12

Re: Search and Close Reading

Michael Hancher

Large digital libraries afford new access to information at the macro scale as well as improved access at the micro scale. The micro scale is familiar enough: the specific datum sought for and found, or not, in an archive of whatever size, large or small: the stuff of philological research. The macro scale is new, and newly revelatory; and being new it preoccupies the stories told in Debates in the Digital Humanities (Gold). Everyone acknowledges that data mining and data mapping are changing the scholarly landscape. But few mention that large digital libraries also afford a greatly improved access to minute particulars—thanks not to data mining but to datum mining. Such particulars, when discovered, can loom large in the foreground of that landscape. Keyword search, now ubiquitous, is taken for granted, like flying in an airplane. And yet, like the airplane, it has been a game-changer.

Anyone of a certain age who stops to think about this can come up with examples of the change. I recall one instance, some seven or eight years ago, when a student alerted me to something he had found in Google Books that had eluded my careful, albeit pre-Google, research several years earlier. I had published an article about a nineteenth-century British broadside ballad, “The March of Intellect in the Butchering Line,” which expressed a butcher’s indignation at the snobbery that his wife had contracted from reading too many “improving” penny magazines. I had (too) conservatively estimated a date for this ballad of about 1850 (Hancher, “From Street Ballad,” 98). When I prepared to discuss it in a course about the history of literacy, I referred students to a facsimile that the Bodleian Library had published online. However, this student found the type too small to read (he overlooked the magnification option provided by the Bodleian website) and reflexively turned to Google to find a legible substitute. Google led him to a facsimile of the text of the poem as it was published in Thomas Hudson’s Comic Songs (London, 1818)—more than three decades before my faulty estimate. Thanks to the student and Google Books, I learned that the song was “Sung by Mr. Sloman at the Coburg Theater” to the tune of “The Irish Washerwoman”: information that belonged in my article, but never made it there. Furthermore, Google Books, especially Google Books Advanced Book Search, and also YouTube, can readily elucidate the singer, the theater, and the song.

More recently, some students and I examined the first volume of All the Year Round, the weekly magazine that Charles Dickens edited from 1859 to 1870. One drew attention to an anonymous poem, “Spinners and Weavers,” in which four overworked women, like “Furies,” praise the utility of their nocturnal labor, which results in candle wicks to brighten the eyes of the nobility, clothing for peasants, rugs for the nobility, and a noose for “a felon’s neck.” The identity of the author, Bryan Waller Procter, can be determined in the traditional way by consulting an edition of Dickens’s correspondence (Letters 9:40). But it takes a search in Google Books on one or another phrase in the poem to discover that Procter had published the same poem years earlier, as “The Flax Spinners,” in Hood’s Magazine and Comic Miscellany (1844)—signed in the table of contents with his pseudonym “Barry Cornwall.” Hood and Procter were friends, and Proctor’s “The Flax Spinners” paid a kind of homage by imitation to “The Song of the Shirt,” the labor-reformist poem that Hood had published in Punch the previous year, which made his literary reputation. Dickens was a friend of both poets; did he know that he was reprinting an old poem when he published “Weavers and Spinners”? Unlike us, he lacked access to Google Books.

Another example: wrapping up an article about a piracy of A Christmas Carol, which I had started some years earlier, I found that many of the obscure literary quotations that ornamented the margins of the pages of that two-penny pamphlet, framing the main text, finally shed their obscurity under the revealing scrutiny of Google’s search engine (Hancher, “Grafting”).

I can cite two other examples that are less personal, although I did organize the conference panel at which they were discussed. At the annual meeting of the Modern Language Association held in San Francisco in December 2008, three presenters spoke to the topic “The Library of Google: Researching Scanned Books.” Two of them, Lisa Spiro and Amanda French, narrated parallel histories. Several years previously each had completed a dissertation, making her an expert on a special topic (Spiro, “Bachelorhood”; French, “Refrain”). Both dissertations were completed before the launch of Google Book Search in 2004, now known as Google Books. In 2008, Spiro and French searched in Google Books for information relevant to their projects, and each found that the new information made available there was at least comparable in scale to the information that she had amassed through conventional library research. That is to say, by 2008, Google Books had at least doubled the particular information that was accessible to a careful researcher. By now, that increase may be geometrical.

The absent guest at Debates in the Digital Humanities, alluded to if not discussed by most participants, was Franco Moretti. “Distant reading,” his brand of literary data mining, shadowed many of the chapters. The findings that I have cited do not participate in distant reading. Instead, they illustrate a different power of search: local, limited, but nonetheless informative. The critical history of “close reading” that follows is similarly enabled by the specific affordances of search in several large digital libraries, including Google Books, HathiTrust, Internet Archive, Early English Books Online, Eighteenth-Century Collections Online, JSTOR, Project Muse, HeinOnline, and British Newspapers 1600–1900.

Such resources have imperfections, which limit their usefulness for datum mining as well as for data mining (Duguid; Nunberg, “Google”; Nunberg, “Counting”). For individual data, however, their failures are usually failures of omission (false negatives, which escape notice) rather than failures of commission (false positives, which can be corrected). It may be that the accuracy of datum mining generally exceeds that of data mining. In any case, a datum, unlike data, can be individually assessed.

When Moretti first mentioned distant reading, in “Conjectures on World Literature” (2000; reprinted in Distant Reading, 2013), he opposed it to “close reading.” At the same time he gave close reading a national affiliation, bracketed it in the recent past, and restricted its application to a narrow set of texts. “The United States is the country of close reading,” he wrote, and “the trouble with close reading (in all its incarnations, from the new criticism to deconstruction) is that it necessarily depends on an extremely small canon” (2000, 57).

All three of these propositions are doubtful. That close reading applied to noncanonical texts of popular culture was evident as long ago as 1948, when Leo Spitzer closely read an ad for Sunkist orange juice.1 Although close reading did thrive in the United States, partly thanks to the influence of Practical Criticism by I. A. Richards (1929) and Seven Types of Ambiguity by Richards’s student William Empson (1930), neither writer was a citizen of the United States and both books were first published in London. Furthermore, the clear affiliation of close reading to the French tradition of explication de texte gave it respectably un-American credentials as well as a genealogy before the New Criticism.2

John Guillory brackets the practice as closely as Moretti does: “By close reading, I do not mean the same thing as reading closely, which arguably describes many different practices of reading from antiquity to the modern era. I assume rather that close reading is a modern academic practice with an inaugural moment, a period of development, and now perhaps a phase of decline.” Guillory locates the “inaugural moment” in the work of I. A. Richards, who “has long been identified with the origins of close reading” (Guillory, 8). However, by 1929 “close reading” was already central to academic practice.

It was a mainstay of classical education:

Philo M. Buck Jr., who published these fond memories in 1914, later became a professor at the University of Wisconsin, where he founded the Department of Comparative Literature—one of the first such departments in the country (Carruth, 221; Cronon and Jenkins, 490).

Thomas Arnold, headmaster of Rugby School—also Matthew Arnold’s father—had enthusiastically described the same process, though he did not call it “close reading,” in a letter of 1836:

“We travel too fast.” The cure for such speed was what Reuben Brower, who had taught Greek as well as English at Amherst College,3 would in 1959 call “reading in slow motion”—that is, “slowing down the process of reading to observe what is happening, in order to attend very closely to the words, their uses, and their meanings” (“The Humanities,” 77). Brower’s slow-motion method implicitly countered the contemporary vogue for teaching “speed reading” in high school.4 An esteemed late practitioner of the New Criticism, Brower accomplished within one language what Arnold did across two. He once acknowledged, “as a teacher of courses in close reading, I have tried to give non-classical students some experience of imaginative but disciplined reading of the text of the kind that once was common in the teaching of the Classics.”5

Thomas Arnold’s half paragraph celebrating “the process of construing” would be quoted in print dozens of times across the rest of the nineteenth century and into the twentieth, on both sides of the Atlantic.6 One function of this litany of citation was to authorize the displacement of Greek and Latin texts in the classroom by British and American ones. Another was to authorize the continuity of close reading, akin to the niceties of classical translation, as a pedagogic staple during the century that led up to the advent of the New Criticism.

By 1897, close reading as such was prescribed to autodidacts, as well as teachers. Sara D. Jenkins, the editor of Popular Educator, fielded a (real or invented?) question that year: “What is meant by ‘Close Reading’? I am trying a course of reading by myself and my author says, ‘Spend thirty minutes daily in “close reading.”’” Jenkins replied, “It might be interpreted variously. My own use and construction of the terms is in referring to diction, figures of speech, translation of allusion, etc.” (Jenkins, 68).7 And she went on to give teachers some advice about pacing such instruction. Evidently “close reading” was part of the toolkit of vernacular secondary education and self-education in the nineteenth century, as well as in classical studies.

When Richards referred to “close reading” in Practical Criticism, he was not coining a phrase but endorsing a standard practice. That his first mention of it is elliptical and casual indicates that he is not inventing a term of art but using common language ready to hand. One reader, he remarks, is “less close in his reading than” another (114). A later remark, that “no close reader will doubt or deny” some proposition (191), presumes that Richards’s actual or ideal reader is already a close reader. “All respectable poetry invites close reading” (203) fully names the practice that has been taken for granted all along, because it was a standard practice. Guillory posits that this “little spatial trope” was “elevated” by Richards “into a disciplinary term of art” (Guillory, 12). But the “trope” was already a term of art, or of standard practice, long before Richards midwifed the New Criticism. (How the trope was “spatial,” or not, will be considered shortly.) Good readers were close readers; and some texts, including some poems, were good enough to warrant close reading.

The text that most required close reading was the Bible—the canonical text, the book of books. In 1805 one “H. H.,” writing from Lambeth, spoke of “giving [Psalm 15] a close reading, and seriously searching into its true import” (399). That same year “Gaius” remarked on a supposedly obvious meaning in Psalm 23 that “it does not require much close reading, nor very deep researches into the meaning of the Psalm, to perceive this” (Gaius, 594; emphasis in the original). In 1795, a cleric wrote to the Gentleman’s Magazine that “for several years past I have devoted my time to close reading and study” (Cleros, 489).

Evidently “close reading” could describe a prolonged course of study as well as the scrutiny of a single text or passage. So William Cave, in 1673, reported that “I set myself to a more close and diligent reading of the first Fathers and ancient monuments of the Church than ever I had done before” (preface, n.p.). “Close study” was another name for it; so John Tillotson, archbishop of Canterbury, “began with a deep and close Study of the Scriptures, upon which he spent four or five Years, till he had arrived at a true understanding of them” (Burnet, 13). “Close perusal,” “close reasoning,” “close meditation,” “close attention,” and “close application” were qualities of both particular and far-reaching study from the sixteenth to the twentieth centuries.

The young philistine Stephen Gosson is suggestive in his satire on such ivory-tower study:

Presented with this passage, Empson would have noted a kind of ambiguity. The “close study” in question here could be the intellectual activity of close reading, or it could just as well be a small room in which such reading takes place—where the university scholars “coop themselves up” to pursue such study.9 Close study is closeted study. Close study, and close reading, have something to do with spatial constraint, with tightness and restraint. Such closeted tightness can be deprecated, as E. C. Stedman once deprecated certain early poems by William Morris as “pieces which repay close reading, but also compel it, for they smack of the closet and library, rather than the world of men and women, or that of woods, waters, and hills” (Stedman, 233).

The tight, taut (taught) attention to the text that is and was close reading was matched, if not indeed motivated, by the “closeness” of the texts that it confronted. If reading, study, reasoning, analysis, and attention could be “close,” so could diction, discourse, or style and the thinking that it displays, and even the handwriting or printing (print or matter) in which it circulates.

Style especially: Quintilian’s analysis in Institutio Oratio, influential for millennia, is worth close study in this regard. The broadest distinction of styles, we learn from an eighteenth-century translation of this treatise, was between the “grand” and the “plain”—styles associated with “the Asiatics and the Attics.” The distinction was “of an old Standing: the latter affected to be close and concise, and the other were blam’d for an empty, bombast Manner.”10 Here the phrase “close and concise” translates pressi: constrained, compressed. Earlier we read, “one Kind of Prose Style is close and compacted; another, such as we use in Letters, in Conversation, is loose” (2:303). Here “close” translates vincta (bound, fettered, confined)—the opposite of “loose,” which translates soluta (unbound, unfettered, free). “Compacted” (which here translates contexta) is a synonym of “close” (vincta); more fully contexta means “woven together, closely connected, continuous.”11 Thucydides had a distinctive style, we learn in another eighteenth-century translation of Quintilian: “Thucydides is close, concise, and ever going on” (that is, hurrying forward).12 “Densus [dense] et brevis”: “close, concise.” Pressi, vincta, densus; compressed, confined, concise, dense: the “spatial trope” of closeness turns out to be a matter not of proximity but of tightness and density.

Such economy of the writer’s means could increase the reader’s labor, as Henry Felton acknowledged in 1713: “Thucydides does sometimes write in a Style so close, that almost every Word is a Sentence, and every Sentence almost acquaints us with something New; so that from the multitude of Clauses, and Variety of Matter crowded together, we should suspect him to be obscure” (Felton, 181).13 Although Felton would acquit Thucydides of such a charge, he did note the hazard of a close style: “while we study to be concise, we can hardly avoid being obscure. We crowd our Thoughts into too small a Compass, and are so sparing of our Words, that we will not afford enow to express our Meaning” (108; emphasis in the original). Later in the eighteenth century Fulke Greville rehearsed a similar apology, apropos La Rochefoucauld’s maxims, for the dangers of verbal compression: “Many persons have thought these reflections obscure, not only in the expression but in the sense. Obscurity, however, is not always the fault of the writer. Reflections, or maxims and sentences, as the world has called these, ought always to be written in a close style, which does not admit the utmost degree of perspicuity” (Greville, xix). A “close,” sententious style is not the writer’s fault, but the reader’s responsibility.14

As for verbal style, so for chirographic and typographic. Close writing—that is, close handwriting—could be hard to read. (“He opened a drawer, and found there a manuscript, of close writing, so that much of it was illegible.”15) And so could “close printing” or “close matter” or “close type”—the typographical equivalents of close writing. In the nineteenth century a closely printed “fasciculus of some cyclopedia” would be cheaper and more compendious than a regular book; and so “it beho[o]ves us to mount our spectacles on nose to con the small letter-press,” even though “the close writing begets close reading, and compels us to endue [that is, put on] the garment of attention, which to many is like the shirt of Nessus to Hercules—consumedly troublesome.”16

“The close writing begets close reading.” Some, if not all, close reading was occasioned by the density (closeness) of the print, or of the handwriting, or, most significantly, of the verbal style. Magnification might help for the close printing or writing. “Close” attention might suit a close verbal style, one rich in implication, like that of the Bible or a sonnet by Shakespeare.17 The book of Proverbs, in particular, was often spoken of as “close.”18 Shakespeare’s sonnet 129, as closely read by Laura Riding and Robert Graves in a chapter of A Survey of Modernist Poetry, was the acknowledged inspiration for Seven Types of Ambiguity, the gospel of the New Criticism.19

Of course, the “close reading” of the New Criticism was usually not just a matter of reading, but also of writing—writing done after or during reading. The “close attention to the printed word” that Brower modeled for his students, “the closest scrutiny of meanings and forms of expression,” would typically find expression in writing, not just reading (Brower, “The Humanities,” 79–80).20 “Teaching of reading is necessarily teaching of writing. The student cannot show his teacher or himself that he has had an important and relevant literary experience except in writing or in speaking that is as disciplined as good writing” (81). And such writing was often abundant, rarely “close.” Empson’s Seven Types of Ambiguity (arguably still apprentice work, written by a recent college graduate at the prompting of his former tutor, I. A. Richards) is considerably longer than the handful of poems it expounds. As Wikipedia puts it, “A truly attentive close reading of a two-hundred-word poem might be thousands of words long without exhausting the possibilities for observation and insight” (Wikipedia, “Close Reading”). Close reading had long been a matter of inward, private study—indeed, of “observation and insight”: more a matter of understanding than of interpretation in the public sense of that word (Hancher, “What Kind of Speech”). However, in tutorials such as Richards’s and in classrooms such as Arnold’s and Brower’s, it became an occasion for semi-public performance. The actuality of that performance, when registered in writing, might achieve full publicity in print.

D. A. Miller has appreciated the enduring charms of close reading, now thought to be obsolescent. Now, relieved of its stern “respectability,” it can “come out as a thing that, even under the high-minded (but now kitschy-sounding) rationales of its former mission, it had always been: an almost infantile desire to be close, period, as close as one can get, without literal plagiarism, to merging with the mother-text” (Miller, 58). This looks like Guillory’s “spatial trope” of closeness raised to the highest degree—the extreme opposite of Moretti’s “distant reading.” But all these spatial readings of “close reading” are misconceived. Distant reading is not the opposite of close reading. The “closeness” of close reading has long concerned not proximity but density and concentration: concentration certainly in the reader and often, as well, in the text being read. So it was for the New Criticism, and so it was for the old. The opposite of close reading is not distant reading but loose, casual, and careless reading.21 Yet, paradoxically, when close reading is recorded in interpretive writing it can rarely be close but must expand and exfoliate to report the unpacking of details.

Miller’s obsequies for close reading were premature in any case. Thanks to the recent Common Core State Standards Initiative, close reading is now central to primary and secondary school curricula in the United States. The Introduction to “Common Core State Standards for English Language Arts and Literacy in History/Social Studies, Science, and Technical Subjects” stipulates that “students who meet the Standards readily undertake the close, attentive reading that is at the heart of understanding and enjoying complex works of literature” (“Common Core,” 3). The first “key idea” for “College and Career Readiness Anchor Standards for Reading,” as framed both for grades K–5 and also for grades 6–12—that is, applying all the way from kindergarten to the high school diploma—is, “read closely to determine what the text says explicitly and to make logical inferences from it; cite specific textual evidence when writing or speaking to support conclusions drawn from the text” (10, 35). A marginal gloss on the K–5 rubric comments that “students . . . acquire the habits of reading independently and closely, which are essential to their future success” (10). Evidently the idea of close reading is fundamental to the Common Core, even though the phrase itself does not appear in the Standards. The phrase does appear twelve times in an authoritative auxiliary statement written by David Coleman and Susan Pimentel, “Revised Publishers’ Criteria for the Common Core State Standards in English Language Arts and Literacy, Grades 3–12.” For example: “Close reading and gathering knowledge from specific texts should be at the heart of classroom activities and not be consigned to the margins when completing assignments” (Coleman and Pimentel, 4–6, 8–9, 15–16). Coleman, who is president and chief executive officer of the College Board, and Pimentel, who is vice chair of the executive committee of the federally appointed National Assessment Governing Board, were principal agents in establishing the Common Core Standards.

As might be expected, such official emphasis on close reading has caused widespread curiosity, even anxiety, among teachers and parents. The question that Sara D. Jenkins raised in the Popular Educator in 1897, “What is meant by ‘Close reading’?” now echoes thousands of times online, if with a more immediate urgency and in a different vernacular—for example: “So, what is close reading, anyway?” or, “What Close Reading Actually Means,” or simply, “What is Close Reading?” (Hodgson; Wiggins, “What Close Reading”; Shanahan). Part of the answer is that the New Close Reading (to give it a name for now) is a rejection of the sixties and seventies vogue for reader-centered reading. Coleman and Pimentel use some understatement in putting the matter: “When examining a complex text in depth, tasks should require careful scrutiny of the text and specific references to evidence from the text itself to support responses” (6; emphasis added). Nancy Boyles is more direct: “the teaching of reading veered significantly off track when . . . personal connections . . . began to dominate the teaching and testing of comprehension, often leaving the text itself a distant memory” (Boyles, 7; emphasis added). Grant Wiggins would go so far as to revive the author of the text: “the goal is to understand what the author is doing and accomplishing, and what it means; the goal is not to respond personally to what the author is doing” (Wiggins, “On Close Reading, Part 2”). Vintage reader-response theory, influential in the latter half of the twentieth century, is now thought to be passé.22

Of course, the methodological debate extends a hundred years further back, to the sixties and seventies of the nineteenth century. Walter Pater neatly framed the alternatives apropos Matthew Arnold, whom he agreed with, only to contradict: “‘To see the object as in itself it really is,’ has been justly said to be the aim of all true criticism whatever; and in aesthetic criticism the first step towards seeing one’s object as it really is, is to know one’s own impression as it really is, to discriminate it, to realise it distinctly” (Pater, vii, quoting Arnold, 6). One does not have to choose between Arnold and Pater to recognize that close, attentive reading, whether of the object itself or of one’s apprehension of it, has long been highly valued in the construction of aesthetic and literary experience, and that it continues to be so.

A simple Google Books Ngram Viewer chart of “close reading” for the two centuries with the best data, 1800 to 2000, drawing on a corpus of more than five million books, shows that close reading has been a matter of growing concern for the entire period, especially after 1870—not contained by the eras of New Criticism and deconstruction, but older than that, and newer than that:

Data for more recent years, if properly adjusted, might show the curve to have lifted even higher, propelled by the Common Core. As things stand, according to an FAQ page published by the designers of the Ngram Viewer, any data after 2004 would be greatly complicated by the launch that year of the Google Books project itself.

The same FAQ page offers reassurance at the very outset, under the rubric “Big Picture,” about the continuing relevance of close reading, despite the advent of “Culturomics,” the Ngram Viewer version of distant reading:

The same is true for particular search.

The revisionary reading of “close reading” that I have proposed here depends (except for Figure 12.1) not on distant reading but on minute particulars discovered by digitized access to published texts. I could not have written this chapter ten years ago. The digital humanities revolution includes unprecedented discovery of texts to read, sometimes closely.

Of course it still pays to read books printed on paper, such as Macroanalysis: Digital Methods and Literary History by Matthew L. Jockers (2013), or Passages from the Life of a Philosopher by Charles Babbage (1864).

In Macroanalysis Jockers proposed an adjustable style of computer-assisted reading, one not merely distant nor merely close, but flexible in its focus. (He, too, conventionally imagines that closeness has to do with proximity, not density.) “The strength of the approach allows for both zooming in and zooming out” (23). Such an approach, he believes, promises more than did older, search-driven inquiries after what is only already known:

This is a limited view of both search and close reading. The new ability to access and organize minute particulars can do more than inform our understanding of an isolated text (though it can do that), and it can generate new knowledge, not merely sustain old prejudices. Thanks to such resources as Google Books, HathiTrust, and the Internet Archive, search has become the Oxford English Dictionary on steroids—at least as regards illustrative quotations, if not definitions. The history of ideas is the history of the language in which ideas are constructed and expressed, and that history has now become available with a revelatory granularity. All that is denied our grasp—no small loss, admittedly—is the lived, oral language of ideas that never did get recorded in writing, in print, or in any other durable medium (Harris).

For more than half a century the British mathematician Charles Babbage (1791–1871) has been celebrated as the “father of the computer,” in part because of the publicity that he gave to his plans to build elaborate calculating apparatuses that would process mechanically coded information, and also because of the credit assigned to his enterprise by Howard H. Aiken (1900–1973), who designed the pioneering IBM Automatic Sequence Controlled Calculator (Mark I) that was envisioned in 1937 and installed at Harvard University in 1944.24 Although neither Babbage’s “Difference Engine” nor his later “Analytical Engine” ever reached a similar stage of completion, those thwarted machines haunt Babbage’s late memoir Passages from the Life of a Philosopher, along with a miscellany of other topics, some more relevant to computing than others. One of these topics, the idea of “forming a universal language,” had preoccupied Babbage as an adolescent. Even before enrolling at Trinity College, Cambridge, he “accidentally heard . . . of [the] idea.”

John Wilkins, later Bishop of Chester, had served as master of that same Trinity College for one year, from August 1659 until August 1660. In November 1660, he chaired the first meeting of an assembly later to be chartered as the Royal Society of London, characterized as a “college for the promoting of physico-mathematical experimental learning” (Birch 1:4). Wilkins’s most important publication, a substantial folio titled An Essay towards a Real Character and a Philosophical Language, was published by the Royal Society in 1668.25 The goal in that book was to devise “a Real universal Character [that is, an objective, transcendental code] that should not signifie words, but things and notions [concepts], and consequently be legible by any Nation in their own Tongue”—thereby remedying the linguistic diversity that mankind had suffered as a consequence of the debacle at Babel, a “confusion” that had impeded both the advancement of science and the propagation of international commerce (Wilkins, 13). An early biographer reported that Babbage “studied with much attention the investigations of his predecessors” in the search for a universal language, “and especially Bishop Wilkins”—the most thoroughgoing of seventeenth-century speculators in this line (Buxton, 347).

Wilkins’s grand project was not a success on its own terms; if it had been, Babbage would not have had as many occasions as he did to call for the establishment of a universal language. Some of these calls were modest in scope—recommending a “universal language” for signals emitted from all lighthouses, for example, or “a universal language” of complex-machine specification (Babbage, Notes, 14; Exposition, 139). But Babbage could also press the matter in earnest. The question of what should be “the language of science” having been raised at a meeting of the International Statistical Congress in London (July 1860), Babbage gallantly conceded “that French is the language to be adopted,” although “other views ought also to be considered” (Report of the Proceedings, 394). In any case, he argued, “the existence of different languages is a great evil; it is the destruction of a certain amount of the intellect of mankind, which is thus consumed by the friction that the different languages create.” The global spread of “the commerce of England” promised to spread the language of England as well, but there was the risk that different colonies might develop different colonial Englishes. “I think there are measures which might be taken to prevent the gradual branching off of our own language into a number of separate dialects or languages, and perhaps of ultimately rendering it the universal language of the world.” Authoritative dictionaries, prescriptive as well as descriptive, would be essential to that task.

If Wilkins failed to establish a universal language, he did address Babbage’s original conundrum, indirectly contriving to “[arrange] signs in a consecutive order, so as to find, as in a dictionary, the meaning of each when wanted.” The arrangement was complex but complete, at least in theory: using the English language, Wilkins’s book provided an elaborate tabular, analytic taxonomy of things and ideas, to which an alphabetic index or dictionary (also in English) provided access; and between these sections a code of constructed ideograms (“real characters”) was proposed to articulate nodes in that taxonomy.26

The account of Babbage’s memoir published by the British Quarterly Review ignored Wilkins and simply paraphrased the young Babbage’s dilemma: “What would be the use of a lexicon to a lad if all the words were printed promiscuously, or drawn up in an abstract array and under ideal heads, as they are in Roget’s Thesaurus?” (“Passages from a Philosopher’s Life,” 97). That is, absent a regular alphabet enabling an index, each of the two possible arrangements for a set of invented signs—randomness on the one hand, or abstraction on the other—would be equally inaccessible and unsatisfactory.

James Gleick aptly characterizes Babbage’s lexicographical impasse at this point as “a problem of storage and retrieval” (90)—that is, search. To enable search, something was needed like the Analytical Engine. The “consecutive order” that gives the most basic access to stored data is not thematic nor even alphabetical but digital: sequential numbers underlying any code and providing access to any of its expressions. The digits might be decimal or binary, or some other number base; the apparatus might be mechanical or electronic; no matter, the principal is the same. “Babbage had all the essential ideas,” as Alan Turing remarked in 1950 (439). For Babbage’s projected machine, decimal digits were most efficient, and for the electronic apparatus of the twentieth century and now, binary. Furthermore, electronic computers can now be networked, vastly enlarging the capacity of storage and the reach of search.

Babbage’s hopes for a universal language, his scheme for a universal dictionary, and his plan for an Analytical Engine having all failed, the father of the computer would have welcomed the binary ordering that now empowers networked electronic search, making it possible (among other things) for scholars to discover and to organize minute particulars in the actual written record. As two editors of the Economist pointed out at the turn of our millennium, “If there is such a thing as a universal language, it is not English but binary” (Micklethwait and Wooldridge, 35–36).

Notes

Bibliography