NamesIssues

Issues to address

Remaining questions and problems

January 9, 2009

This is a list of questions (in no particular order) that have come up during the process of cleaning up our current FileMaker data. A few were discussed briefly at a WWP staff meeting on January 8, 2009.

Surnames of monarchs

Under what circumstances should we consider a monarch's dynastic or family name to be a surname? Does the treatment of monarchical surnames remain consistent throughout the WWP's historical period, or does it change at some point? Is it consistent across all (European) nations?

Example: The Plantagenet monarchs. "The Plantagenets" is a standard way of referring Henry II and his various heirs who ruled England from 1154 until 1485. In modern usage, it sometimes is treated as the exact equivalent of a modern surname (e.g. Henry Plantagenet), but it's unclear to me whether this is historically accurate. To make matters slightly more complicated, the word "Plantagenet" was originally merely the nickname of Geoffrey V of Anjou (according to Wikipedia) and was later adopted in a somewhat more formal fashion by some of his heirs. Unlike later English monarchs, the Plantagents are rarely if ever referred to as "the House of Plantagenet".

Example: The Tudor monarchs. It is actually somewhat more common to treat Tudor as a surname (e.g. Henry Tudor, Mary Tudor, etc.) --- these _seem_ to be more or less historically accurate ways of referring to individual members of the family. However, the functional similarity to a modern surname does not necessarily mean that it is one in fact.

Example: The Stuart monarchs. It is extremely common to use Stuart as a surname, and it appears to have functioned in that way among their contemporaries.

Example: The Hanoverian monarchs. Once we get to the 18C, English monarchs no longer appear to have anything that looks like a modern surname. No one ever calls George I "George Hanover," etc.

Example: Other European monarchs. Most other European monarchs, particularly those from France and Spain (and their various possessions) don't appear to have had modern surnames either (e.g. the Hapsburgs, the Bourbons, etc.). However, I believe that some English texts _do_ treat some of these dynastic names as surnames --- I'm fairly sure that there are English references to, say, "Louis Bourbon." But that usage raises other questions: is it self-consciously incorrect for satiric or derogatory purposes? Is it an attempt to make European monarchical names conform to what was only an English practice?

Surnames of certain nobles and lines of descent

There are some cases in the WWP's current data where the relationship between surnames and titles/rolenames are ambiguous, particularly in the centuries immediately following the Norman conquest of England. It seems that it wasn't terribly uncommon for an individual to take as his surname the title of a predecessor --- or is it that the predecessor's title was actually created from his surname?

Multiple surnames and their order

For Spanish and Portuguese names (and possibly others as well), people are typically given one or more forenames and two surnames. The concept of a middle name is uncommon and few people (none?) have them. Spanish names follow a pattern in which an individual's first (primary) surname is the father's primary surname, and the second (secondary) surname is the mother's primary surname. For indexing purposes, the primary surname is always used first, followed by the forename and then the secondary surname:

Juan Gutierrez Guerra = Gutierrez, Juan Guerra

The problem is that this convention has historically been poorly understood by English speakers/writers, such that the name "Juan Gutierrez Guerra" might actually be "Englished" as "Juan Guerra Gutierrez." Similarly, English writers often use only a single surname when referring to Spanish/Portuguese characters, and it may not always be possible for us to know whether that surname is being used as the correct primary surname or a mistaken primary surname (actually the secondary surname).

In practice, this may not be a very large problem, as we currently have very few texts that name a significant number of Spanish/Portuguese people. However, it bears consideration as we may wish to encode texts that do (several English and American women traveled extensively in S. America in the early 19C and wrote about their experiences).

Geographical role of scriptural and mythological people

Do Biblical (or other scriptural) and mythological figures have a geographical role? Or, more precisely, is their geographical role determined by the geographical region in which the scriptural/mythological sources circulated or by the locations mentioned in the scripture/mythology?

Example: Atlas, said to be holding up the sky in northern Africa --- but a figure from Greek mythology. Is he classified as being associated with Africa, Europe, or both? Example: Some of the Titans (e.g. Rhea, Cronos, etc.) who don't exactly belong to any geographic place (they are essentially "the world," after all) but who are figures in Greek mythology.

More broadly speaking, is Greek mythology simply "European"? Or does it, by virtue of its influence and extension to parts of northern Africa and western Asia during certain historical periods also require classification as "Asian" and "African"?

I suspect that in any single instance, it is fairly easy to make this judgment (Atlas could be reasonably classified as belonging to Europe and Africa, Daphne probably belongs only to Europe, etc). The real question is whether consistency across all mythological or scriptural figures is something that matters to us, or whether deciding on a case-by-case basis is good enough.

Modernization/Regularization of geography

What approach to modernizing or regularizing geographic locations should we take? Many people in the database are referred to by names that include locations that no longer exist, may never have existed, or now have different names. Should we always supply the modernized location when we know it, or should we instead always use the original name when possible? If the latter, what counts as the "original" name --- is it, say, the name used by the Greeks? By the Romans?

Example: Troy. Is this an actual site presumed to be the same location as the modern archaeological site? If so, is this part of Europe (because of its role in Greek literature/history) or Asia (because it is located in present-day Turkey)? Both? Do we call this place Troy or Illium? Etc. Example: Lydda. Saint George, for instance, is said to be from the ancient city of Lydda --- what is now called Lod, located in modern-day Israel. (He was often referred to as "George of Lydda," making it essentially a part of his name.

In some sense, this is a very minor question --- the people who it will most likely apply to are those about whom we will be recording the least detailed information (i.e. a minor historical character in a dramatic text, etc.) To extend this issue a little further, however, we also should consider the problem of places that are now part of a geopolitical entity that did not exist at the time --- for instance, Germany, Belgium, Italy, the Netherlands. Both of these examples might, in fact, apply to textbase authors who may have resided in places now contained within those countries.

This is complicated slightly by the fact that the term "Italy" was widely used long before any single country of that name existed; it simply designated the geographic region where the individual principalities and kingdom that would later become the nation happened to be. It's even more complicated by the fact that large parts of what are now Italy were actually part of the Spanish Empire during the early modern period (the same is true for the Netherlands, and parts of Belgium) --- but we don't really want to use that as a naming criterion, do we?

I suspect that, as with the geographic roles of mythological figures, this will be fairly easy to figure out using common sense in any given instance. However, that approach raises potential problems for consistency when applied to several hundred individual records...

Nonbreaking names (and their relation to name variants)

There are several issues related to our current identification and use of nonbreaking names:

1. Who qualifies for nonbreaking name treatment? Many people in the textbase are referred to by only one name --- but that doesn't meant they don't actually have others. In these cases, their full name is the one we want to record as the canonical or authoritative name, but presumably we also want to record the name that actually appears in the text. Is this a nonbreaking name? Or is it a name variant?

Example: Plutarch (Lucius Mestrius Plutarchus) is always referred to simply as "Plutarch" in English texts. Current practice in the names database is to add "Plutarch" to the nonbreaking name field and leave other name fields empty, placing the full name in the name variants field. This strikes me as both inaccurate and inelegant.

2. Use of nonbreaking names for monarchs (and sometimes other nobles)? Is the standard name of a monarch (e.g. Queen Anne, King William II, King Louis XIV) a single unit which should be treated as a nonbreaking name? Or is it simply a convenient form of reference that should be broken down into its constituent parts (honorifics, titles, generational markers, etc.)?

Example: Henry II (Henry II, King of England) --- current practice puts "Henry II" in the nonbreaking name field, then supplies additional information in other name fields ("King of England" in the rolename field; "Henry" in the forename field; "II" in the generational name field; "King" in the honorific field). In cases like this one, there is redundancy in our current practice: we place content in the nonbreaking name field in order to have the calculated _name field read "Henry II" instead of "Plantagenet, Henry, King of England," and that content is then replicated in other fields. This is essentially a FileMaker issue, however, and presumably shouldn't determine our practice when dealing with the XML personography.

3. Use of nonbreaking names for saints? Is "Saint" part of a nonbreaking name? Current practice is somewhat mixed, with it often being included (Saint George, Saint Bennet) --- but sometimes it is excluded (Mary Magdalene) and instead treated as an honorific (Mary Magdalene, John the Baptist), and sometimes it is included as part of the nonbreaking name and *also* treated as an honorific (Saint Stephen, Saint Crispian). I don't know that it much matters which approach we take, but greater consistency would probably be a good thing.

Additional names, nicknames, epithets

What do we consider to be the "additional" part of the name? In the case of nicknames, it clearly should simply be the nickname (e.g. "Bill" if the person's forename is "William"). Epithets seem less straightforward: do we wish to include only the part of the epithet that isn't already part of another name field (e.g. "the Hammer" in the case of "Charles the Hammer") or do we wish to include the full epithet even when it repeats name information stored elsewhere ("Alexander the Great")?

Ideally, avoiding redundancy might be a good idea if we could later reconstitute the full epithet algorithmically. But because we are using the same field for several different kinds of additional names, I don't know if that would work. An algorithm that correctly puts together William "Bill" Bennet will fail with William "The Bard" Shakespeare or Ethelred "Ethelred the Unready" I.

Honorifics

Under what circumstances should honorifics be supplied? In all cases? Only for people for whom honorifics cannot be inferred from other name information (such as titles/roles)? Only for individuals whose names appear with honorifics in source texts? The latter presents further consistency problems, since some people in our textbase sometimes are referred to in honorifics in some texts, but not in others --- so which do we take as authoritative?

Additionally, does the inclusion of "Mrs." or "Miss" as honorifics for women writers have patronizing, dismissive, or derogatory overtones --- even if the usage is, strictly speaking, historically accurate? Earlier generations of (mostly male) literary scholars often wrote about "Miss Austen", "Mrs. Shelley", or "Mrs. Behn" --- but no scholar would do so now except to make a point about earlier scholarship.

Example: Eliza Haywood, who sometimes published under the name "Mrs. Haywood", sometimes "Mrs. Eliza Haywood", and sometimes "Eliza Haywood." Do we say that because she was sometimes referred to in print as "Mrs. Haywood" we should include "Mrs." as an honorific? Would we then be required to apply the same logic to her male contemporaries: Mr. Samuel Richardson, Mr. Defoe, etc?

Example: Susannah Gunning, who published exclusively under the name "Mrs. Gunning" (her daughter, Elizabeth, published under the name "Miss Gunning")

Multiple surnames resulting from multiple marriages (for women)

What is the best way to handle multiple surnames for women who had multiple marriages? Which of the surnames do we treat as canonical? Should publication history be a consideration (if she published most work under a particular surname, that is the canonical one)? Should the other names simply be listed in a notes field, or are they more properly considered name variants (even though, strictly speaking, they aren't "variants" at all, but actual names by which the person was known at some point)? How does all this work with our typical treatment of surname vs. birthname?

Example: Anne Clifford --- who, following our practice elsewhere should be either Anne (Clifford) Sackville or Anne (Clifford) Herbert. Yet the name record lists her as Anne Clifford, and then supplies both Sackville and Herbert as name variants.

Authors who published under their birthname after marriage, or who were married after the conclusion of their literary career

There is an inconsistency in the way we treat female authors who marry after all or part of their literary career has ended (or who married early in their careers but continued publishing under their birthname). This is compounded by the way scholars sometimes refer to these writers using other name combinations:

Example: Judith Sargent Murray, who is typically referred to by that name by scholars. Following typical WWP usage, she should actually be Judith (Sargent) Murray, yet few scholars do refer to her in this way --- at least among Americanists. (Actually, I should check this out to be sure --- now that I think of it, there are some feminist scholars from the 1980s/early 1990s who very well might.) She had originally married John Stevens but is never called Judith Sargent Stevens. But neither is she called simply Judith Murray.

Example: Elizabeth Gunning, who published under the name "Miss Gunning" prior to her marriage. After her marriage, she became Elizabeth (Gunning) Plunkett --- but scholars do not refer to her in this way. Instead, they either disregard it (because she did not publish under that name) or make a passing indication of the fact (as the Feminist Companion does): "Elizabeth Gunning (later Plunkett)"

Example: Mary (Scott) Taylor, who we currently identify as Mary Scott and include "Taylor" as a name variant.

The main concern in all these examples is consistency. I suspect that someone is going to be offended or displeased by whatever approach we take to the marriage/naming nexus --- which I don't mind in the least so long as we have a clear, defensible rationale.

Roman (and possibly Greek?) names that do not follow modern name conventions

Even though Roman names follow a pattern that looks exactly like modern English forename/middle name/surname usage, the similarity is misleading. According to the helpful Wikipedia entry, the first part is the praenomen which does correspond more or less to a modern forename. The "middle name" is actually the nomen, which is a "gens" or clan name; the final part of the name is the cognomen (family within the clan). The latter two don't map cleanly onto the modern surname category --- in this sense, Roman names share the double surname problem with Spanish/Portuguese names.

This can sometimes be complicated by the fact that English writers didn't always get the name functions and order right, particularly since names written in formal inscriptions (as opposed to more informal or personal contexts) sometimes omitted or abbreviated the cognomen. Also, some names included an agnomen (apparently the equivalent of a nickname?), which sometimes preceded the other three but at other times (depending on context???) was intermingled with other name components.

Examples: Marcus Junius Brutus, Gaius Julius Caesar, Sextus Pompeius Magnus Pius.

Biographical notes

What standard kinds of information ought to be included in the notes field? This may depend in part on whether we imagine making the contents of this field available to readers in some form, or whether we imagine its primary role being for internal disambiguation purposes.

If we think it is reasonably likely that we will want to publish the biographical prose in some form, it may be worthwhile to consider what basic elements we would want each note to contain. This could include (but certainly isn't limited to):

• Additional details of education, socioeconomic background
• Major works for which they are known
• Literary, cultural, religious, or political movements they are associated with
• Associates or other members of their literary circle (for coterie writers, etc.)
• Additional information about titles or honors (e.g. "1st Earl of Catlehaven, Second Creation")
• Other things?

Name/role classification: imprint

Currently, we are using the general term "imprint" to classify any person whose name appears on a title page or who is otherwise associated with the publication and circulation of our texts. We may wish to consider granularizing this by adding more specific categories --- perhaps using our new values for @type on <docRole> (publisher, printer, bookseller, etc.)? One potential problem with doing so this is the fact that there was so much overlap between these more fine-grained categories. Many booksellers also acted as de facto publishers, many printers also sold books at their shops, etc.

Name/role classification: authors

Currently, we identify anyone who is the author or translator of a WWP text as a textbase author. However, any historical person who is known today for his/her role as an author and who is mentioned in one of our texts is identified only as a textbase figure --- meaning that there is no distinction at all between people who scholars might care about for their roles as important literary figures and people who simply happen to be mentioned in a particular text. Should we attempt to identify prominent literary figures as authors, even if they did not write any of our texts?

One advantage of doing so is that it puts us in a better position to trace patterns of literary influence across the textbase by making it possible to see what other authors our textbase authors found sufficiently important or notable to mention. A potential disadvantage is that the threshold for authorship may seem arbitrary: do we a person as an author if he published one poem in a newspaper? If he wrote horticultural letters to the Royal Society that remained unpublished during his lifetime? If he was believed to be the author of a scurrilous pamphlet during his lifetime that he did not, according to scholars, actually write?

Examples: William Shakespeare; Cicero; Homer; John Wilmot, Earl of Rochester --- currently all are merely classified as someone mentioned in the textbase; the fact that they are important literary figures is not recorded in any machine-readable way.

Sept. 16, 2008

Nationality

In an earlier meeting we decided that if we choose to record information about nationality, we would do so only in a very broad way and treat it less as a category for designating national allegiance, whatever that may have been in the early modern period, than as the "country of habitual residence." In other words, we would simply be recording information about where a person happened to reside the longest during his/her lifetime.

If we want to proceed with this plan, we need to address how we'll deal with some foreseeable complications:

1. Authors for whom it is difficult to specify a single nationality, even if we apply the "habitual residence" criterion fairly loosely
   • Susannah Rowson as an example of this problem: Rowson (not yet published in WWO, but a few of her works are in progress right now) was born in Portsmouth, England, but moved to Massachusetts at the age of 5. She remained living in the colonies until 1778, when her family returned to England. She continued living in England until 1793, during which time she published a number of her most successful works. From 1796 until 1824, she lived in Massachusetts once more. This means that Rowson lived in England for 20 years of her life and in British America (later the United States) for a total of 42 years.
   • If we accept that "habitual residence" means greater than 50% of one's time is spent in a particular place, then we could reasonably say that Rowson lived in the United states. But doing so fails to register a couple of important pieces of information: (a) the fact that Rowson's multiple places of residence are themselves important to scholars; (b) most of the texts for which Rowson is best known were actually written during the period when she wasn't living in her "country of habitual residence" but was living in a state of semi-exile in England.
   • There's also the issue of precision of terms: in Rowson's case, she only lived in the United States after 1789, since before that no such place officially existed. When she resided in North America prior to that point she was living in either "British America" or "the British colonies".
2. Place names that appear, disappear, or change during our textbase period
   • There's also the question of what to do with names for designating a "country of habitual residence" when those names don't really mean the same thing over the entire textbase period. For instance, depending on your source and the particular period, the exact definition of what counts as "Scotland" (versus either "England" or "Great Britain") varies widely. (To say nothing of places like "Germany" or "Italy," which didn't acquire formal definition as nations until after the period we deal with -- yet English writers routinely used both those names to refer to the same general geographic regions they apply to today.)
   • Places that were once British colonies are equally problematic: It's easy to say when parts of "British America" become "the United States," but at what point do places like the British W. Indies, Canada, or other territories become (or cease to be) "Britain," "Great Britain," or some other term. This may not be much of an issue at the moment, since we have few if any texts by writers who spent most of their lives in any of these places, but it seems desirable to make our system of classifying nationality or place of residence flexible enough to accommodate texts we may want to add in the future

Based on these points, there are several options we may wish to consider:

1. Forget about encoding information on nationality/residence as part of the personography
2. Keep the idea of "habitual residence" as our sole criterion for determining nationality, and use a 50%1 test for deciding a person's nationality
3. Retain the idea of "habitual residence" but allow it to have multiple values (a sort of faceted approach)
4. Shift the notion of "habitual residence" so it becomes a list of all countries in which a person resided at any point during her lifetime (with some arbitrary minimum duration)
5. Some other option...

[We decided that option 4 in the preceding list is the best approach; we are now using the residence field in FileMaker to record all places (at the country/nation level) inhabited by the person for a reasonable amount of time. "Reasonable" is essentially left up to our names researcher's judgment --- but as a rule of thumb, the amount of time required to make some place a residence for our purposes should be on the order of one or more years.]

Socioeconomic status

We haven't yet discussed in any detail ideas for working with information about individuals' socioeconomic status or profession. In order for this kind of information to be useful, given the spottiness with which some of it is probably recorded in historical sources and modern reference material, we would probably need to take a fairly broad view. A quick survey of a handful of our textbase authors suggests that most of them have very little in the way of detailed socioeconomic information included in standard resources like the Feminist Companion to Literature in English.

There is probably some information that we have a reasonably high likelihood of finding:

1. When individual authors are noble by birth or marriage
2. When individual authors practiced a known trade
3. When an individual's financial/economic status caused him/her to experience extreme poverty (this is something scholars seem fond of mentioning whenever possible)

Beyond these three broad areas, references --- of the sort that we'll have the time and resources to examine, at any rate --- to socioeconomic status tend to be fairly haphazard and incidental. For example, the _Feminist Companion_ notes that Mary Astell's father came from a family of coal merchants, but says little beyond this. Inferring Mary Astell's own circumstances from this passing comment seems problematic at best --- to say nothing of the numerous cases where there would be even less information to go on.

Based on the lack of consistency in reporting this kind of information, I'm inclined to think that actively seeking it out goes rather far beyond the scope of our present grant. Any information we find is going to be very incomplete without a LOT of extra research --- meaning that it won't be very useful as a category of analysis, since the lack of information for most authors will make it hard to make valid comparisons across the textbase. In order to get complete information for all our authors, we'd need to make our categories extremely broad: say, "noble" versus "common" or something like that. (Even if we wanted to use some other general term like "trade" or "profession", we'd run into trouble. For a decent part of textbase period, things like law and medicine might reasonably be considered "trade," but that was still very different from other professions that would fall under the "trade" heading -- blacksmith, say, or cabinetmaker. Those very real distinctions would be lost unless we were to granularize them using categories like "manual labor," and even there I think we run into a whole different set of problems.)

[We have decided that for the time being we are not recording socioeconomic data in a systematic way. If there appears to be an important piece of information about an author's socioeconomic status, it can be included a brief prose note at the researcher's discretion. This will typically take the form of statements like "washerwoman to the queen," "daughter of prominent coal merchants", "owner of several ships in the East India trade", or "died in extreme poverty".]

Developing ideas for user interaction with name data

Overarching questions and problems

One of the desired outcomes of the present grant is the development of a set of tools that will allow users to interact with our personographic data in a variety of ways. Before we can begin developing those tools (or selecting existing tools for use in the WWO environment), we need to identify the kinds of scholarly needs we hope such tools will accomodate or support. This means asking at least three different questions:

• What kinds of things do scholars already want to do with texts, and what tools will help them?
• What kinds of things _might_ scholars do with texts if they had the right tools?
• What do we mean when we talk about "interaction" in the context of WWO texts?

This final question may seem obvious, but it's worth thinking carefully about the kinds of interaction (beyond the level of "interaction" that our present full-text view permits) that can be involved in reading/searching/playing with texts and textual data.

• "Linear" modes of interaction, involving the ability to perform searches --- including a variety of assisted or predefined searches --- that can be incrementally refined or modified based on the results of previous searches. Simple forms of such interaction might include:
  • the ability to select from a list of name types (e.g. "author," "publisher," "printer") and view a list of names that belong to each category
  • a set of filters that can be applied to any set of name-related search results to separate them according to name type
  • options for searching texts according to a limited set of user-selected names (e.g. find all texts that contain the phrase "wit and fancy" in content and are published by Robert Dodsley)
  • options for comparing the results of multiple name-based searches side by side (compare the results for the search string "wit and fancy" in works published by Robert Dodsley with the results of the same string in works published by Wilkie and Robinson)
  • the ability to search according to the location of a name within a text or set of texts --- for instance, searches that are context-aware, such that a user can find only those instances of "Sappho" that appear in titles, headings, or salutes
  • show all names of a given type, sorted by date/author/publisher, etc.; allow users to select individual names to see a list of all texts in which they appear
• "Non-linear" modes of interaction, involving the presentation of aggregate data in ways that may defamiliarize it, expose previously unnoticed patterns, encourage playing around, etc. These types of interaction don't present results in familiar search formats, and they may not even look like searches (though in many cases they)
  • existing interface tools like the mandala browser, tokenX, word clouds, tag clouds, etc.
  • social network or "degrees of separation" visualizers that show connections between authors whose work is published by the same person, characters who exist in the same text, or name references that occur in the work of a given author (users might be allowed to expand those connections to second\- or third-\degree associations, or mix different types of associations: see all classical references made by other authors who were published by the same person as the present author)
  • timelines and maps (whether animated or not) that respond to user input

Obviously there is a good deal of overlap between portions of these two lists, but the idea is to think about the different kinds of interaction that might be possible, identify the contributions each form of interaction might make to scholarly research, and also hammer out what exactly we mean by "interaction."

Desiderata for interaction (however we define it)

This is just a rough initial list of the kinds of things people might expect to do with name information, and can be edited or expanded as needed. It is likely that individual users will want to:

• view basic biographical details for a selected person (birth and death dates, place of birth/death, dates of primary activity, parental and marital status, religious affiliation, place of residence)
• find texts of passages in texts according to basic name-related criteria (all texts by authors born between 1690 and 1720; texts written by authors born in England but living in America)
• browse or search texts by name of publisher, printer, or bookseller
• observe patterns in name usage across multiple texts, texts by a single author, or even a single text (the frequency of classical name references in works of poetry versus novels; the ratio of historical names to mythological names in texts by Katherine Philips; etc.)
• identify structural patterns in the use of names in individual texts or entire genres (frequency of historical names in poems versus frequency of historical names in religious tracts or sermons)
• associate names of authors, printers, and publishers with the places where they lived or worked
• find names with multiple layers of reference (the same name referring to multiple people who have different roles)
• disambiguate multiple people with the same name
• find out what historical and literary figures were being referenced by particular writers in a given historical period

Original planning notes

These notes are from the grant proposal planning phase.

Name and person issues

• Generic names (Flora, Phoebe) that don't denote a specific person, even a fictional one: can these always be identified confidently? should they be treated differently?
• Issue of fictionality and in general the kinds of ontological space people in texts inhabit
• What kinds of information about people do we intend to map? (Practical scope question as well as theory question.) What are the reasonable boundaries one might draw in a project like this?
• What kinds of relationships between people, and between people and places, do we intend to map? (as above)
• What relationships are we creating between metadata and content with this project?
• Disambiguation and how to treat unresolvable ambiguities of reference
• Issues of multiple reference: cases where a name refers to multiple people, and cases where a name contains multiple layers of reference: allegory, typological references, etc. Robinson as "Sappho", a lover as "Laura" (accessing Petrarch and also Greek myth), etc.
• NB our observation about "genealogical" connections rather than Moretti-style aggregation: looking at connections between specific items in the network, rather than at linkages that emerge statistically

Encoding issues

• What do we need/want to model about persons that isn't provided in TEI?
• Where should the Big Personography in the Sky live?
• How can we link to name authority records outside our project?
• Can Library of Congress name authority records be exploited for our project?
• What is the relationship between LoC name authority records and possible name authority records of textual persons?
• How can our encoding reflect our ideas about the name issues above?

Budget items

• Brian at X% FTE
• Brian travel to conferences (DH, TEI)
• Student programmer or STG time (cost-sharing)
• Syd time (cost-sharing? Time availability is an issue here.)

Theorizing

Thoughts on <persName>:

The WWP currently offers no distinction in its encoding between the names of fictional and historical characters.  Our present model directs encoders to tag anything with "a significant portion" of human anatomy as a <persName>, making it so that the list of personas eligible for <persName> ranges from mythological hybrids to real, historical figures.  The encoding protocal, then, implicitly argues that all occurrances of <persName> possess the same textual value.  We don't need to apologize for this view, but only to understand that we have put the fictional world in dialogue with the historical.

In charting or mapping <persName>, we will be attempting to define two main types of relationships:

?    relationships between figures in the historical world as represented within texts: which printers, in which cities, and when, were printing which women's writing? Which historical figures were being given space within texts (and how? Can we answer how they are represented---where in the text, for how long, in what capacity?)  This could also be as simple as which authors were writing about their grandmothers, which about other authors, which never extending beyond the religious world?

?    relationships between texts or authors, as represented by overlap (intentional or otherwise) in persons or characters: which issues (if we encode personificaion of virtues and vices) were author's grappling with? Which source texts or stories (i.e., which traditional characters) were most popular for allusions and appropriations? What other books had impressed them?

<persName> refers to a number of groups.  The following list represents the spectrum from historical to the far end of the fictional:

?    the material personage: author, printer, publisher, dedicatee, etc., who exist in the paratexts, and are immediately mappable onto a material, historical world.

?    the historical personage: other writers, government officials, family members, etc., who exist within the text and are referred to non-metaphorically or non-allegorically, in an intentionally historical tone.  Yet even here, embedded within the text, many historical figures will shade toward the fictional: Shakespeare, any British king or queen, etc.  Figures of the past are corrupted by mythology.

?    the historical character: a grey area. Characters drawn from historical figures, but heightened or intensified by the shape of the text - any of Shakespeare's historical characters fall into this field, Bowdler's poem on the death of David Garrick which features his voice from beyond the grave, etc.  This group might expand to include family members, friends, and acquaintances mentioned within the text but about whom we have no historical data.  The question looming over this group is how to meaningfully and usefully either assign primacy to one persona - either the fictional or historical - or record the existence of both.

?    the actually fictional: fictions.  Despite the apparent straightforwardness of this group, some interesting connections could be drawn intertextually---characters carried over from previous works, borrowed from or based upon historical figures (think Falstaff and Sir John Oldcastle).  This field also includes the mythological, provided the text is refering to the characters as they exist within traditional mythologies. Here's a sticky question: where do biblical figures fall?

?    the blank character: characters with names typically drawn loosely from mythology or the pastoral tradition who serve as code for supposedly historical figures.

?    the non-person persona: while not currently tagged, personifications make up a significant part of the characters (defined as renditionally distinct agents) in the WWP's poetry.  If one of our interests is to chart overlap between texts, the personifications of vices and virtues could be an interesting area of study, because like direct references to historical figures or characters from other texts, they reveal what works and ideas an author is grappling with.

This list shows (I think) that unique, fictional characters will be the least interesting for our project (unless they are examined on a very general level) because they are the area of least overlap.  For example, it won't be that interesting if Eleanor Davies has a character named Brian---a name presumably unassociated with history, religion, politics, or other fictional characters.  On a general level, it would become interesting as a data point: many women have more male characters than female, draw their male names from Irish origins, etc.

If overlap is our interest, a profitable testbed would use a group of works related by period or location but also by mode of reference.  That is, they would either

?    refer to the historical and fictional worlds in the same ways (through the pastoral mode, historical accounts, mythologies, personifications, etc.), or

?    refer to the same historical or fictional worlds

Scholarly Interest in <persName> and names in general

What are the scholarly interests in the <persName> project?

?    Tracing the genealogy of character names.  How are figures from Greek and Roman mythology, the classic world, the Bible, etc., manifested across the textbase?  Do different names become more popular references during different periods?  Do references to types of names - classical, historical, fictive - appear regularly in certain parts of texts (dedicatory epistles, addresses to the reader, the body)?  A scholar could trace the appearances of figues like Odysseus or Athena through the WWO collection, linking their appearances (i.e., where and how in the text they are used) with the period and locations of their authors. This work allows scholars to trace the rise of particular literary characters through the popular culture and to relate these trends to the social contexts in which they operate, hence making visible the kind of work that literary characters do in the cultural imaginary of a particular period.

?    Establishing historical communities.  Were there intertextual relationships between writers?  Are there paratextual figures - patrons, printers - who textually overlap? Which printers were most popular?  Which patronesses? One can imagine, for example, mapping the authorial landscape surrounding the Countess of Pembroke.  How far - textually, temporally, and geographically - did her influence extend? n This kind of historical work is particularly important during the ages of the scholarly self-fashioning and intellectual salons and societies which our textbase covers. Scholars, particularly those extending the boundaries of cultural materialist work, are highly interested in the networks of exchange and dialog which helped to develop strong intellectual and literary trends. There remain a number of historical paradoxes or lacunae to be addressed and our work on names may offer significant insight in this direction. For example, what is known as the "Newcastle circle" is an intellectual network which drove the uptake of atomism and other natural philosophical ideas in seventeenth-century England. While the coterie that was established around William Cavendish's table has been well documented, the names which appear in the work of these writers and thinkers has not. There remains scholarly confusion about "the Chymist van Helmont" with whom Margaret Cavendish engages in her Philosophical Writings and elsewhere. While the elder John Bapista van Helmont was a great influence on Martin Mersenne and other emerging chemists, his son, Francis Mercurius van Helmont was a major figure in chemical practice, Cabalistic theory, educational practices and had extensive connections with Lady Anne Conway and William Penn, the founder of Pennsylvania. Because much of JB van Helmont's work was published posthumously by his son, who also published his own work, there is confusion amongst the readership regarding which of these influential writers is being referred to in individual texts. The implications regarding intellectual networks are significant, not to mention that clarity in name/person identification bears significantly on the ability of the reader to interpret a given text.

?    Unpacking double-references or coded characters.  What happens when historical figures are addressed using classical names?  Identifying contemporary figures with historical/fictional personae hints at a sense of history or the ideals of the author's world.  What historical or classical women were held as figures of such idealization? Double-reference names also imply interpretations of the source texts - how and what the authors were reading - and the uses of the fictional world. If we include personifications, are women identified with specific virtues or vices?  (What are the genders of this group?) Scholars continue to be interested in the use of emblems in the early modern period and in their eventual transition out of popular usage. Tracing such coded character names will allow further research into the possible displacement of emblematic signification into personages. At the same time, such tracing would make it possible for scholars to see how the Laura figure of Petrachan sonneteering tradition reappears in the political discourse of the British or how the figure of Eliza operates as both a model and critique of the historical Queen Elizabeth.  

(Here's the url for MaPs http://makingpublics.mcgill.ca/)

In the last several years the collaborative "Making Publics" research project has been investigating the nature of early modern publics, their formation, function, and the relationship of these to a recognizable cultural formation known as the 'early modern'. Such questions require large-scale detailed information about the constituents of the various publics under consideration; they require a detailed personography. As an example, Michael McKeon has used information about individuals, collectives, and larger intellectual networks to detail what he has called "the secret history of domesticity," a history which establishes the role of the private and the public in early modern epistemology.  Other scholars have considered the roles that personal networks have played in civil governance and its relationship to revelry laws, the circulation of knowledge systems within and across mercantile groups in England and on the Continent, and the interactions between artisan, sailors, and collectors in the development of early modern mapping technologies. In each of these projects scholars have used large quantities of information about the people and places involved in knowledge production to reshape our understanding of early modern publics. 

Roger Chartier's work in the history of the book and print culture depends upon knowledge of the print shops, printers, editors, authors, and booksellers of sixteenth\- and seventeenth-century Europe. Chartier's work extends this historical landscape of people and their circulation of books into its engagement within fictional texts as well. He describes in great detail the interaction of historical print society with Miguel Cervantes' character Don Quixote, and in doing so offers a new sense of how the fictional character also functions as one node in the complex network of seventeenth-century Spanish book history.

very brief biblio on these two

Chartier, Roger. Inscription and erasure : literature and written culture from the eleventh to the eighteenth century. Arthur Goldhammer, trans. Philadelphia : University of Pennsylvania Press, 2007

--    The order of books : readers, authors, and libraries in Europe between the fourteenth and eighteenth centuries. Lydia G. Cochrane, trans. Stanford, Calif. : Stanford University Press, 1994

McKeon, Michael. The secret history of domesticity : public, private, and the division of knowledge Baltimore : Johns Hopkins University Press, 2005