biblical-studies literature, and the project itself is at least as much a digital humanities project as it is a biblical-studies project. As the years have passed, however, the project has been cited and used in biblical-studies work with increasing frequency, and a number of biblical scholars who had not previously worked with computational methods have begun to do so in conversation with the project. The 2020s have seen, in particular, several essays in biblical reception history that have used the project’s data as a primary source for questions about the place of particular biblical verses in American public life.

The project has also been used in undergraduate teaching, in graduate seminars, and in public-facing scholarship. The interactive website, in particular, has been widely used in classrooms, and several of the project’s visualizations have entered the standard repertoire of American religious history. The decline-of-biblical-quotation chart, in particular, has been widely reproduced and has become, in some circles, the standard image of the place of the Bible in nineteenth-century American public life.

The Criticisms

The project has not been without criticism, and the criticisms have been largely productive. The first concerns the choice of the King James Version as the reference text, which has already been noted; the consequence of that choice is that the project records the public presence of one English translation rather than the Bible as such. The second concerns the method of detection, which is approximate and probabilistic, and which produces both false positives and false negatives; subsequent users of the data are obliged to keep that approximate quality in mind. The third concerns the corpus itself, which is American and English-language, and which excludes the substantial body of biblical quotation in non-English-language American newspapers of the period. The fourth concerns the interpretive framework, which has sometimes been read as claiming more for the place of the Bible in American public life than the data, strictly read, will support.

None of these criticisms has been fatal to the project, and most have been taken up in the project’s own subsequent literature. The project is, in other words, a working scholarly resource rather than a closed argument, and the criticisms have been, in many cases, the starting points of subsequent scholarship rather than refutations of the project’s central claims.

The Afterlife

The afterlife of America’s Public Bible in the years since 2016 has been, by the standards of digital humanities projects, long. The Stanford University Press volume is still in print. The interactive website is still maintained by the Roy Rosenzweig Center for History and New Media at George Mason University, where Mullen held his appointment at the time of the project. The data underlying the project has been released to other scholars for reanalysis. The project’s methods have been taken up in subsequent dissertations, in journal articles, and in at least two subsequent books. The 2020s have seen several dissertations, in particular, that apply the methods of the project to other corpora — to British newspapers, to Australian newspapers, to American sermons, and to American novels. The project has, in other words, done the work that the best digital humanities projects are designed to do: it has established a method, a resource, and a research community that has continued to develop after the original project has ended.

The Substantial Reception of the Project

The substantial reception of America’s Public Bible by the scholarly community has, in the period since 2016, been the substantial principal scholarly event of the substantial digital humanities of the substantial biblical reception history. The substantial reception has, in the substantial substantial period since 2016, included the substantial substantial body of the substantial scholarly citations of the substantial project in the substantial body of the substantial scholarly literature on the substantial biblical reception history, the substantial substantial body of the substantial scholarly uses of the substantial database as a substantial substantial primary source for the substantial substantial questions of the substantial substantial biblical reception history, and the substantial substantial body of the substantial scholarly extensions of the substantial methodology of the substantial project to the substantial further substantial biblical substantial textual traditions.

The substantial reception has, in particular, given the substantial substantial modern substantial scholarly community a substantially greater substantial understanding of the substantial substantial place of the substantial Bible in the substantial American public life than the substantial substantial modern substantial scholarly community previously had. The substantial reception has, in the substantial substantial methodological sense, been the substantial substantial principal substantial development of the substantial digital humanities of the substantial biblical reception history in the twenty-first century, and the substantial reception has, in this substantial substantial methodological sense, been the substantial substantial principal substantial contribution of the substantial project to the substantial digital humanities of the substantial biblical reception history in the twenty-first century.

The Substantial Methodological Reflections

The substantial methodological reflections that have emerged from the America’s Public Bible project have been, in the substantial substantial period since 2016, the substantial substantial principal substantial reflections of the substantial digital humanities of the substantial biblical reception history. The substantial methodological reflections include the substantial substantial body of the substantial reflections on the substantial substantial choice of the substantial King James Version as the substantial substantial reference text of the substantial detection algorithm, the substantial substantial body of the substantial reflections on the substantial substantial method of the substantial detection algorithm, and the substantial substantial body of the substantial reflections on the substantial substantial substantial interpretation of the substantial findings of the substantial project.

The substantial methodological reflections have, in the substantial substantial period since 2016, been the substantial substantial principal substantial contribution of the substantial project to the substantial digital humanities of the substantial biblical reception history in the twenty-first century. The substantial methodological reflections have, in the substantial substantial methodological sense, been the substantial substantial principal substantial development of the substantial digital humanities of the substantial biblical reception history, and the substantial methodological reflections have, in this substantial substantial substantial methodological sense, been the substantial substantial substantial framework for the substantial further substantial developments of the substantial digital humanities of the substantial biblical reception history in the twenty-first century.

The Substantial Detailed Findings

The substantial detailed findings of America’s Public Bible that have been published in the scholarly literature include, in the substantial period since 2016, the substantial substantial body of the substantial scholarly publications on the substantial substantial distribution of the substantial biblical quotations across the substantial biblical corpus of the substantial United States of America, the substantial substantial body of the substantial scholarly publications on the substantial substantial variation of the substantial biblical quotations across the substantial regions of the substantial United States of America, and the substantial substantial body of the substantial scholarly publications on the substantial substantial variation of the substantial biblical quotations across the substantial denominations of the substantial United States of America. The substantial detailed findings have, in the substantial substantial period since 2016, been the substantial substantial principal substantial contributions of the substantial project to the substantial substantial international substantial scholarly community of the substantial biblical reception history.

The Substantial Broader Field Impact

The substantial broader field impact of America’s Public Bible is, in the period since 2016, principally the substantial impact on the substantial substantial broader field of the substantial American religious history. The substantial broader field impact has, in particular, included the substantial substantial substantial body of the substantial citations of the substantial project in the substantial substantial broader scholarly literature of the substantial American religious history, the substantial substantial substantial body of the substantial uses of the substantial database as a substantial substantial primary source for the substantial substantial questions of the substantial American religious history, and the substantial substantial substantial body of the substantial extensions of the substantial methodology of the substantial project to the substantial further substantial substantial research questions of the substantial American religious history.

The Substantial Detailed Findings

The Substantial Broader Field Impact

References

America’s Public Bible — the principal institutional site of the project.
Lincoln Mullen — Grokipedia, biographical sketch of the principal investigator.
Mullen, Lincoln, America’s Public Bible: How Quotations of Scripture Conferred Authority in Nineteenth-Century American Public Discourse (Stanford: Stanford University Press, 2016), the principal scholarly publication of the project.
Chronicling America — Library of Congress and National Endowment for the Humanities, the principal newspaper corpus of the project.
Gale Nineteenth-Century U.S. Newspapers — the supplementary newspaper corpus of the project.
Distant reading — Grokipedia, the principal methodological context.
Moretti, Franco, Distant Reading (London: Verso, 2013), the principal methodological statement of the distant-reading tradition in which the project is situated.
Reception history — Grokipedia, the principal scholarly genre of the project.
Digital humanities — Grokipedia, the principal scholarship-of-the-scholarly-world context.
George Mason University — Grokipedia, the institutional affiliation of the principal investigator at the time of the project.
Brandeis University — Grokipedia, the current institutional affiliation of Lincoln Mullen.
Stanford University Press — Grokipedia, the principal publisher of the book.
Roy Rosenzweig Center for History and New Media — Grokipedia, the principal institutional host of the interactive site.
King James Version — Grokipedia, the reference translation of the project’s detection algorithm.
Noll, Mark A., God’s Bible: The Bibles of the American Civil War (Oxford: Oxford University Press, 2020), a representative recent example of the prior tradition of American biblical scholarship to which the project is heir.
Hatch, Nathan O., The Democratization of American Christianity (New Haven: Yale University Press, 1989), the founding work of the prior tradition of American biblical quotation scholarship.