Section: Controversy

THE IMAGERY OF A MYTH: COMPUTER-ASSISTED RESEARCH ON LITERATURE

Doubts have often been expressed about computer-assisted research on literature, but until about seven years ago these doubts were voiced mainly by those who did not themselves practice it.[ 1] Since 1988, however, real worries have increasingly been expressed by specialists in books and periodicals devoted to computer-assisted research on literature (henceforth CARL). They ask questions not only about methods, but, more seriously, about the assumptions behind the methods. These questions are starting to be put with a harshness unprecedented in CARL. Thus Willie van Peer, in a 1989 article in Computers and the Humanities, goes so far as to say that the view of language underlying quantitative studies is "utterly naive" (303), while in 1991, in Literary and Linguistic Computing, Thomas Corns talks of certain hopes for CARL as resembling a "fanatic's dream" (128). These doubts have culminated in a recent special issue of Computers and the Humanities (1993-94) in which contributors call still more urgently for internal revaluation and ask more pressingly than before whether CARL should not be moving in new directions.

During these last seven years, CARL researchers have voiced a special unease about the fact that almost all literary critics of repute ignore the results of CARL investigations. This unease was first highlighted in a colorful and provocative article by Rosanne Potter ("Literary Criticism"), and the recent special issue of Computers and the Humanities once more points out that in its present form CARL is marginalized by most literary critics.[ 2] The CARL experts who have asked why this should be so have usually provided what is no doubt the correct answer: that the conclusions of most individual CARL projects have simply been too trivial or too obvious to attract attention.[ 3] A second reason put forward for the marginalization of CARL is the rebarbative presentation of its research; this is a particular concern of Potter's ("Literary Criticism" 91, 94, 97), but others have echoed her.

There may be a third contributory factor, one that has not yet been singled out by any practitioner nor, I believe, by external critics. Some surprising figures of speech infiltrate the critical diction of many CARL analysts--figures of speech that tend to mythologize their own enterprises. This self-mythologization is, perhaps, both a symptom and a cause of the problem. It may have stopped CARL experts evaluating properly the results of their own research, and it cannot but be off-putting to the non-CARL critics they are trying to win over.

It is perhaps natural that some CARL experts should see themselves as pioneers or navigators, going downstream to check information at the confluence of the valleys and fortunately not looking upstream at all the vertiginous tasks ahead:

II a donc fallu remonter en aval du Dictionnaire des frequences et pulser l'information tout pres de la source, au pied des textes, au confluent de 15 vallees ou convergeaient les formes brutes des 15 tranches chronologiques . . . en amont que de taches nous attendaient, qui auraient pu nous donner le vertige si nous avions porte assez loin le regard. Heureusement le brouillard enveloppait notre depart. (Brunet, Le Vocabulaire francais I 19)[ 4]

We therefore had to go back downstream from the Dictionary of Frequencies, and extract information very close to the source, at the foot of the texts, at the confluence of 15 valleys where the raw forms of the 15 chronological sections converged [ . . .] upstream, how many tasks were waiting--tasks which might have made us dizzy if we had looked far enough ahead. Luckily fog swathed our departure.

Or:

[T]he present study constitutes an entree en matiere, a first voyage into the quantitatively uncharted waters of contemporary narratological pragmatics. (Frautschi 280)

But it is rather more surprising to find them as mountaineers:

[L]es epigones auront-ils plus de courage que les devanciers et suivront-ils la pente raide ou Gunnel Engwall s'est engagee? (Brunet, "Review" 251)

Will imitators be braver than precursors were--will they follow up the steep slope Gunnel Engwall has embarked on?

Or as horse-riders:

As an experienced rider controls his horse with ever more subtle commands, so the critic may control the exact routines of the computer with more latitude. (Smith 39)

Or as big-game hunters:

[Computer-knowledge is] unrestrained curiosity, drawing people to safaris in the jungles of the unknown. (Busa 71)

They may even share this outdoor life with their chosen authors: certain stylo-statistical tests show that Marlowe is like a "young Olympic quality athlete" who "may dominate the track in all events" (Baker 37). More charismatic still, they are wartime-leaders: Gunnel Engwall ends her description of her work hoping that Churchill's words will have "quelque validite egalement ici: 'Cela n'est pas la fin. Cela n'est pas non plus le commencement de la fin. C'est la fin du commencement'" ("some validity here too: 'This is not the end. It is not even the beginning of the end. But it is, perhaps, the end of the beginning'" [lvii]).

CARL experts may also be such characters of modern legend as the detective (Sherlock Holmes, of course: H. and G. Logan 247), or heroes of Greek myth--fighting hydras or rolling eternal rocks like Sisyphus:

Gunnel Engwall a ose affronter l'hydre aux mille tetes bourgeonnantes [ . . .] heroique jusqu' au bout. (Brunet, "Review" 251)

Gunnel Engwall has dared to confront the hydra of a thousand sprouting heads [ . . .] heroic to the end.

II fallait doric pousser un peu plus loin le tocher de Sisyphe. (Brunet, Le Vocabulaire francais I 24)

So we had to push Sisyphus's rock a little further.

And they give their projects fairy-tale dimensions: computer-users are magicians and the computer gives its adepts seven-league boots or new eyes--better bodies, in short. Thus Charles Muller refers to "formules magiques" ("magic formulae"), "apprentis sorciers" ("sorcerer's apprentices"), "maitres sorciers" ("master sorcerers") (153,157). And for Helmut Schanze, "the theme is one of a fairy-tale development from printed literature to the PC Age" (171); "tiny desktop micros" are "winning the mythical war against weak and old computer giants" (171). The electronic age has taken its first steps in "hazardous Siebenmeilenstiefeln--seven-league boots" (171), and systems for analyzing data "form an expansion of our eyes" (173). Schanze may be to some extent self-consciously jocular here; yet there is no note of irony when elsewhere in the same article he expresses another "better-body" hope that computers might become "the new memory device, a function which the brain and the book have served for a millennium" (172).[ 5]

This evocation of myth and fairy-tale spills over into other supernatural and even religious references. Again, while there can be an element of humor here, at times it is difficult to know whether the tone is genuinely tongue-in-cheek or verges on something more uncomfortable. L. D. Misek-Falkoff, in an otherwise sensible article, remarks that "What one will now be able to learn about critical axioms" will "surely lead us to states of awe" (20); indeed, for Etienne Brunet, CARL becomes potentially a twentieth-century deity: "quand depuis quinze ans on a voue sinon son culte, du moins son temps, au dieu moderne, ce serait une cruelle desillusion d'apprendre qu'il se trompe" ("when, for 15 years, you have devoted to the modern god if not a religious dedication, at least your time, it would be cruelly disillusioning to discover that this god is wrong") (Le Vocabulaire de Zola I 3). Similarly, Louis Milic writes: "We are at the feet of the goddess of truth, but instead of beseeching her for help in finding the right way, we seem to be struggling among ourselves" ("Progress" 399). And Muller, particularly prone to religious analogies, says the problem of homograph-disambiguation "se dresse, tel l'ange au glaive flamboyant, a l'entree du paradis quantitatif" ("rises up, like the angel with the flaming sword, at the entrance to a quantitative paradise"); the opponents of disambiguation are taking up what is for them "l'attitude la moins entachee du peche originel" ("the attitude that is the least stained with original sin"), and they regard disambiguation as "damnable" (115,133,140). Finally, Literary and Linguistic Computing published in 1992 a short piece by one of the doyens of CARL, Roberto Busa SJ, in which Busa remarks:

Being a priest, people often consider my presence in computer science as exotic . . . But it is precisely as a priest that I am doing what I do. In fact analysing texts leads to realizing the presence of the mystery of God at the roots of human understanding and speaking . . God is the grandfather of the computer. (70,73)

In fact, Busa's presence in literary computing is less unusual than he suggests, for the number of humanities computing articles about Biblical texts is striking.[ 6] Could some of the imagery just cited be inspired by the religious origins of the concordance? As Joseph Raben says, it was hoped that Biblical concordances would reveal the "heart" of the truth underlying overlapping and partly contradictory versions of Christ's life as recorded in the Gospels ("Humanities" 343); possibly some of these hopes have spilled over idolatrously into secular uses of CARL.

There are other figures of speech that inflate the value of CARL running through the publications of many CARL experts. It would be almost as interesting as the images I cite to consider images of new-era revolution, of golden ages and Utopias, of burdensome tasks bravely undertaken, and so on. But those already described are the most striking, and do have an obverse side in the mythologization by a few CARL experts of their "opponents": those literary critics who attack or ignore CARL. Here, the vocabulary used tends to imply that non-CARL critics are degenerate, shoddy, domineering or dangerous. Thus Paul Fortier accuses Genette of "intellectual decadence," and says: "I am appalled by the flabbiness of most contemporary writings about literature" ("Analysis" 81, "Theory" 193); according to Ross, non-CARL critics "make up yarns" (45); to Sutherland, the commitment of professional critics, and students, to certain types of reading is a "distorting ideological commitment" that needs "breaking" by activities like CARL (308); and, for Burrows, "the structuralists" during a certain period "roamed at large across the world of ideas"--like wild beasts, presumably ("Computers" 187).

Even the well-respected Louis Milic uses similar imagery in a response to Stanley Fish, published in 1985 as a review article ("Contra"). Milic compares Fish, in his attack on stylistics and stylometrics, to an onlooker scoffing at early aeronautical adventurers ("It will never fly, Fish would have said to the Wright brothers" [388]); Fish is presented by Milic as superstitiously afraid of CARL, forgetting that "it was not Satan but men who invented mathematics, linguistics, machines and all he rails at" (390). In Milic's words, Fish believes he has "devastated" stylistics (391), was "throwing a bomb at Jakobson" (388). But to Milic, the intellectual behavior of Fish is disreputable or even promiscuous. Milic says disparagingly that "Fish used to hang around" with philosophers of language (389), whereas, Milic suggests, CARL users have their sexuality under control, for they are "not necessarily lusting to be literary critics" (390). If it is not lust, then it is inebriation, for Milic also argues that a stylistic analysis in another essay of Fish's evokes a reader who "stumbles around reeling from one possibility to another" (392). But the bottom line is that the only kind of reader who would really puzzle over this material analyzed by Fish would be a reader "in Fish's employ" (392). As for Fish himself, Milic regards him as sanctimoniously "swathing himself with a hieratic mantle" of a particular kind of "piety" (390). But Fish later turns his attention to the law, and in this respect, says Milic, he has "undoubtedly progressed . . . I have heard that he is now flailing the misdeeds displayed in the prose of the legal profession" (393).

Milic is not the only CARL expert who judges opponents harshly. A similarly extreme image is created in that same 1988 article in which Potter argues for a rapprochement between CARL and non-CARL ("Literary Criticism"). Potter claims that

[m]any literary researchers moved to scientific approaches to texts because they were unconvinced by the high-wire, speculative balancing acts that passed for criticism during their years of graduate education. (92)

Their literature professors, she says, were "oh-so-unsystematic" (94), and critical assertions of the 60s and 70s "appeared to be acts of sorcery . . . . Studying with these high-wire artists," she claims, "was like watching Sherlock Holmes solve crimes without hearing his explanations to Watson"; this kind of criticism was "'dancing on the ceiling'" and the professors "mental Fred Astaires" (92). Furthermore, says Potter, "Literary critics have been glad to borrow terminology from anywhere and everywhere . . . . [A]ny well-argued assertion is as valid as the next one" (93). Finally, at the end of the article Potter reminds us that the non-CARL literary critic is a quack Sherlock Holmes, for now the real Holmes can be rehabilitated: "Let us reclaim the brilliance of Holmesian deduction by being solidly rooted in evidence; then our intellectual discoveries will seem 'Elementary, my dear Watson'" (97). Elementary?

Thus, the picture built up by Milic, Potter, and others suggests that non-CARL critics belong to unrespectable, even seamy, strata of society. In Milic, they become bomb-throwing terrorists, groupies, Tartuffes, devil-fearers, lustful; they assume a kind of drunkenness in readers who "stumble around reeling", and they gain support corruptly by having those readers in their "employ." Their reputation proceeds by rumor and cruel showmanship ("I have heard that he is now flailing the misdeeds"). In Potter, the showmanship moves to the fore with the images of the circus and Fred Astaire; literary critics are charlatans (fake detectives); they are spongers, maybe even promiscuous, "glad to borrow assertions from anywhere and everywhere." And still in 1991 Potter unrepentantly continues to use words like "sham" of non-CARL criticism, while any errors early CARL may have made are pleasantly attributable merely to its youth, to "toddler and teenage thinking . . . . It was an innocent process . . . . We've been playing, which is natural for young folks" ("Statistical Analysis" 427,428).

We may begin to suspect, by now, that in these critiques there is a reasonably coherent image at work, not one totally coherent, but we need not expect that of fantasy or myth. On one side, there is the over-sexual, disreputable, and even dangerous literary critic. On the other, there is the clean-limbed hero who sets off on outdoor adventures that demand physical skill and privileged access to the supernatural.

What provokes this tendency to mythologize? First, no doubt, the glamour of technology. Northrop Frye wryly remarks, in an introduction to a humanities computing conference, "I often find that, when I read books about the technology available in the near future, the author's eyes are starry while mine are still glazed" ( 8). Indeed, as a matter of record, in all the CARL publications I have surveyed, only one writer reminds the reader that the computer is infinitely less powerful than the brain (Koch 183), CARL workers do sometimes indulge in sheer wonderment at machinery, suggesting, for instance, that we live in eras in which the "uniting of human efforts to machine capabilities" makes us see that "there is indeed something remarkable left beneath the visiting moon" (Widmann 63).

Not just the glamour of technology, but a more generalized glamour of science plays a part. Many CARL practitioners still evoke a scenario in which they are "scientists" while other literary critics are imprecise, impressionistic, intuitive (this latter word never used as a compliment). The ambition to be considered scientists is perhaps apparent in the very obtrusiveness with which CARL researchers put graphs, tables and mathematical formulae in the body of their discussions. This ambition shows too in the unrealistic exhortations to other literary critics to learn programming and statistics.[ 7] But it also emerges in quite specific statements and comparisons. CARL users are like "chemists who carefully construct their series of successively excluding tests only to make their greatest discoveries inadvertently" (Potter, Literary Computing xxviii); they could even be like Wilkins and Franklin discovering DNA (Bailey 3-5). On the other hand, "the traditional literary critic or scholar asks questions and forms hypotheses he is reluctant to test about data he is reluctant to measure" (Beatie 190). Furthermore, says Potter, "In the field of literary criticism, the truth of assertions is not valued," and literary criticism "has great need of the scientific methods that can be learned in the quantitative disciplines" ("Literary Criticism" 94). Potter goes so far as to link CARL with behaviorist views of the mind and to assume that these can and should prevail over others. Urging CARL users to be patient while waiting for recognition, she reminds them that the changeover to scientific methodologies has not completely succeeded even in a social science like psychology (in the 50 years since Watson changed the terms of the discourse). ("Literary Criticism" 92)[ 8]

Examples could be multiplied; a last one is Fortier's claim that, on the one hand, "much of modern criticism [i.e. non-CARL] is somewhat closer to science fiction than to science," but with the computer, on the other hand, we can begin to adopt a "scientific model of critical discourse" and "aspire to the type of certainty taken for granted in the sciences" ("Theory" 194, 195). All this is of course bolstered by the very widespread assumption that CARL is "objective" whereas non-CARL is "subjective." (Much could be discussed here; suffice it to say that simply on its own terms, CARL is marked by "subjectivity" at many levels: not only in the choice of material and the type of test applied, but also in various subsequent decisions to ignore some results and promote others. Finally, no database yet exists that is completely lexically disambiguated; thus many CARL results are not only "subjective" but may be misleading or inaccurate.[ 9])

The attractiveness of machines and science, then, plays a part in the myth. But other motives may enter into its construction. No doubt the stress on heroic or scientific firmness expresses some kind of anxiety about the ambiguities of literature. Perhaps, too, it is a compensation for the low professional status of CARL. Yet, further, behind the wish for mythical and professional status, there may be an economic motive. One would hesitate to advance this notion if it were not plainly stated in CARL publications themselves. One CARL expert says,

Experience shows that the creation of software will delay rather than accelerate the creator's chance of promotion in an academic establishment. This is an unfortunate fact of life which grant awarding bodies such as research councils should take note of and perhaps award contracts for the production of such packages. (Roper 132)

Fortier more optimistically foresees that an overturning of professional criteria will allow CARL critics to rise to the top and push non-CARL critics to the bottom:

As such [computerized] information becomes widely available, we can look forward to a new structure in the academic critical establishment---one which distinguishes clearly between fact on the one hand and opinion on the other, and gives pride of place to those practitioners who best account for the former. ("Analysis" 91)

And Paul Delany and George Landow are no less explicit:

[A] more literal presentation promises to disturb theoreticians, in part, of course, because it disturbs status and power relations within their--our--field of expertise . . . . Unfortunately, software development is often not recognized as equivalent to conventional publication, with its known rewards of tenure and promotion. (10; 40)

In short, "a literary scholar interested in hypermedia must operate more like a scientist, with substantial research funding for equipment and salaries" (40; my italics). Is it too cynical to wonder whether not only the self-mythologizing of CARL, but also its new self-questioning are a half-conscious, half-unconscious strategy to win the acceptability and rewards so far denied?

If the computer can offer something to stylistic studies--and it is still an "if," this contribution will probably always be a minor one, simply a pointer to larger questions; moreover, whatever the contribution, it must surely be one that keeps tone and syntactic context to the fore. CARL analysts have not yet fully confronted these two problems. They do have an admirable expertise at their fingertips, but it may be difficult to use this fruitfully until the myths surrounding it are deconstructed.

Notes

1 For example, by Barbara Herrnstein Smith and Christian Mair, and, most swingeingly, by Stanley Fish.

2 Mark Olsen's essay is especially telling on this point.

3 For arguments proposing this hypothesis, see the essays in the special number of Computers and the Humanities. Even author-attribution, so far the most "headline-catching" aspect of CARL, has increasingly been thrown into doubt, with commentators such as David Robey suggesting that these studies have "had their day" (241), and with such subtle CARL users as John Burrows showing that great writers can write so differently from themselves that they can fool the tests into deducing that 'they are another author ("Not Unless").

4 Here and elsewhere, translations are my own. The paraphrase I provide before the quotation clarifies the metaphorical landscape somewhat. One surely could not go upstream from a "source" located in a confluence of valleys. For other "pioneer" images see, e.g., Peter Denley (15, 17).

5 For other "magic" vocabulary, see Etienne Brunet on the CARL "conjuror" ("pres-tidigitateur") who should not hide his data under the table (Le Vocabulaire de Zola 14 n. 5); and Jean-Yves Tadie on the "fairy story" ("conte de fees") of CARL (in his introduction to Brunet's Le Vocabulaire de Proust, I i). In both cases, the rhetoric is deliberate: Brunet is using the conjuror analogy critically, and the goal of Tadie's introduction is to highlight the admirable aspects of the project; but it is still interesting to see what type of imagery occurs to both writers.

6 Two examples among many are the articles by H. H. Greenwood on St Paul's epistles and by B. A. Nieuwoudt on Old Testament texts. Additionally, David Holmes even suggests that CARL can help us towards the revelation of the voice of God speaking through such prophets as Joanna Southcott and Joseph Smith.

7 For example Raben, who says that "within reasonable limits" user and programmer will have to be the same person ("The Humanist" 7); and Milic, who says that "those who keep reminding us of the need to be some kind of statistician if we are to be any kind of stylistician are right" ("Progress" 396).

8 Although Potter gives no references here, she must mean the John B. Watson who is regarded as the precursor of behaviorists such as B. F. Skinner.

9 This is unfortunately the case with, for example, the work of Brunet; his database is that of the Tresor de la langue francaise (now FRANTEXT or ARTFL), but this database does not even disambiguate words such as "danse" (meaning "dance," either noun or part of the verb), and "or" (meaning either "gold" or the narrative conjunction "now"). More generally, Geoffrey Leech and Michael Short argue tellingly against the view that "style can be defined objectively in terms of the frequencies of the linguistic properties of a text or corpus" (70).

Works Cited

Bailey, Richard. "The Future of Computational Stylistics." Potter. Literary Computing and Literary Criticism. 3-12.

Baker, John. "Pace: A Test of Authorship Based on the Rate at which New Words Enter an Author's Text." Literary and Linguistic Computing 3 (1988): 36-39.

Beatie, Brace. "Measurement and the Study of Literature." Computers and the Humanities 13 (1979): 185-94.

Brunet, Etienne. Le Vocabulaire de Proust. 3 vols. Geneva: Slatkine/Champion, 1983.

--. Le Vocabulaire de Zola. 3 vols. Geneva: Slatkine/Champion, 1985.

--. Le Vocabulaire francais de 1789 a nos jours. 3 vols. Geneva: Slatkine/Champion, 1981.

--. "Review article." Literary and Linguistic Computing 2 (1987): 251-53.

Burrows, John. "Computers and the Study of Literature." Computers and Written Texts. Ed. Christopher Butler. Oxford: Blackwell, 1992, 167-204.

--. "Not Unless You Ask Nicely: The Interpretative Nexus Between Analysis and Information." Literary and Linguistic Computing 7 (1992): 91-109.

Busa, Roberto. "Half a Century of Literary Computing: Towards a 'New' Philology." Literary and Linguistic Computing 7 (1992): 69-73.

Computers and the Humanities 27 (1993-94). Special no. ( 5-6): A New Direction for Literary Studies ?

Corns, Thomas. "Computers in the Humanities: Methods and Applications in the Study of English Literature." Literary and Linguistic Computing 6 (1991): 127-30.

Delany, Paul, and George Landow ed. Hypermedia and Literary Studies. Cambridge, MA: MIT P, 1991.

Denley, Peter. "The Computer Revolution and 'Redefining the Humanities.'" Humanities and the Computer: New Directions. Ed. David Miall. Oxford: Clarendon, 1990, 13-25.

Engwall, Gunnel. Vocabulaire du roman francais (1962-1968): Dictionnaire des frequences. Stockholm: Almqvist, 1984.

Fish, Stanley. "What Is Stylistics and Why Are They Saying Such Terrible Things About It?" Is There A Text In This Class? Cambridge MA: Harvard UP, 1980, 68-96.

Fortier, Paul. "Analysis of Twentieth-Century French Prose Fiction: Theoretical Context, Results, Perspective." Potter. Literary Computing and Literary Criticism. 77-95.

-----. "Theory, Methods and Applications: Some Examples in French Literature." Literary and Linguistic Computing 6 (1991): 192-96.

Frautschi, Richard. "Focal Patterns in Textual Samples from Balzac, Flaubert, Zola and Proust." Literary and Linguistic Computing 4 (1989): 274-80.

Frye, Northrop. "Literary and Mechanical Models." Ed. Susan Hockey and Nancy Ide. Research in Humanities Computing. Selected papers from the ALLC-ACH Conference 1, Oxford: Clarendon, 1991, 3-12.

Greenwood, H. H. "St Paul Revisited--a Computational Result." Literary and Linguistic Computing 7 (1992): 43-47.

Holmes, David. "Vocabulary Richness and the Prophetic Voice." Literary and Linguistic Computing 6 (1991): 259-68.

Herrnstein Smith, Barbara. "Surfacing from the Deep." On the Margins of Discourse: The Relation of Literature to Language. Chicago: U of Chicago P, 1978, 157-201.

Koch, Christian. "Metaphorical and Analogical Understanding in Reader-Text Interaction: A Connectionist (Neural Network) Approach." Research in Humanities Computing. Ed. Susan Hockey and Nancy Ide. Selected papers from the ALLC-ACH Conference 1. Oxford: Clarendon, 1991, 177-203.

Leech, Geoffrey, and Short, Michael. Style in Fiction: A Linguistic Introduction to English Fictional Prose. London: Longman, 1981.

Logan, Harry and Grace. "The Case of the Canterbury Pilgrims: Sentence Semantics and World View in Frag. I of The Canterbury Tales." Literary and Linguistic Computing 5 (1990): 242-47.

Malt, Christian. "The 'New Stylistics': A Success Story or the Story of Successful Self-Deception?" Style 19 (1985): 117-33.

Milic, Louis. "Contra Fish: The Arrogance of Misreading." Style 19 (1985): 385-94.

-----. "Progress in Stylistics: Theory, Statistics, Computers." Computers and the Humanities 25 (1991): 393-400.

Misek-Falkoff, L. D. "Data-Base and Query Systems." Association for Literary and Linguistic Computing Journal 4 (1983): 1-30.

Muller, Charles. Langue francaise, linguistique quantitative, informatique. Geneva: Slat-kine/Champion, 1985.

Nieuwoudt, B. A. "Beyond CATSS: Utilizing Relational Databases for Text-Critical Research." Literary and Linguistic Computing 4 (1989): 254-59.

Olsen, Mark. "Signs, Symbols and Discourses: A New Direction for Computer-Aided Literature Studies." Special no. Computers and the Humanities 27 (1993-94): 309-14.

Potter, Rosanne, ed. Literary Computing and Literary Criticism. Philadelphia: U of Pennsylvania P, 1989.

-----. "Literary Criticism and Literary Computing: The Difficulties of a Synthesis." Computers and the Humanities 22 (1988): 91-97.

-----. "Statistical Analysis of Literature: A Retrospective on Computers and the Humanities, 1966-1990." Computers and the Humanities 25 (1991): 401-29.

Raben, Joseph. "Humanities Computing 25 Years Later." Computing and the Humanities 25 (1991): 341-50.

-----. "The Humanist in the Computer Lab." ALLC Bulletin 1.1 (1973): 3-9.

Robey, David. Report on the 1992 ALLC-ACH conference. Literary and Linguistic Computing 7 (1992): 239-42.

Roper, John. "The New Humanities Workstation." Literary and Linguistic Computing '6 (1991): 131-33.

Ross, Donald, Jr. "Differences, Genres and Influences." Potter. Literary Computing and Literary Criticism. 45-57.

Schanze, Helmut. "A European Perspective for the PC Age." Literary and Linguistic Computing 5 (1990): 171-73.

Smith, John. "Computer Criticism." Potter. Literary Computing and Literary Criticism. 13-43.

Sutherland, Kathryn. "A Guide Through the Labyrinth: Dickens's Little Dorrit as Hypertext." Literary and Linguistic Computing 5 (1990): 305-09.

van Peer, Willie. "Quantitative Studies of Literature. A Critique and an Outlook." Computers and the Humanities 23 (1989): 301-07.

Widmann, R. L. "The Computer in Historical Collation: Use of the IBM 360/75 in Collating Multiple Editions of A Midsummer Night's Dream." The Computer in Literary and Linguistic Research. Ed. Roy Wisbey. Cambridge: Cambridge UP, 1971, 57-63.

~~~~~~~~

By Alison M. Finch, Merton College, Oxford