"Natural Language" Guide to Editing Biographies
BESS Biographical Elements and Structure Schema

  1. Introduction
  2. Persona Type
  3. Stage of Life
  4. Events
    1. Event Type
    2. Agent Type
    3. Location Setting
    4. Location Structure
    5. Time of Day
    6. Dates
    7. Season
  5. Persona Description
  6. Discourse
  7. Topos


Introduction

This project is based on an international annotated bibliography of 1200 books published since 1830 that include three or more short biographies of women: http://womensbios.lib.virginia.edu (CBW). CBW provides access to these books, some Featured Subject pages (conceived as starting points for research on individual women), a Pop Chart, and other continuations of the study by Alison Booth, How to Make It as a Woman: Collective Biographical History from Victoria to the Present (U Chicago P, 2004).

BESS is a system for interpreting, marking, analyzing, and in due course visualizing the patterns in short biographies (TEI texts in XML). Some of our concepts and terms resemble common-sense understanding of narratives both fictional and historical; all are relevant for scholars and theorists of narrative. Other terms we have developed from studying the varieties of brief biographies of women published in the nineteenth century (UK and US). An initial focus on 18 books that include a life of Sister Dora has generated our schema to date, but we are beginning a study of books that include a very different persona, Lola Montez, and have considered some of the “values” or vocabulary as it might apply to a scientist, Caroline Herschel.

BESS and CBW are not dedicated to a more common practice in literary digital humanities, variorum textual editions. Our schema is not for editing the specifics of the printed texts themselves. It is a “stand-off” markup schema. In order to focus on narrative structure, we set aside a great deal that interests us in the project on prosopography. The BESS analysis is coordinated with a particular text by identifying units or a range of units in that text, that is, numbered paragraphs. For example, a stage of life, “beginning,” will have the range of paragraphs 1-10 in a particular text. A different version of the same life might deal with the “beginning” in paragraph 1 only.

Such markup of individual texts is aimed at comparative analysis of: various versions of one persona’s life, revealing changing reputations across decades or different contexts (e.g. Hannah More praised by Evangelicals till about 1830; disparaged or ignored by feminists from about 1880); all the variations of that life as it compares with other BESS analysis of narratives in the category of the persona’s type (e.g. philanthropy); these variant versions of one persona in comparison to other short biographies in the same collection (thus characterizing the presenter—biographer, publisher, illustrator—or aggregate aim of that book); patterns of representation in different cohorts or persona types (e.g. performers vs. philanthropists); changing and constant form and style of short narrative biographies of women; hypothetically, a narrative theory of biography useful for interpreting biography in general, the monograph, the life of a man (setting aside the minimal dictionary or database entry.

Editors will read texts differently. In analyzing a biography, standard kinds of data (date, place, etc.) produce little debate. We distinguish observable elements such as events that contemporary witnesses would confirm from more abstract and interpretative elements, such as typical scenes (referred to as topos in singular, topoi in plural) or rhetoric and imagery. Stages of life will be open to interpretation (where does the persona’s life shift from “middle” to “culmination”?) within some range of paragraphs. Allusions may or may not be recognized by different editors. Nevertheless, across numerous biographies (often of some 50-75 paragraphs), the social and narrative conventions will emerge. Many other kinds of information (e.g. word frequency and proximity searching) can be gleaned from these texts and perhaps correlated with these conventions, in due course. It remains to be seen how much any of the interpretative analysis could be developed by other means than close-reading by humans.

One aim of this project is to experiment with digital editing in collaboration and to demonstrate ways that digital tools can be applied to a large archive of narratives that were treated in many ways as common property.

All biographies are interpretative to begin with, even when they closely follow primary evidence. Many of these biographies plunder or rewrite previous versions using fictional technique (scene, dialogue) and self-help or sermon-like commentary. This guide attempts to bring our editing judgments into some range of similarity by explaining some of the elements of BESS and the “controlled values” that are available to an editor of one of these biographies in our collection.

It is helpful for editors to remember that our task is not to identify the particular data in a text such as proper names or actual locations. We are tagging kinds of people (either the persona or the “agents” whom she interacts with), and whether they are individual or group, named nor not; kinds of setting; and so on. An exception is stated dates. We should also resist the tendency to fine-tune editing of the paragraph as such; again, this is not an edition of the text but an analysis of the elements we find at different points (no finer than a paragraph unit). This means that an editor may need to go back, for example, to associate a particular figureOrImage with a range of paragraphs rather than one.

As a method, some will want to read through the TEI file, roughing in the stage of life paragraph ranges. Some will want to proceed paragraph by paragraph, identifying the different kinds of elements (event, persona description, narration, topos). Others will want to pursue all events, all persona description, and so on. It is possible to search quotation marks to identify dialogue or quotation, search the use of “I” or “you” or “we”; similarly, some persona description or figures or dates are searchable in a first pass. A final confirmation of the BESS analysis for the whole will be needed to be sure it is logical, economical as well as thorough, and accurate. It will be difficult to interpret some categories without considering others.

Our primary categories of analysis for this study of narrative structure:



Persona Type

In BESS analysis, each XML file corresponds with a single biography in a specific collection and with a unique person or persona (all of which have unique identification in the CBW database). Each BESS file should identify not only the title of the biography and the name of the persona or subject of that biography, but also at least one of the Persona Type values in the schema. These types are largely occupational, and indicate the main reason the woman's story has been promoted. There are many more types of person among the 8,000 or more gathered in the database, but this more limited vocabulary serves our comparisons of more in-depth analyses of samples of texts using the BESS schema.



Stage of Life

Stage of Life is a small range of phases that might be understood in relation to plot in drama or fiction, but they refer to the parts of the persona’s life in its actual occurrence (as we understand it), not the parts of this particular narrative. In other words, we identify units in this text that deal with what happened “before” person was born; that deal with her early development; etc. In general, each stage will refer to several “units” in any one text, that is, a range of paragraphs identified by number. It is possible for the same paragraph to include more than one stage of the life, for example both the beginning of the persona’s life and the middle—the transition is narrated in the midst of that paragraph. While many biographies are chronological so that the sequence of the historical life and the stages of the narrative are in the same order, the narrative might begin with the aftermath of the funeral, i.e. that part of text would be marked “after.” Notes: Culmination: crisis, climax, best or worst of the narrative: some difference from the ongoing practice of a woman’s established occupation. In general, the narrative will do more than summarize repeating action but focus on an event or events that fulfill the persona’s narrative, perhaps bringing about the “end.” The “end” is not the same as the last word of the text, but the death of the persona, even if it is not described as a scene. “After” includes funeral and other posthumous events, but also may be statements about public response after death and ongoing reputation or more than passing references to history or society since death. We may find “after” in opening paragraphs, for instance.



Events

Events are observable actions in time and space that would raise little debate among interpreters. Some events have no clear outline or end point, but can be tagged to one or more paragraphs, such as "childhood or upbringing," provided the text explicitly narrates actions or occasions that belong with that value. Statements about activity such as "cooking," public events such as "battle," and transformations such as "conversion" all are identified in the event element.

Event Type

"Event_type" has "child" elements that go along with the definition of an event: action (by certain agents) in time and place. BESS assumes that the "persona" or subject of the biography is usually active in events in the narrative. Sometimes we have specified agent in the event type, thus "crying, persona's," but "crying, adult female" is used when another woman besides the persona cries. Editors should not tag an inference (a logical guess that some kind of person was involved in an event), but they should note in BESS the type of agent for any textual reference to participants in events (see "Agent Types").

Events and Dates, Time of Day, Season: Events often include indications of time. In BESS, we only record explicit statements of dates, using yyyy-mm-dd format. That is, although biographical information about the persona's life would allow us to make a complete timeline, we are responding to the stated facts of this version. If the text specifies time of day or season, we include that in the Event markup (see Time of Day and Season, below).

Events and Locations: Events also refer to space. BESS analysis of any event should indicate explicit narration of "Location Structure" and "Location Setting" (see below). Structures are part of the built environment, including vehicles or enterprises; settings occupy more space on a map than structures, and include natural formations as well as landscapes and countries. The named countries are an exception to the general aim of BESS to provide an abstract of types; in Anglo-American narratives, certain major national settings become significant signals of a different kind of life for a woman. Thus, some countries do not have named values in BESS, and England or the United States are only mentioned "as traveler or immigrant."

Event vs. Topos: A few event values closely resemble values for a topos or typical scene, for example “calling or vocation.” As an event, “calling or vocation choice, considered” or "calling or vocation choice, decision" are narrated actions by the persona; when she joins a sisterhood or becomes a dancer, it is a known fact. Similarly, “recognition, receiving award or prize” is a documented event which we interpret as recognition. Both “calling or vocation” and “recognition or tribute” are also important typical scenes or topoi, and may never produce a narrated scene such as joining a sisterhood or winning a prize (the narrator may summarize or assess; editors may detect an underlying theme). The same text unit may be marked as both an event and a topos of recognition, for example. The topos might be discerned across a range of paragraphs and a wider variety of actions.

Agent Type

Agent type is a fairly reductive categorization of the kinds of human or divine actors in the narrative events of a biography in this archive. As noted, it seems important to notice naming or anonymity, the single and the group. “Object” is a way to avoid “inferior,” and almost invariably the people of lower class status interacting with these women are the “objects” of their work, charity or reform or teaching. At times, the weather is a kind of agent, and frequently deity is addressed by epithet and interacting in persona’s life, hence sacred or divine, epithet or indirect reference to known (e.g. “the Divine Spirit came to her”).

Location Setting

Location Setting is usually larger than a structure, and is sometimes not created by humans. Types of settings beyond the home or building include landscape features, cities, and countries. As with agents, the usual rule is to identify the type not the specific instance: city, not Paris. We have made exceptions for some countries because in the collective biographies of women, France or Ireland seem to represent certain options for women or associations with persona types. Some biographies will characterize their personae as exceptionally mobile through the number and variety of settings included in the text, at times within one paragraph. Note that if an entire life transpires within the "default" country (e.g. England) without mentioning it, BESS analysis has little to say about national setting. But "village," "slum," or "colony" may tell us a lot about the actions of certain types of women in these texts.

Location Structure

Location Structure values indicate a variety of limited spaces for action, such as vehicles, buildings, and institutions or enterprises. These can be typical of social status because of property ownership or the means and power of mobility or access. Any of these structures would be familiar in all kinds of life narrative in recent centuries. Some structures determine the characteristics of a persona's life, as an occupational space. Again, BESS does not tag what a unit of text does not specify.

Time of Day

BESS editors should note whenever the text specifies time of day, as in "One night," or "every morning." If the text is silent about this, so is the markup. Some events may seem to imply time of day, but we make no inferences for purposes of analysis.

Season

Biographies sometimes specify season; if the text states the season, this should be indicated in BESS analysis.



Persona Description

Interpretation of personal qualities is going to be variable. The text will often supply specific terms, but we believe the positive or negative qualities for these examples of womanhood will tend to be normative even if the individual woman is described realistically as contemporaries knew her. We want to go beyond word searches to interpret the general kinds of attributes that these biographies, as a group, are identifying. Arguably the entire text is a characterization of a persona, but we identify units in the text that include statements about these traits.



Discourse

This interpretative set of elements will mark paragraph units or ranges that use different narrative techniques, components of text such as quotations, imagery, rhetoric and other features of narration. Included here will be characterization of the discourse—the behavior of the narrator—but not the physical features of the text per se (that is, we don’t include headers, paratextual features such as layout, illustration). An exception will be when we interpret an emphasis not only in wording but such effects as exclamation point, italics, capital lettering; although these may be decided by presenters such as publishers, it becomes an effect of the way the story is told by the teller.

We include here figures of speech, which are verbal equivalents of topoi. Extended metaphors go beyond one phrase to affect word choice (verbs, adjectives) in longer sentence or several sentences. FigureOrImage may include similes or metaphors or imagery of various kinds.

Analogies and figures of speech may of course also be identified as persona description, in which case it may be warranted to also indicate that element associated with that text unit or unit range. Not all references to named persons are analogies, which should explicitly compare the persona to a kind of agent.

Allusions may require some knowledgeable recognition and/or research; they are references to literature specifically or as paraphrase; calling something quixotic is not worth marking as allusion, but Rocinante, Dulcinea, or “the knight who tilted at windmills” all are “allusion, literary, reference.” A quote from Don Quixote is of course a “quotation, fiction.” Quotations will usually be indicated by quotation marks or italics, but sources often will not be identified, again requiring research. Dialogue (among historical persons represented in the narrative) may or may not be indicated in separate paragraphs of quoted speech. Not all questions are rhetorical; some questions are rather free indirect discourse, and focalization, attributing thoughts (“Would she never find a true love?”). It is possible to focalize without using free indirect discourse.

Evaluation specifically refers to a paragraph or range of paragraphs in which we find statements of approval or disapproval of the persona’s actions or character (in one sense these biographies are broadly evaluative). This may be closely allied with precepts, which we apply to a narrator’s statement of general advice. “She was wrong to avoid confrontation” is evaluation; “It is unwise of avoid confrontation” is a precept. This may also occur near or with direct address, use of I or direct address, use of we. But a narrator may write: “Reader, I hope you all remember the days before railroads,” and it is not an explicit evaluation or precept in our terms, and it is both direct address, use of I and direct address, use of we. Editors may disagree on what constitutes digression, but the techniques in this paragraph will be common in digressions. Let’s agree that a digression should be longer than a sentence, and not simply an allusion or analogy to a text or other person. It should be in some way about a topic other than the persona herself, e.g. the Church of England in general, even if the persona is C of E. We expect digressions and precepts to coincide frequently.

Complex matters of narrative time may be clearly stated (e.g. references to dates, time spans) or implied in such things as verbal tense. It is very important for narrative structure to note flashbacks (retrospective), flashforwards (prospective or foreshadowing), or the even more frequent condition of ongoing or continuous conditions, iterative, repeating or persistent: “For weeks she continued to search for the lost book.” This may cover a great deal that is not an event, and even may be most common where there is no topos or value-laden structure of relations with other people.



Topos

Topos is a type of scene, not necessarily narrated as an observable historic event. It may or may not correspond with an aspect of discourse such as "precepts" or a "figureOrImage." Indeed, you could think of topos as a blend of event and discourse with social convention; we do not suppose the topos was apparent in the persona's immediate experience, but that the text has adapted this version of the life to the structure of feeling or social template that gives meaning to the life as an example. "Topos" comes from the word for place, as in topography. Topoi (plural of topos) can be thought of as like the common scenarios in classical paintings, or as patterns of relation in myth (a temptation in a garden is a topos, heavily influenced by Genesis). In our work, we have gleaned common topoi from a large archive of a certain genre, brief collected biographies of women, but many of these scenarios are likely to appear in long biographies, and representations of other kinds of people including men. Situations become recognizable as a topos with a lot of repetition; experience, and just one life story, don't create a pattern that we recognize. Although events or observable actions can be interpreted as topoi, the text does not have to spell out the topos for us to mark it in BESS.