Monday, July 30, 2018

Linguistics and Textual Criticism

I think we have not seen enough discussion about the uses or importance of linguistics to textual critics of the modern era.  It seems to be generally accepted as essential in Classical and Medieval editing.  There the critical pursuit of scribal error usually lacks the types of corroborating evidence often found in abundance in the letters, autograph manuscripts, and marked proofs relatively common for modern works.  Hence, reliance on one’s thorough acquaintance with linguistic details is essential for the detection of error, or, more important, the detection of legitimate but rare usages.
     Even editors of modern literature, whose primary goal is to represent the texts of extant documents precisely, must feel the need of linguistic sophistication in determining obscure or unusual manuscript readings where ambiguity might either be discovered to exist or possibly be resolved by appeal to linguistic knowledge or principles.
My own interest in linguistics was strictly amateurish and limited, nor would I ever claim to be a linguist or even very knowledgeable about it beyond a graduate course taken before I realized its importance and a magnificent syntax course, experienced after I realized it was important to my work as a textual editor.
     It was my privilege to have had a friend whose interest in linguistics far exceeded mine but who nevertheless let me into his thinking over a forty year period, during which he gnawed and chewed on ordinary language and how meaning was created and expressed and taken up by speakers and writers and listeners and readers.  His thinking influenced my ideas about “script act theory” elaborated in Resisting Texts and From Gutenberg to Google.
     Price Caldwell’s writings on linguistics have been gathered and edited posthumously by Oliver Cresswell and Robert J. Stainton as Discourse Structure and Linguistic Choice: The Theory and Applications of Molecular Sememics (Springer, 2018).   I was privileged to write the forward, which I append here in its pre-publication form.   Blatant self-promotion revealed, can I just suggest that you ask your library to order the book.  (

Peter Shillingsburg
Price Caldwell and I were best friends for forty years, beginning when he moved in across the street and needed help sanding the floors of his house. At the time I fancied myself a good arguer. I had won a high school medal in debating. Price just looked tired when I got in that mode. He was not interested in conducting or winning a competitive argument. Winning a debate, he said, leaves one exactly where one started. Losing, on the other hand, means one has at least had to give up some untenable proposition. He taught me to use argument to discover what I did not know. He was never satisfied with what he thought he knew. He was seldom satisfied with what he had written – hence, I think, this unfinished book. If I were to be his friend, I would need to stop debating and start examining arguments for weaknesses to be fixed and for strengths that were hidden. In conversations with him, if both of us moved off of our opening positions, we both won. I miss him.
            In two ways I represent potential readers of Price’s book. One is as a layman fascinated with linguistics and theory and philosophy, driven by a hope that pursuit of these subjects will lead, if not to truth, at least to an escape from (some) self-deception of the sort that makes one so certain. Certainty is often, if not always, supported in part by one’s ignorance (sometimes willful ignorance) of that which would undermine it. Being undermined in one's certainty is a release from the burden of ignorance, of partial understanding. Price always wanted to know how an argument could hide that which undermined it. Time and again, in this book, he suspects the standard view to be based on some assumption hidden by the rhetoric of the arguments supporting it. His work is an exploration. He is not flogging a fixed position. He was always ready to concede in light of better argument.
            The second way I represent potential readers is my desire in particular to know how meaning works, how language makes meaning, and how people make language, driven in my case by a somewhat esoteric need to understand how revision works in the process of writing. I remember the first time Price tried to explain molecular sememics to me in the cafeteria of the college where we both taught. I was skeptical. Another acquaintance had been bending my ear about a modified version of phlogiston. I was not buying. About an hour and a half into the conversation with Price the penny dropped. I understood the basic notion: that the functional meaning of a word in a sentence as it was used by a person in a situation where speaker and listener both understood nearly everything that went without saying might, just might, be the same meaning to speaker and listener (intender and interpreter), even though the words were not defined so in any dictionary. Yet, the flexibility and instability and dexterity of language was such that there was no guarantee a word would mean the same to both speaker and listener, to both writer and reader. I revisited my college linguistic books, re-read J. L. Austin’s How to Do Things with Words, ate up the writings of John Searle, Quentin Skinner, Steven Pinker, and Paul Hernadi, sat in on a syntax course taught by my colleague John (Haj) Ross, and mulled Price’s ideas, and fancied myself quite the ‘pragmaticist’.
            The truth is I don’t think Price ever believed that I had understood him fully. I did not believe so either. I fancy he wrote this book to help me get there. Over the years he sent me drafts of some of the chapters. Any layman can understand them. He hated big words as much as I did. Big words too often hide things rather than expose or encapsulate ideas. His interests were in how ordinary language works in speech and writing, in propositions, jokes, lies, innuendos, hints, deceptions, inarticulations (how does one know the difference between a doohickey and a thingamabob? how can one say one thing and mean another?) – everything; mine was in how language works in writing, in understanding the writings of inaccessible (dead) authors. Price gave me new tools for studying drafts, manuscripts, revisions, proof alterations, and revised editions.
            I found the concept of molecular sememics invigorating and useful in elaborating a variation on speech act theory, which I called script act theory. There are a variety of important differences between speech and writing, not the least of which is that, in speech, speaker and listener usually occupy the same place and time, while in writing, it is usual for writer and reader to be leagues and ages apart. The contexts of speech that form one important aspect of the molecular sememe, are shaped and modified in crucial ways by what goes without saying – understood to influence meaning, though not explicit or even acknowledged. In writing, with gaps of space and time separating writer from reader, what went without saying ceases to do so. The controlling sememic molecule is eviscerated. In addition, molecular sememics explains how, as composition proceeds, triangulating processes narrow the range of acceptable and expected words at each next point in a sentence, such that each chosen next word means what it means in contrast to the few remaining other possible words that could have been chosen. (Umpires in baseball do not call a runner “in” as opposed to “out”.) When a writer chooses, crosses out, chooses again, crosses out and finally decides, each potentially expected word from that triangulated set of possible choices affects the meaning of the chosen word and simultaneously limits the range of options for the next word. We recognize typos and malapropisms, in part, because they fall outside that set of acceptable / expected words triangulated for us by what went before, and the contexts of the speech act. In script acts the writer has more time and space for experiment, and, yet, the final choice is not necessarily the perfect choice. The crossed out words belong to a limited set and help a reader know how the writer was thinking. Knowledge of each abandoned choice helps us understand the last one more precisely. These ‘Caldwellean’ concepts stand at the core of my books on textual criticism, Resisting Texts: Authority and Submission in Constructions of Meaning and From Gutenberg to Google: Electronic Representations of Literary Texts.Price’s ideas helped me understand the significance of revisions for literary criticism and for scholarly editing. Others will find his ideas intriguing and useful in other ways  
            Twenty-five or more years ago, for example, my enthusiasm for molecular sememics led me to explain the ideas to an artist friend, who at the time was also reading Foucault and Levi Strauss. She painted a triptych in which the first panel was a fairly faithful representation of an academic gown, the second panel was a portrait of a man, and the third a somewhat chaotic combination of the colors used in the first two panels in which a pattern emerged when viewed beside the first two panels. She said, “I got the idea from your friend’s idea about the capsule, you know, the module”. Maybe she got the words wrong but she got the concept clearly enough. Recognition of sameness and difference is crucial to meaning making.
            I do not mean to imply that this book is for laymen only, and yet, it might strike some linguists reading the early chapters that the way Price has nested his ideas in the history of linguistic thought, the ways he has explained his views by contrasting them with alternatives common in the field of linguistics, have left something to be desired. That is, it could be argued that the ‘potted histories’ of linguistics in the early chapters are insufficient to give gravitas to his views. He read more than shows up in the footnotes. He took an NEH sponsored course with George Lakoff. This sense of haste in ‘covering’ existing thought in linguistics is remedied in the second section of ‘the book’ (Chapter 5). I knew Price and the way he thinks. He had a vision of how language works that made more sense to him than did the existing views. His idea of conversation or discourse was that one did one’s best to be clear, and then one listened for the responses. Persuading others that he was right about linguistics was not his main aim. Offering as coherent an argument as he could, he hoped for feedback, even blowback, that took him seriously, genuinely considered his views. At every point, he was willing to concede his position in the face of new information or better argument. Too often, I saw him at times when he felt his ideas were dismissed a priori because he questioned rather than built on received wisdom or because he was deemed an amateur.
            The structure of this posthumous book reflects the processes by which Price thought his way into molecular sememics. Understanding that process will help explain the somewhat unorthodox structure of the book. The first four chapters, being collected for the most part from previous stand-alone articles, consist of four introductions to the subject, each developing some new aspect of it, but each covering some ground covered elsewhere. This is a good thing, because Caldwell’s overall intent is to introduce a new paradigm for understanding how ordinary language works, and that requires stripping away much that many of us have adopted as unquestioned truths about how language works. Chapter 5 consists of five sections of what Caldwell intended to be ‘the book.’ It is in some ways more formal, more methodical than the first four chapters, and because it represents his attempt to organize and present the idea of molecular sememics as a whole, it takes up in a formal organized way the ideas he has been introducing us to in the first chapters. No doubt, had he lived to complete ‘the book’ it would have stood more securely on its own without the essays he produced as he was working out the details of molecular sememics. Although some ideas are repeated more often than required in a single publication, I find the result of these multiple approaches and re-explanations very useful in making the ideas understandable.
            Though for years Price and I talked endlessly about MS and molecule-selection-and-execution structures, I did not know the full range of his thinking in the deep and interesting ways that show up in Section 4 of Chapter 5, “Qualities of the Sememe”— his clearest expression of the difference between MS as an arena in which meaning gets expressed in a dynamic exercise of rhetorical skill (even by the nearly inarticulate), on one hand, and the structuralist, formalist view of language as a pre-existing set of categorical choices offered by the Langue, on the other. Molecular sememics explains ordinary exchanges of intention and interpretation (speaker/listener) as skill in innovation of expression rather than as dexterity in the application of rules. And though he wrote it first many years ago, I did not know how well he applied his sense of how language works to the unpacking of nuance in Wallace Stevens’ poems, in Section II. Anyone unaware of MS before reading Caldwell on Stevens might be none the wiser about MS but surely would be wiser about Steven’s exploration of the relation between language and one’s sense of the ‘real’ world. But, having read this book, one can see MS lurking in the background of Caldwell’s analyses of Stevens, Hemingway, and Ford. The difference between applying with dexterity the categorical options of structured language by contrast to the momentary dynamics of constructing meaning is echoed in Caldwell’s comment on Richard Ford’s writing, that “Every writer should distrust the received meanings of words and concepts. He or she should take on the obligation to make meaning...”
            In the last fifty years or so, Samuel Beckett’s question about authorship and voice in literature: What does it matter who is speaking?, has been reiterated by Foucault, Derrida, and numerous literary critics. One of the reasons for this attitude toward voice is that interpretive despair and the absurdity of modern life have made the question uninteresting. Another is that when the text is read, it is the reader who is speaking, making up the tone and feeling about what is written. Caldwell focuses on how readers make up the tone and feeling of what they read. He begins with explorations of how writers and speakers create meaning using not just words but tone and feeling and context and expectations – providing meaning-determiners that sensitive listeners and readers can use to ferret out intended meaning, or at least to avoid the literalness of tone deaf readings. Reading this book brought home to me more clearly than ever that molecular sememics, as a theory about practical meaning making, is an enormously useful tool for well-informed reading. In short, Price gives us grounds for thinking that ‘who is speaking’ does indeed matter.
            While he is no longer with us to receive feedback, Price Caldwell would have been delighted to know that his ideas entered or provoked a conversation, a reconsideration of mainstream thinking on whether competence precedes or follows performance, whether language precedes or derives from speech, whether the rules of a language system predict or circumscribe language acts or if local speech, through repetitions and incremental conventionalization first creates and then modifies the rules, whether the rules of syntax determine word order or if the salience order of discourse creates syntactic paths that become conventionalized into rules. His primary insight, from my point of view, was the dynamics of meaning creation or construction, for dynamic meaning making, whether primitive or sophisticated, provides a plausible explanation for the innovative nuances that speakers and writers accomplish on the fly; as well as for the evolution of language competencies; and for how rules develop and how they can be broken to good effect. Beyond that, when he says, “meaning belongs to the molecule, and not to the word”, he opens a path for literary critics to seek historical meanings in writings as well as exploring the limits of misprision. Using what we know about the contexts of origination, the audience, the cultural expectations of the time and noting the actually trialed but rejected words in drafts, manuscripts, and revision, a critic can often come close to recognizing the molecular sememe that determined meaning for the writer. That one cannot always do that and that one cannot do so with certainty was not, for Caldwell, a reason to abandon hope of approximating intended meanings. His conclusion (that “there is as much good reason in examining the readers’ assumptions – as reader-response criticism insists – as there is in searching for historical evidence of the writer’s intentions”), is not startling. What is new is his methods for understanding how we go about that business.
            The important issue is not whether one or the other of the possible explanations for how language works is right, but how they can help us understand meaning and meaning making. As Caldwell demonstrates and as is well known, many grammatically and syntactically ‘correct’ sentences make no sense – are meaningless. Discourse salience in molecular sememics is like an owners’ manual for deploying the tools of language effectively, not just acceptably. That there are other explanations, Price obviously knew. That his ideas gave him at least a temporary sense of satisfaction and clarity of sight is also apparent. His ideas arose from the sense of dissatisfaction with the standard explanations. He was less concerned with whether he was right than he was with whether his thinking about language, running as it does against the grain in so many ways, could stimulate further thought and analysis. He would be happy to know that he had nudged the conversation forward, even if just a bit.

Wednesday, March 7, 2018

Woodcraft and Textual Studies, Pt II:

Flaws and Errors vs Character of the Material

        I recently acquired an oak burl that was riddled with carpenter ants and some rot.  Burl wood is the result of disease or damage.  The tree attempts to repair the damage and isolate the disease or invading fungus or insect.  The result is a sort of gall of wood fiber without a set grain direction.  The patterning can be beautiful, but the juxtaposition of rot and insect damage with solid burl wood is erratic.  One could call it natural materiality, if one were aping aca-talk.
       I cut pen blanks from my burl.  A pen blank is about three inches long and 3/4inches square.  One drills a hole through the blank, inserts and glues a thin metal tube, mounts it on a lathe with appropriate bushings, and shapes a pen body.  What could go wrong?
       Well, first I ran into some soft sap wood that flecked off, leaving bits of exposed metal tubing.   In another effort I ran into the channels left by the carpenter ants.  And in another effort I gouged the wood by holding the lathe tool at the wrong angle.  It got me thinking about the differences between flaws in the material, errors of execution, and discovery of the character of the wood.  Woodworkers often casually refer to "character of the wood" to render beautiful that which appears anomalous.  Knots are the chief such objects.  In some woods the knots are surrounded by grain that curls and chips during planing, leaving dips and small divots chipped out of the wood.  Sanding might remove evidence of very shallow divots, and a smooth dip might be said to follow the grain.  Here is where a fine line exists between flaw and error.  Did the carpenter mar the wood by breaking fibers or was the carpenter exposing and displaying the subtle undulations of the wood?
       It made me think of the materiality of texts and the carpentry of editing.  It might make more sense to compare carpenters to writers, but writers create material objects from words, where the material is merely a medium, not the thing itself (whatever that is).  Editors, like carpenters, find the material already in existence and seek to find how it should best express itself.  Understanding the character of the material makes it possible to work with the material to expose and display its subtlety.   Some carpenters, like some editors, are not interested in the material.  Their eye is on a conventional outcome: a text without flaws; a chair or a desk that looks exactly like its mates and in which no undulations, subtle or otherwise, can be detected.   
       My pen blank with soft sapwood that flecked off the lathe declared itself unfit for the use I sought.  The ant channels offered a different problem.  They were not, per se, character of wood.  Yet, they were endemic to my particular piece of wood.  Did their existence render my pen blank inadequate for the job?  Or did they become a natural aspect to be appreciated and cared for in the pen-making process?   Is care, devoted to such material, a vain attempt to salvage inadequate material, or is it an homage to the material?
       Answers to those questions vary according to the expectations and inclinations of the person setting up as judge.  In editing, if one's expectations, one's predispositions and assumptions, about the material are strong, one can run roughshod over the subtleties.  If one's principles are strong, they might prevent one from seeing how the material resists the editor's tool.
       It has become common these days to lump Anglo-American editorial practice into a catch-all bin called intentionalism or idealism in order to contrast it (usually pejoratively) with European editorial practice in a bin called materialism.  The former is said to drive toward an unachievable idealism, while the later pursues and cares for material text(s).  Both bins are damaging oversimplifications: damaging because anyone buying into such stereotypes is encouraged to become inured to the differences between flaws and errors and between character of the material and character of the editorial principle. 
       It is no defense of realism or scholarship to say that the intentional character of material texts is outside the purview of editing.  Likewise, it is not defense of idealism or criticism to say one can treat documents as mere temporal snapshots of intentional works in progress.  Both statements contain truth and are thus more than a little bit attractive, but a true wood worker respects the character of wood and rejects the flaws.  To do otherwise is error.

Wednesday, June 14, 2017

Textuality and Knowledge: the book

Update:  $44.95 paperback issued June 2018

See Order Form and Discount, below.
         In a time when the distinctions between fact and opinion have become blurred, I thought it would be good to review what we think we know about knowledge--about what we think passes as knowledge.   I think about the differences between knowledge, on one hand and opinion, speculation, guesswork, and wishful thinking, on the other.  I think about why the "fact" that uncertainty dogs every proposition is not an excuse for thinking that "uninformed" opinions are just as valid as "informed opinions."   I think about the need to validate arguments by showing sources of evidence and revealing the sources of the sources.   This book is about textual knowledge; knowledge of texts.  It is not philosophy; it is practice.
     Textuality and Knowledge
39.11  1990  >  1990s
96.4  authorial,  >  authorial
96.5  intention-clear  >  intention clear
NB: the sense of that sentence ending would have been clearer had I written:   “. . . aims of producing clear texts representing final authorial intentions.”   Rather than what I did write:  “. . . aims of producing final authorial intention clear texts.”   
100.9-10  McGann’s applications to the editing of McKenzie’s bibliographical observations.   > McGann’s applications to editing of McKenzie’s bibliographical observations.
NB: the sense might have been clearer had I written: “. . . McGann’s applications of McKenzie’s bibliographical observations to editing.”  

Textuality and Knowledge: Essays is now available from Penn State University Press!

If you think that this book should be part of your library’s collection, please forward the information to the library.

Textuality and Knowledge


Peter Shillingsburg
In literary investigation all evidence is textual, dependent on preservation in material copies. Copies, however, are vulnerable to inadvertent and purposeful change. In this volume, Peter Shillingsburg explores the implications of this central concept of textual scholarship.
Through thirteen essays, Shillingsburg argues that literary study depends on documents, the preservation of works, and textual replication, and he traces how this proposition affects understanding. He explains the consequences of textual knowledge (and ignorance) in teaching, reading, and research—and in the generous impulses behind the digitization of cultural documents. He also examines the ways in which facile assumptions about a text can lead one astray, discusses how differing international and cultural understandings of the importance of documents and their preservation shape both knowledge about and replication of works, and assesses the dissemination of information in the context of ethics and social justice. In bringing these wide-ranging pieces together, Shillingsburg reveals how and why meaning changes with each successive rendering of a work, the value in viewing each subsequent copy of a text as an original entity, and the relationship between textuality and knowledge.
Featuring case studies throughout, this erudite collection distills decades of Shillingsburg’s thought on literary history and criticism and appraises the place of textual studies and scholarly editing today.

Please send me Textuality and Knowledge
_______ cloth @ $80.50 subtotal: _________
U.S. (S&H)
$8 first book, $4 each add’l: _________
Canada/Mexico (S&H)
$12 first book, $6 each add’l: _________
Outside North America (S&H)
$25 first book, $10 each add’l: _________
PA residents, add 6% sales tax;
Canada residents, add 5% GST: _________
Total: _________
Payment: Check* Money order MasterCard VISA Discover American Express
Card# Exp. Date___________
*Make checks payable to Penn State University.
Please mention the PLS17 code when ordering by phone.

Friday, March 3, 2017

"Agency" and "Authority"
   In discussions of textual criticism, the word "agency" often is used to answer the question: Who wrote this or who edited this or who made this change in the text.  The answers often are: the author or the publisher or the censor or, perhaps too often, I don't know. Answers to these questions supposedly lead to understanding why the changes were made and whether the changes were authorized or have authority.   
   Authority is a word with a variety of connotations.  If it means having the power to control how a reader understands the text, the reaction of many if not most readers has often been, Oh, yeah, watch me.  But if it means, exercising the rights of an author to have the words and punctuation of a text be what he or she wishes them to be, the word authority suggests instead that the change is not spurious.  Literary critics since the 50s have rebelled against the authority of authors to determine what they meant by a text.  They have rebelled against the notion that an expert critic can determine the meaning of a text to be the author's meaning and therefore the meaning that is to be taught to students.  The author is dead as the authority over he meaning of the text.  The author, at best, is just a function that initiates a discourse.  The author is a convenient fiction of the reader's imagination.  Authority, in that realm of meaning is abhorrent.  For critics, the word authority means what it means by contrast to liberty, freedom, flexibility, imagination.  Authority is an "anti-" term.
   That attitude has troubled textual critics whenever they forget or fail to see that their own interest in the authority of a textual change, in the author of a sequence of words, and in the authentication of sources of texts has nothing to do with the authority abhorred by critics.  Instead, for textual critics the word authority means what it means by contrast to spurious, fake, inadvertent, error, mistake, or interference by some interloper.  The contrast between the literary critic's objection to authority or authorities and the textual critics preference for authoritative or authentic texts is sharply seen in censorship.  Censorship is the exercise of the kind of authority literary critics hate.  Censorship is also the interference and introduction of unauthorized work that textual critics object to.
   In literary criticism, agency has had at least one additional meaning and function, which also has a bearing on the questions textual critics ask.   The question of agency in literary criticism is often connected to epistemology, how do we know, or is it possible to know, and who or what is it that is doing the knowing.   This is a double problem: our human distance from the objects of our attention, a gap that makes it impossible to be objective about objects, on one hand, and our human inability to pin down who it is that knows what we think we know, a problem of one's self as subject.  The question of agency is thus limited by the uncertainty of knowledge: both the knower and the  (supposedly) known.  Thus, even if one answers the textual critic's question, who wrote this, by saying, the author wrote this, the literary critic's response is likely to be two-fold.  First, what is an author or who is the author or how can you know what is meant by the author, on one hand, and secondly, who are you who claims that you know the author wrote this.  And having answered that you do not know for certain, the critic might wish to know why you tried to ask and answer the question in the first place.  That is, if you were trying to pin down agency in an effort to understand the author's meaning, you would have to contend with a series of questions about agency that undermine it as a means to that end.  In short, regardless of how you define authority, they will have none of it.
   Put a slightly different way, uncertainty about how one knows anything and uncertainty about one's own base of knowledge are together a definition of the human condition.  All social interactions, indeed all interactions that anyone experiences are rendered uncertain by these observations about agency.  And there is (as yet?) no way round it.  Many, maybe most, people proceed unbothered by this fundamental uncertainty.  Either they proceed unaware that there is a problem because, pragmatically, the uncertainty factor is small enough not to be noticed.  They believe they understand for the most part what was said to them and they believe for the most part what they say is understood by their audience.  Perhaps with less force, this contentment with some slippage also applies to writings, letters, laws, contracts, and even poetry and fiction--which are generally successful in being "taken for what they say."  So the problem, if it is a problem, is not noticed by many and felt to be negligible by most. They do not perceive it to be a significant problem.   Or, others proceed fully aware of the problem of uncertainty and fully aware that there is no way round it, but manage not to go crazy because they also know that it is a condition shared in common by all, and that life must go on as best it can. 
    For this second group, those who have learned to live with uncertainty, there are also two sub-groups: those, who having accepted the uncertainty of life, decide that any choice they make is as good as any other choice and that success is measured, not by truth, integrity, or right and wrong, but by dominance--survival of the fittest or might is right.  The other subgroup, having accepted the uncertainty of life, choose to proceed tentatively, always ready to concede that new evidence or a different point of view might have value.  The first sub-group is capable of cruelty--where there is no certainty, they merely have to assert their own certainty and impose it on others; the second sub-group is incapable of cruelty--where there is no certainty, one must always be ready to admit one's own limitations and self-deceptions.  That is to say, although recognition of the fundamental uncertainty at the base of all knowledge is a significant problem, it is rather  like gravity, neutral and universal in its effects and does not give grounds for any new kind of behavior.  The discovery that everything that we had taken to be at least relatively determinable and understandable is in fact always up for question, does not give us grounds for saying that nothing means anything determinable or that everything can have any meaning we are able to ascribe to it.
    Nevertheless, in textual criticism, acknowledgement of uncertainty about agency entails three areas of uncertainty.  First, evidence is often insufficient to establish who (which agent) was responsible for the text found in a document.  Second, it is not determinable with certainty what was meant by the agent of change. And third, it is not certain that you (or the reader) has taken up the words and their meaning in a the same way that motivated the writing.
   It is important to understand that the literary critic's revolt against the hegemony of what has been called author meaning is not a rebellion that textual critics need object to.  Textual critics can also object that the imposition of author meaning is or can be authoritarian in the worst meaning of that word.  Textual critics are not interested in author meaning as a goal of their work to be imposed on all readers.  They are not interested in restricting the liberty of readers to do whatever they wish to do with text in whatever way they want to do it.  
   But, it is telling to note that some literary critics, embracing the freedom to do what they like with their texts, have ceased to ask, what might the originator(s) of this text have meant by their text?  They have been so keen on establishing their freedom to ignore the author meaning that they seem to have forgotten how to seek author meaning out as one of their options.  The textual critic, more than any other branch of literary investigation, is devoted to providing the tools to increase access to author meaning as one of the options readers have.  Understanding which text of a wok one it reading, knowing where the text came from and who prepared it and how it was prepared, knowing how that text differs from other texts of the same work, and knowing the contexts surrounding its creation and the audience(s) to which it was directed--all these are crucial to the reconstruction of what could have been author meaning(s) and the elimination of meanings that could not have been author meaning.  The point is that it takes work and discipline to make author meaning even plausibly available.  It is not a question of imposing the correct author meaning; it is a matter of making author meanings available as one of the reader's options.
   Finally, it is good to remember that when textual critics speak of authorial intention, they are not speaking about author meaning; they are speaking about the choices authors make for the text: the choice of words and punctuation may be the most obvious, but they include the choices of paper and writing instruments, the choice of publishers, the choice of audiences.  What does the surviving evidence indicate were the author's choices.  We can argue about what was meant by these choices after we have determined who made the choices.  That is why agency matters to textual critics.

Thursday, December 15, 2016

Disinformation and False Facts
     The rules of evidence require that demonstrations of the plausibility of opinions stand or fall according to cogent arguments based on the facticity, comprehensiveness, and relevance of demonstrable evidence.  Persuasion cannot be "by any means."  It must follow rules, that over centuries, have been devised to protect us from falsehoods and deceptions.  I do not mean, specifically, the "Federal Rules of Evidence" that apply in courts in the United States, but rather the rules of evidence that govern rational thought, that avoid fallacies, the support our best efforts as human beings to seek truth and shun error--forlorn and imperfect as those efforts might, in practice, prove to be.
     Recent events in the political history of the United States suggest that there is no necessary link between evidence and argument in general matters of life--faith and practice.  Perceptions, regardless of source or support, appear to be a sufficient base for beliefs and actions.  Of course, that was always the case, even when evidence was sought, examined and tested, and even when arguments were probed and questioned before leading to accepted conclusions.  All rational conclusions, as well as all prejudicial, superstitious, or otherwise biased conclusions, are based on perceptions.  Humans seem not to have any absolute way around the uncertainties of knowledge or the frailties of perception.   But are all perceptions of equal value?
     The difference between the ideals of deliberate, evidence-based argument, on one side, and bias, superstition and prejudice, on the other, has not been that evidence and argument leads inevitably to truth, but rather that evidence and rational argument have proven to be the best detectors of error.  One does not prove truth; one exposes error.
     Whenever it has been posited that religion and authority and dogma are the best guarantors of truth, it has been rational thought and investigation of evidence that has exposed the errors and won the day--at least for those who value evidence and argument over blind faith, submission, or indulgence in the self-deceptions arising from what Francis Bacon called the "Four Idols:" the tribe, the cave, the marketplace, and the theatre.  (Very much worth a re-read, btw.)
     Yet, the success of misinformation, false facts, deliberate bias, hyperbolic claims, and blind hope in the manipulation of public opinion suggests that in matters of faith and practice, that is, in matters of ordinary day-to-day life, we live in a place and time that does not require or use evidence and argument or the rules of evidence to sort out the way we live now.
     When I retired in 2013 I thought I should wrap up my scholarly life and thinking by collecting and honing what I thought to be the best papers of my last decade of work.  The resulting collection, I saw, focused on the relationship between critical understanding of literature and the soundness of the evidence upon which criticism had play.  The collection, titled "Textuality and Knowledge," explores the implication of the facts that in literature, evidence is all textual, texts are all documentary, and no two texts or documents are exactly alike, either bibliographically or lexically.  Sometimes the differences in the appearance or in the texts of documents purporting to represent the same work of literature are significant--are capable of stimulating different or even contradictory responses.  It would seem to follow logically that any critical argument based on a documentary text would apply, well or ill, to the text in that document and might not apply equally well or ill to the text of a different document bearing the same title.
     Textuality and Knowledge will be published by Penn State University Press in mid 2017.  The writing was completed before the presidential campaigns for the 2016 election and before that campaign made us all painfully aware, first, that we could not trust the sources of our news information, and second, that we tended to trust those stories that corresponded with what we already wanted to believe--even when those stories were proven false.  I think that was unfortunately too true across the entire political spectrum.  
     In our culture, our DNA, and our basic education we lack the habit of questioning sources, doing research, and finding out if a source is reliable in its use of research and facts.  The problem may start as early as the day a parent or teacher says, "Because I said so."  This appeal to authority is meant to squash the questioning spirit--usually on behalf of some form of desirable order or peace.  But the long-range effect may to be to quash curiosity and critical thinking, except, of course, in the rebellious few.
     So, in spite of a life of devotion to the idea that "the unexamined life is not worth living" and "every conclusion should be based on sound, comprehensive, relevant evidence and sound argument," I question whether that attitude is still important for the way we live now?  So many people seem not to care.  So many opinions hold sway without visible support.  All that seems to matter is the number of people who hold similar opinions based on similar perceptions, regardless of facticity or argument.
     Or is that just the expression of a local, temporary despair?  Somehow, I am not tempted to exclaim, "Oh, brave new world."  My eyes seem not to have been opened by a new truth.  They seem instead to have been clouded by a thicker than usual fog.
     Yet, the metaphors of "eyes newly opened" or "eyes clouded by fog" are rhetorical flourishes.   Am I adding to the disinformation about our place and our times by expressing disapproval of the way persuasion works?  I'm reminded, by his death this week, of Robert Scholes call several decades ago for a shift of focus in English Departments from high art, dissociated from daily life, to the practicalities of survival in a language-based culture.  English departments should teach "Textual Self-Defense"--students should learn enough about how language distorts and manipulates our culture so that they can see through advertisements, political speeches, and sermons and to protect themselves against what Vance Packard called "Hidden Persuaders."
     Careful attendance to sources, accuracy, and deliberate argumentation does not serve as a guide to truth but, rather, as protection against falsehood.   Nothing can guarantee truth.  Plausibility and persuasion is what we go on.  Investigation is always questioning, and trying to falsify propositions.  We, as thoughtful, questioning, rational beings hold tentatively to the probable truth of those proposition which have as yet not been debunked by argument and evidence.   There is much that has not been debunked--many "eternal verities"--that form the ground on which we walk.  Verities, despite the etymology of the word, cannot be verified.  Many have stood the test of critics, debunkers, rebels, and nay-sayers for centuries.  We feel that they are verities.  We live our lives by them. We are ready to give them up when evidence and argument warrant it.
     But in matters that can be verified, please, can we exercise caution and verify, to the best of our abilities before we spread rumors, gossip, or anything that comes to us through the transom that is the internet.   Let there be no more cases of, "I have not had time to see if this is true or not, but here it is."  Stop it!   Let there be no more, "I stand by what I said, regardless of the fact that I have no proof--or regardless of the fact that I made no effort whatsoever to determine if it was true."
     May peace and a clear head be with you, always.

Monday, October 10, 2016

Enduring Distinctions in Textual Studies
[Address to the European Society for Textual Scholarship meeting at the University of Antwerp, October 5, 2016, on the occasion of receiving a prize for my work in textual studies.]                       

     I was asked to look back in order to look ahead--to offer a brief retrospective on textual studies.  That could take a while since I have been a part of ESTS from its inception in 2001 with Peter Robinson and because I have been in textual studies since 1966.  I first confronted European textual studies seriously in 1999, when Dick Van Vliet invited me to a conference of German textual scholars in The Hague.   It was an eye-opening experience, but one I did not immediately understand.  The English language is spoken on both sides of the Atlantic but we do not always understand the same thing by the words that we use.  Words like authority and mixed authority, scholarly edition, emendation, copy-text, and even apparatus all had different meanings for me before ESTS because the historical contexts of textual studies in our different cultural traditions introduced unfamiliar values and nuances.
      The passage from Past to Present and into the Future may be punctuated with new insights, but there is continuation as well.  The enduring theoretical concepts of textual criticism and scholarly editing include certain basic distinctions and sensitivity to relations between the concepts so distinguished.  One such distinction is that between the material document and the texts inscribed on them.  We can treat the document as a unity, where the material and the symbolic are an undivided whole, or we can treat the inscribed text as a recipe that can be endlessly replicated in other documents.  We have come to label these two ways of looking at texts in documents as the distinction between the bibliographic code, which is documentary, and the lexical code, which is symbolic.  Despite all the ink that has been devoted to this topic, there may well need to be more.
     Another distinction depends to some extent on how one sees the first one.  It is between the material, visible text as found in a document and  the verbal conceptual content or experience of the literary work which the text represents for, or stimulates in, readers.  Documents, by themselves, apart from how they represent literary works, are of sufficient complexity and provide textual scholars with sufficient difficulties that we could spend all of our time with them without lifting our eyes from the material pages to the conceptual works and mental experiences enabled by the books.  I doubt, however, that many of us ever forget that, as Emily Dickinson said, there is no frigate like a book to carry us away from our present time and place.  Like John Keats, we have boarded those frigates and "travell'd in the realms of gold, / And many goodly states and kingdoms seen".  With Keats we have all looked into Chapman's Homer and seen islands.  We do that by seeing through the text to the world created by the imagination.  Without that, we probably would not be discussing bibliography, textual criticism, or editorial theory.    As textual critics, however, we look at the books, as well as through them.   The relationships between literary works and the reader's experience of the them are dependent upon the material texts.  We know that different readers construct or experience the same text differently; we also know that, when two material texts differ bibliographically or lexically in significant ways, they influence the reading experience in different ways.   It is important not to lose sight of the literary work, the aesthetic object and our experience of it, while examining the literary documents, just as it is important not to skim unheedingly over and through the literary documents on our way to the literary work.  
    Some people love simplicity more than they love clarity of understanding, and so there have been a number of simplifying generalizations made about the relation between documents and works, but simplifications grasp one truth about the textual condition at the expense of others.  The close observer of texts in documents and texts of works is forced to abandon simplifications.  Just two common examples will clarify.  It has been said that each document is the work because without the document we would not have the work.  This simple equation of work and document seems to require, however, that each different document be a different work.  If one does not want to say that every copy of a work is a different work, then one must not say that the document and the work form an inseparable unity.  If, on the other hand, one says a single work is represented differently by the variant texts in different documents, it seems necessary to also hold that one cannot apprehend the work as a whole without somehow holding its variant iterations in mind.   These difficulties point to textual complexities that resist simplification.   The way one conceives of the relationship between documents and works influences the practice one follows when editing a work; it is, therefore, important to have a sense of the complexity of that relationship.   Dirk Van Hulle and I collaborated on an analysis of this point in an article called "Orientations to Text Revisited" published earlier this year in Studies in Bibliography.   I recommend the parts Dirk wrote.
           A third distinction in my list is so obvious we have all internalized it and may not think it worth mentioning.  It is the distinction between methods and goals.  My first encounters with European, particularly German Historical Critical editing confused me because the general tendency of European editorial practice was different from that in which I had been trained.  Each side thought we wanted the same goal, so, of course, the other side’s methods must be wrong.  Turns out the goals were different, too.  The European model emphasized assembling a record of a work’s historical forms, providing an orderly representation of textual history by combining a clear text, accurately representing one historical document's text, with an apparatus that codified accurately the texts found in other historical documents.  This model has as its goal the compression between the covers of one volume of the historical textual information from the archive of documents representing a work.      By contrast, Anglo-American editors--at least those who followed the principles set forth by Greg, Bowers, and Tanselle--conducted all the research into the archival forms of the work, compiled all the data for the historical apparatus, and attached that information to a critically edited text.  Europeans accused Americans of contaminating the historical record; Americans accused Europeans of stopping before the real work of editing was begun.  In their own way, each was right, but for the most part many of us failed to understand why the other side was editing in a way we thought was wrong.
       It strikes me now, that while on the surface, the Historical /Critical approach appeared to Americans as narrow, rigid, and unimaginative, it was in fact fundamentally liberal and expansive because it presented its strictly factual product as a basis, not only for understanding the history of the texts of a work, but for new critical imaginative editing without trying to dictate what new editorial goals should be.   By contrast, Scholarly Critical Editing in America, while on the surface seeming to be open and imaginative, offering to do the careful detailed textual work on behalf of other readers who could then focus on the Work without getting bogged down in the textual minutiae of the history of documents, was in fact a bit tyrannical, for it offered texts as accomplished, established facts, saying "This is the text you should use; the other data is of historical interest only."  I think that Historical / Critical Editing and Scholarly Critical Editing both have honorable places in textual criticism.  Their methods are different because their goals are different.  No editorial method is the one correct approach for all editorial goals.  An editor should not assume that his or her method will fulfill every goal, nor should editors assume that their particular goal in editing is the goal everyone should have.   The test of a good edition is whether its methods actually fit its goal.  And, also very important, though seldom mentioned in editorial scholarship, readers need to learn to use scholarly editions rather than assuming that they are all alike and work in the same way.
     Another distinction, that between original material documents and representations of them, displayed itself in print in the distinction between new editions (especially scholarly editions) and facsimiles.  New print scholarly editions claim to be only the lexical equivalent of original editions.  Facsimile editions provide a simulacrum of the bibliographic code by imitating the physical aspects of original editions.  
    The digital age has both muddied and clarified our sense of this important distinction, first between originals and reproductions, and, second, between texts and images.  Digitally, we can retype or re-key texts for representation or we can take scans or photographs, digitizing images of documents for representation.   It is an ignorant assumption that the symbols of a text can be represented in any font or any medium and still represent the same work without significant loss.  Nevertheless, digitally, it is often the case that a retyped text, marked up and tagged so that new forms of analysis and representation can become functional, is presented as if it were a sufficient representation of the historical text from which it derives.  We have all seen 21st century transcriptions on line labeled as if they were the 19th century originals from which they purport to derive.  Bibliographers call a re-keyed text a new edition.  It may represent the same work but it does not do so in the same way that the original document represented it, to say nothing of the fact that every keystroke is an opportunity for textual error.   We can also now, far more easily and accurately, create digital images of the original, including pictures of the document the text is printed or written on.  In digital representations of original documentary texts, the distinction between (the lexical) text and (the bibliographic) image is stark because it requires different files, thus, clarifying a distinction that has always been there.  
    For large modern digital projects we have, once again, to acknowledge two truths:  No one person knows all of our periods and areas of specialization, and everyone needs to know more than one area of specialization or have access to someone who will help us.  Despite our long tradition of literary and textual scholars working alone in small carrels in libraries, our projects now require team efforts.  I wish finally to distinguish team models.  In one model the chief of the project says to his helpers, Do this; do that.  The helper may be a graduate research assistant, a secretary or clerk, a librarian, or a computer technologist or programmer.  Like a great secret, the chief sits in the middle and knows.  Except that mostly the chief in the middle does not know--does not know how to do what the helpers know how to do and does not know if the helpers have actually done it the best way or an acceptable way or have just covered up not doing it at all or not well.   This hierarchical model is well known to those of us who have been the helpers or have graduated into the chief position.  I believe the hierarchical model of project conduct is counterproductive, limited, and I hope doomed to extinction.  Another team model involves a group of persons with similar and overlapping interests who conceive of a project and lay out a system of collaboration.  The tasks are various and should go to those best suited for the task.   Being a good fundraiser makes you important but does not make you a chief.  Other tasks focus on bibliography; materials collection; compilation of analytical data; analysis of data; elaboration of textual principles; organization of the work flow; and selection of existing software or development of improved software--which itself is complex, involving tools, data storage and retrieval, interfaces, navigation, and exit and portability strategies.  When the relationship among the people who undertake these tasks is that of partners in an enterprise, the project becomes not only the fulfillment of an initial concept but the development of that concept in ways the chief initiator did not know and could not imagine.   Unfortunately, a project cannot be better than its weakest contributor.   But tell me, who amongst you is yourself the best fundraiser, bibliographer, librarian, codicologist, book historian, editor, and software developer that you know--you alone, all in one?   And if you are not that, do you imagine you will do best by hiring underlings to do your bidding or by finding collaborative partners who can stretch your capabilities?   Of course you should vet your partners, just as they should vet you.  Partnerships like chains are only as strong as their weakest link.   You should strive to be that weak link--which is to say, you should strive to partner only with those who are better than yourself.  It is amazing how good that will make you look.
      In a moment of weakness, and although he should know better, Dirk asked me to look into the future.   In my own future I see fishing, woodworking, travel for pleasure, and the superintendence of a growing array of grandchildren.  In your future I see your tasks and your accomplishments through a lens that reveals that knowledge is not knowledge if it is not verified; that in editing, the facts are all documentary.  If you do not have the original documents, you cannot be absolutely sure of your facts.  I see that methods of editing are not facts; instead they are ways to organize and present facts.   I see that editions are arguments about the facts and are susceptible to the same faults and shortcomings that attend all critical arguments.  I see that you will be tempted from time to time to believe that your discoveries, your methods, and your arguments are the best in the world and that you should tell others what to do and how to do it because, like the secret, you are sitting in the middle and know.  
     But let me leave you with a quotation from Richard Flanagan's magical novel called Gould's Book of Fish, in which the protagonist says:
To be frank, though I have painted all I know, it's clear that what I know is two parts of bugger-all.  All that I don't know, on the other hand, is truly impressive & the library of Alexandria would be too small to contain the details of my ignorance.
    In your future, cultivate productively the bits that you do know; and try to understand what others do before dismissing them or criticizing them.  The world is a big place with room for many truths, but is too small I think for error, for unsupported argument, and for attempts to make everyone see and do the same thing the same way.  We just don't know enough.

Tuesday, September 27, 2016

Confronting the Archive: Seven Questions


(Conference paper delivered at Univ. of Pittsburgh, March 2016.  Many of the illustrations are well-known already; some I have used in other presentations.  Together here, I hope to illustrate my comments on the functions of archives.)

A key distinction needed for my argument is not always made by students of literature today.  It distinguishes works (which are verbal mind constructs created in acts of reading) from books (which are material objects created by authors, publishers, and printers). Literary critics confront literary works (verbal mind abstractions) by looking through the printed page.  Textual critics confront archival books (textual material objects) by looking at the printed page.  The physical copy of the book, which the critic holds, represents the literary work, verbal art, created in the reader's mind as a reading experience.  It is interesting to note that the book the critic holds is just a copy; it is not the original nor "the thing itself".  It might be interesting to ask, "From which previous copy was the present copy copied?"   That is different from asking, "Of what is my copy a copy?"  

      If I hold the Norton Critical Edition of Moby-Dick edited by Hayford and Parker, the answer to the first question (what was my copy copied from?) would be "It is a slightly altered copy of the Northwestern-Newberry Edition, edited by Hayford, Parker, and Tanselle, which itself was copied eclectically from a combination of  [Slide] the first American and first British editions, with additional editorial emendations thrown in."  

The answer to the second question (what is my Norton Critical Edition a copy of?) would be, "This is a copy of the novel Moby-Dick by Herman Melville." (I'm sorry, there is no image of the novel itself , whatever that is when it is experienced in the minds of readers.)  It is rather important to distinguish, therefore, the work as a verbal art experienced in reading from its representation in any given physical copy.  This distinction is subtle and problematic when one observes that the manuscript of Moby-Dick was (it no longer exists)--was just the first copy of the work.  What are we to do with the facts that, first, every copy of the work differs from every other copy? and, second, no copy of the work is the work itself? and, third, every copy of the work represents the work in some sense?  The first step in sorting out this tangle is, I think, to say that a literary work is a cognitive construct, a mental construct or abstraction created by reading, while the book is a physical object.  The science/ art of exploring the relationship between the physical book and the verbal work is called textual criticism.  Literary critics who prefer soft swampy fields of play or foundations in shifting sand do well to avoid textual criticism.
There is a sense, however, in which we are all both literary and textual critics, and we are best when we are both.  The more I know about each field the more their supporting evidence, methods, and goals invade each other's territory.  I've come to think that separating literary from textual criticism is murdering to dissect.  Yet, it is common for critics and students to treat the copy in hand as if it were the work, untroubled by the evidence from the archive.    
What are the questions that literary / textual critics can ask when confronted by the archive.  (Wouldn't it be nice if  the archive came to us and confronted us?)  I have made a list of seven questions from FAQs to SAQs (Frequently asked to Seldom asked).  As for which questions are the most important or the most interesting--each of us is likely to choose a different order for the list.    
Probably the most frequently asked question when confronted by the fact that early copies of a work (indeed all copies) differ in wording and punctuation is:
1.  Which is correct?   The assumption is that one reading is the correct one and all the rest are errors.   Sometimes this is so.  This question reflects a simple desire:  Give me the right text please, and don't bother me with the boring details of textual comma counting.  Of course, sometimes nobody knows which is right or which is wrong.  Fredson Bowers called these situations "indifferent"  not because he did not care but in allusion to the center of indifference--that point between the poles of a magnet where the pull in each direction is equal.  Nevertheless, errors have been made and some can be identified.  We divide them into innocent and sophisticated errors.  An innocent error is an inadvertent error which usually strikes the reader as an error and can usually be fixed in one's mind while reading.  We often call them typos.  It is surprising how easily readers auto-correct innocent errors.  
Aoccdrnig to a rscheearch at Cmabrigde Uinervtisy, it deosn't mttaer in waht oredr the ltteers in a wrod are, the olny iprmoetnt tihng is taht the frist and lsat ltteer be at the rghit pclae. The rset can be a toatl mses and you can sitll raed it wouthit porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe.
Sophisticated errors, by contrast, create readings that are real words.   Sometimes, like an innocent mistake, the error is obvious but the correction is not. An example comes from William Thackeray's Vanity Fair. At the sale of the bankrupt Sedley household, the auctioneer offers for sale the picture Becky Sharp had drawn of Jos Sedley riding an elephant.
William Dobbin, remembering the occasion of the drawing, reacts in a way that the auctioneer takes to be a bid. Dobbin immediately blushes and looks confused, at which the text says, "the auctioneer repeated his discomposure."  No critic is on record with a successful interpretation of "the auctioneer repeated his discomposure."   It is obviously an error.  But the manuscript has disappeared, there is no surviving proof page, and so the only record of what Thackeray meant to write is in this botched physical survivor.  We want a correct text but we can't have one.  We can guess that the auctioneer respected his discomposure, which would be a bit odd, given Thackeray's own drawing of the scene, but at least "respected" makes sense.  Nobody claims that it is correct.
Unfortunately sophisticated errors do not always strike the reader as an error; they are plausible, sensible, and therefore invisible or even attractive.  I won't mention the wicked Bible in which we are commanded to commit adultery--though most editors insist that the word NOT was inadvertently left out.  Another example of a sophisticated error is the change in Herman Melville's White Jacket (1850) where the sailor, having fallen overboard and sunk into the sea has just about given up the will to live when a "coiled" fish of the sea brushes against him, reviving him with fear and love of life.
brushed by a coiled fish of the sea   1850
                        soiled                       1922
Melville died in 1891
A compositor for the Constable edition of 1922, mistakenly set "soiled" instead of "coiled", which not only made plausible sense, but led F. O. Matthiessen to exclaim that the metaphorical resonances among the soiled white jacket, the sailor's soiled soul, and the sea's soiled fish could hardly have been produced by anyone but Melville.  It wasn't.  Matthiessen didn't know he had mistaken an error in the Constable edition for a bon mot; but some other critics have opined that if Melville did not write "soiled" he should have.  So, perhaps after all we don't care if we have a correct text.
Which leads to the second question a literary critic can ask of the archive:
2.  When was the textual variant introduced?  This might be a variation of the first question--an attempt to determine  which reading was first or older and therefore perhaps the most authentic one.  If, however, we are looking for progression in the author's thinking or revision or adaptation for new audiences, we are beginning to seek facts about the textual history and potentially interesting explanations for textual difference instead of just dismissing that which we consider wrong. In the case of White Jacket, Melville died in 1891 and "soiled" made its first appearance in 1922, created by a compositor who had sophisticated the text--purposely or not.  
The third question may be yet another variation of the first two questions, asked in an attempt to determine the correct text so we can ignore all the incorrect ones--unless they appear to be attractive sophistications.
3.  Who made the change?  The likely suspects include the author, the secretary, the publisher's copy-editor, the compositor, a censor, someone from the marketing department  hoping for more sensation, or someone from the liability department hoping for less.  If we can pin the textual change on one of these suspects, will we then be able to value one reading over the other?  Are we still stuck in the first question, trying to find out which text is the one and only one we have to deal with and which ones we should avoid--now, perhaps not because they are incorrect, but because they are "less authentic"?  Or, perhaps better, are we now asking interesting questions about the voices in the text and the effects of who said what?
In the case of White Jacket, it is difficult to argue that the 1922 compositor had the "authority" to introduce changes, though, of course, he had the power.  [Slide]What about the case of Shakespeare's Hamlet in which Hamlet laments either his "too sullied flesh" (Q1 1603), his "too sallied flesh" (Q2 1604), or his "too solid flesh" (Folio 1623).  Who did it is hard to say, for though Shakespeare was alive for Q1 and Q2, he is not known to have had anything to do with those publications, and he was dead by the Folio.  And yet, might the copies the editors of the folio used have been seen by Shakespeare?  We do not know.
What about an aspect of the book that is not textual but is significant?  Lord Byron's Don Juan was first published in a large, expensive edition[Slide] with large margins and elegant type.
Childe Harold’s Pilgrimage. Canto the third. London: John Murray, 1816, pages 12-13. KCL Rare Books Collection PR 4372.F2.
 It was, therefore, literary art--although the publisher, John Murray, did not identify his publishing house on the title page or anywhere else in the book.  But the millions of copies of pirated editions that followed almost immediately were small things, on cheap paper, at cheap prices.
Were they, therefore, soft porn instead of high art?  Who did it?  In both cases, the decisions about format, and therefore the impact that the material presence had on readers, was more likely to have been made by a publisher than an author.  Should the question be "Which is the correct form?" or should it be "What impact does a book's physical form have on a reader's creation of the literary work, the verbal mind object, represented by the book?"
Or take the case of two poems by the Tasmanian poet Gwen Harwood, ``Eloisa to Abelard" and ``Abelard to Eloisa,"  under the pseudonym Walter Lehmann, published in and then immediately suppressed by The Bulletin (a Sydney literary magazine).

Several years ago an Australian friend gave me a photocopy of these poems and told the following story: the author had already published under the same male pseudonym in The Bulletin and had written a bitter complaint to the editor that her poems were distorted in publication by having all lines pushed over flush left, destroying, thereby, significant levels of indentation.  The editor replied that the magazine's column format influenced the policy, which was applied indiscriminately to all Bulletin poetry.  According to my informant,  Harwood then wrote and submitted the two poems reproduced here.  She submitted them, it was said, with varying degrees of indentation, and the editors predictably and indiscriminately ``suppressed" the author's intentions by printing all lines flush left---creating, thereby, the ``inadvertent" acrostic message, readable vertically down the left column: `` Fuck all editors.  So long Bulletin."   From that story I concluded that it is important to ask "Who did this?"
Facsimile of The Bulletin, 5 August 1961, p. 33.
  I subsequently found that no one, including the author, Gwen Harwood, remembers the sequence of events in the way I have described them.  In fact, no one has the same story to tell.  Another Australian acquaintance actually asked Ms Harwood if the account I have given here was true, and she denied it, albeit forty years after the fact.  Regardless of the accuracy of the explanations for the physical poems,  my point is that a sense of agency (whether the historically correct one or not) is a determining factor in every reader's interpretation of the text.  
The questions prompted by the first, probably erroneous, tale remain the important ones:  Whose intentions are fulfilled?  Is this the social contract at work?  Has cultural materialism expressed the mind of society?  Has language spoken the author?  Every interpretation of the poems implies or states answers to these questions.  A biographer's, a literary critic's, and an editor's question's all converge:  is this an innocent pair of poems with a structural element of  bad taste and bitter recrimination? or are there other interpretive matters at work here?  When Eloisa says to Abelard ``Solace and hope depart" is that the author's explanation for her decision to say So Long to the Bulletin?  When Eloisa calls Abelard ``the faint ghost of a remembered dream" and concludes that she ``reap[s] the harvest of my own desire," is the author really referring to this act of bibliographical revenge?  Is the final line a warning to the editors of the Bulletin that ``No heart escapes the torment of its choice"---in this case the choice to ram all lines flush left?  These examples demonstrate that we frequently take poems, and probably novels and short stories and plays, to be hermetically sealed art objects, isolated from their context of origination, perhaps because we do not have enough training in confronting and reading the clues in the archive.

4.  Why was the change made?  Of course, it will help in determining why a change was made if we know who did it, but that is not always necessary.  Corrections in  spelling, punctuation, and grammar are sometimes thought to have self-evident motivations.  That works okay unless on is reading or editing  Finnegans Wake, or The Sound and the Fury, and such.  But when one perfectly good word is changed to another perfectly good word, or when one sentence or phrase is dropped from or added to the text, the critic in us, one hopes, starts by searching for answers.  If we don't know who made the change, when it was made, or why it was made, we might be tempted to make up explanations that pass a plausibility test, not a test verified by the archive.  Given the archive, we are not always thrown upon the resources of our own imaginations to answer the question: Why was the change made?  
     For example, in Thackeray’s The History of Henry Esmond (1852) there is a scene near the end of volume one of this “three-decker novel” depicting Henry coming home on holiday from the university and being greeted with joy by his foster mother, Rachel Esmond, and her two children, all of whom soon troop up to see Henry’s room, freshly gotten up by Rachel. In every edition from 1852 to 1989, we read that Henry and Rachel go up “hand in hand.” The phrase is not unusual, but Rachel is in a difficult marriage; Henry is only a few years younger than his foster mother; she soon becomes a widow; and in volume three, they marry. So, a careful reader might be justified in seeing this hand-holding as an early indication of things to come. In the manuscript that served as setting copy, however, the first word in the phrase “hand in hand” is squeezed in, almost illegibly, at the margin and in fact says “hat”—Henry goes up “hat in hand.” As a significant phrase about the relationship between Rachel and Henry, “hand in hand” comes far too early and the material evidence of the extant manuscript points strongly to the conclusion that the compositor created an adventitious reading—or let’s just say it: an error.  That is to say, the reason the change was made is that the compositor did not have time to decipher a cramped word squeezed into a tight corner and instead set the first thing that came to mind as a potential reading.  Needless to say, he got away with it for well over a century.
    On the other hand, in Thackeray's Vanity Fair, the surviving manuscript says of Becky Sharp that "Ill-natured persons say that her birth preceded the lawful celebration of her excellent parents' marriage."  This hand-written indication that Becky is, what at the time would have been thought of as a natural born bastard, not just a self-made one, is deleted and replaced by the comment: "And it was curious to note how as time passed this young lady’s ancestors advanced in rank and splendour"—a comment referring to Becky's ability to enhance her past as needed.   There is little question about who made the change (Thackeray) or when it was made (during proofreading); and the reason for the change is not a wild speculation, just a mild one--that it is a case of self-censorship, appropriate for a mid-Victorian mainstream novel. I think, however, we are happy to know what Thackeray's first thought about Becky was.

5.  What do I do if both (or all three or four) variants make sense and I cannot pin any of them on a nefarious interloper?  Here, on a grand scale, we have an "indifferent variant"--one where our search for evidence leaves us with an unguided choice.  A E Housman describes the situation memorably of the reader/editor caught between two equally attractive textual possibilities by comparing him to an ass equidistant from two bales of hay.  Such critics think, Housman said, that by arbitrarily choosing one hay bale or variant over the other their ears will get shorter.
     Emily Dickinson's F1460A is an untitled poem [Slide] which illustrates her way of trying alternative words without, in the end, choosing a "correct" version.   One could be frustrated by such an indecisive poet, as if she couldn't finish.  Or one could learn to read as she wrote.  We could perhaps read her poem as if it were a bit richer than conventional poems.
     We can explore together.
A chilly peace infests the Grass  
maybe that does not quite get it right, so let's try
A lonesome peace infests the Grass
and then let's try
A warming peace  
These three alternative forms are not resolved.  Let us revel in all three without actually making the poem longer; it is either a chilly Peace, a lonesome Peace, or a warming Peace or a bit of all three.  One's mind can easily grasp the suggestive complexity.
The Sun respectful lies -- Not any Trance of Industry
These shadow scrutinize--
Or is it: The shadows?   Why, having written These Shadows did ED think that The shadows might be closer to the mark?  or is it?  She does not say.  But we sense the difference between These and The and the ambivalence that distinguishes the specificity or generality of shadows.
Whose Allies go no more astray
or wait, perhaps astray is a bit strong, let's try Whose Allies go no more abroad
For service or for Glee --
But wait again, perhaps not for service, perhaps for Honor, or wait again, perhaps for welcome.
Is it astray or abroad?  Is it for service, honor or welcome -- which is the best or most evocative or most appropriate contrasting alternative to Glee?
But all mankind deliver here from whatsoever Sea
Or should that be Though all mankind deliver here?
No, perhaps mankind cruise softly here from whatsoever Sea--
Or once again, perhaps they row softly or sail softly
or finally perhaps mankind do anchor here from whatsoever Sea.
I do not see any mistakes.  I see no errors.  I see variant readings that each have or had an appeal to the author.  The temptation for many editors has been to do the choosing for ED and end up with one continuous text to be read linearly.  My point here is that the choices are indifferent--held motionless between the poles of the magnet, pulled equally by the surviving evidence.  Any choice made by any editor will be arbitrary--based on personal predilection, not on archival evidence.  The choice will have no shortening effect on the editor's ears; but it will shorten and impoverish the poem for readers.
Before moving to the sixth question to ask the archive, can we just let go of the notion that there is always a correct or even a preferable text that is the one we wish to represent the work, world without end, amen.  You do not need to let go of the notion that some texts are wrong.  Some texts are wrong.  Many texts are wrong.  But it does not follow that one text is right or even that, with enough diligence we can create one text that will be the right one.  Error we abhor, so, off with its head--that is, if you are sure it is an error--or that it is not an interesting and useful error.  But, can we just learn to live with viable alternative texts and ask some decent questions when we confront the archive?  Can we not, in full textual awareness, just say which text of the work am I talking about? and also say, here are some reasons I've chosen to talk about this text rather than that other one? or even to say, actually I want to talk about both or several texts and show how knowledge of the alternative texts has helped me as a critic. These seem to me better than pretending that there is one correct text, the well-wrought urn.

6.  What effect on my reading is had by each viable alternative text?  This question is similar to but crucially different from the question: What motivated this change in the first place? --which is a question about authorial intention, to which you do not have access.  But the question you can answer, both accurately and in a variety of ways, is: What effect on my reading is had by each viable alternative text?  Perhaps it will coincide with what motivated the author, perhaps not.  Perhaps you will contemplate the effect, not of each reading in turn, but of each reading held in tension against its alternative.   In any case, it will have caused you to think deeply about a passage of text which had also troubled the author.
The poem we've just looked at, to illustrate the situation when we have alternatives that all make sense, also illustrates the richness of seeing alternatives.  Read in a way that holds the alternatives as pending choices can make the poem speak more eloquently in their ambivalence than any one of them would have spoken in marble-smooth singularity.  I do not think I want a correct text of this poem.  I think rather that this complex dialogue between author and paper is very clearly written already--correct in its multiple ambiguities.
7.  If I scrutinize each variant, including errors and nefarious interferences, and if I try to see where some of the work's locutions came from, as in allusions to, or phrases quoted (or misquoted) from other sources, and if I try to trace the genetic progression of the text, would I be able to see an authors mind a work? would I be able to see the birth and growth of a work? would I be able to watch as trial passages were replaced by new efforts?  would I be able to participate both intellectually and emotionally in the momentum of creation? might I also get the sense that when the archive gives out, and the librarian says there are no further forms of this work to look at, might I feel that the author has abandoned, rather than finished, the work--all the copies amounting to a series of efforts pursued until exhaustion or distraction interfered?   What then would you say to the critic who just wanted the correct text and please do not bother me with the details?
    Of course, authors do not always hold choices pending and unresolved.  They cancel (that is, they reject unequivocally) manuscript passages short and long for many reasons.  Were they errors? were they less felicitous locutions? did the author change her mind?  did a new creative surge wash out a previously interesting but perhaps less interesting effort?  How can we answer such questions without seeing the archive as a rich resource for critical thinking on our part and not just as a way to resolve textual cruxes?  And if the author on a death bed requests that manuscripts be burned, are we willing to comply?  Or do we register the request and hold it as one of the lights by which we examine the manuscripts?
    The scholar for whom the chair in Textual Studies I held at Loyola University Chicago gives us a brief example of how an author's mind mulls a word over time.  Of the changes Martin J. Svaglic traced in the textual history of John Henry Cardinal Newman's Apologia Pro Vita Sua, perhaps the smallest and most interesting is the triad of variants revealing Newman's developing attempt to make the doctrine of eternal punishment seem less terrible.
In 1864 he wrote that the doctrine could be "less terrible to the 'reason'"; the next year he changed it to "less terrible to the 'intellect'"; and, finally, in 1886 he wrote, "less terrible to the 'imagination.'"  None of these readings is incorrect; each was the word Newman chose for each publication.  If one tries to trace the possible motives for these changes or, with more confidence, the differences they make upon our understanding of what part of our being engages with a doctrine of eternal punishment (reason, intellect, or imagination), it should be much clearer why an interest in textual histories is not focused primarily or only on which is the correct text.
    For a far more detailed and convincing example, I will just refer you to a new book about to be published call How Borges Wrote by Daniel Balderston.  There is a huge problem with confronting the Borges archive.  His manuscripts are scattered to the winds and found in bits and pieces here and there.  Daniel has spent forty years on what is a bit like a global Easter Egg hunt and has created a surrogate archive of Borges manuscripts, complete with hovering guidance helping readers look at the myriad photos he has produced.  He has exposed so much richness there that other books could / should be written to tease out the joys of his archiving work.    
Confronting the archive is not boring or trivial; it is not just a duty or responsibility; it is not just for the bibliographers, collectors, and textual critics; it is exciting and critically rewarding--for those who have eyes to see and at least seven questions to ask.  If you doubt that, take a look at
Dirk Van Hulle ,Manuscript Genetics, Joyce's Know-How, Beckett's Nohow . University Press of Florida, 2008.
Daniel Balderston.  How Borges Wrote. University of Virginia Press.  Forthcoming, 2017.
Paul Eggert.  Biography of a Book: Henry Lawson's While the Billy Boils. Penn State Univ. Press, 2013.
Luca Crispi. Joyce's Creative Process and the Construction of Characters in Ulysses: Becoming the Blooms.  Oxford Univ. Press, 2015.