FAQ: Frequently Asked Questions about
Data Entry in the Genre Evolution Project 

FAQs relating to:
All Genre Evolution Databases
The Text Database
The Author Database
The Context Database 
(see also Online data sources; Guide to Access Queries)
Worksheet for Text Coding
Worksheet for Context and Author Coding
Worksheet for News Periodical Cover Coding in Excel 2000 format
Spreadsheet of Missing Context Data in Excel 2000 format

Online data sources:
Most reprinted Authors and Stories (from Contento's Index to Science Fiction Anthologies and Collections [through 1983])
The Locus Index to Science Fiction (continues Contento's Index, 1984- )
Internet Speculative Fiction DataBase (ISFDB)
ISBNdb.com Publishers Search
Gale Literary Databases (including Contemporary Authors)
Reda's combined search engine
Pulps & Magazines Americains (includes many magazine covers)
The Philip K. Dick Bookshelf (includes books covers)
TIME Magazine: 100,000,000 words, 1923-2006 (corpus linguistics dataset)

top

All Genre Evolution Databases

Q: Which categories are entered automatically?

A: The ID number for each database (e.g., Context ID in the Context Database) is created and entered automatically when you click the New Record button. Similarly, entry dates and times and calculated values (e.g., %3DimChars) are created and entered automatically. When an Author ID is entered in the Text Database, the AuthFirstName and AuthLastName fields are filled automatically. Do NOT edit automatically entered data.

Q: In what format should I enter dates?

A: Enter all dates in XX-XX-XXXX format, to be able to differentiate between 1901 and 2001.

Q: How do I enter years as dates if I know only the year (e.g., 1967)?

A: Use the first day of the first month as a placeholder (e.g., 01/01/1967).

Q: How do I enter seasons as dates (e.g., Summer, 1998)?

A: Use the first day of the first full month of the quarter of each season to represent seasons:

+ Winter: 01/01/1967

+ Spring: 04/01/1967

+ Summer: 07/01/1967

+ Autumn: 10/01/1967

Q: How do I enter months as dates (e.g., May, 1967)?

A: Use the first day of the month as the day placeholder (e.g., 05/01/1967).

Q: What do I do if I do not wish to enter data for a field that uses radio buttons, but I have already selected one and cannot unselect the field?

A: Either clear the form and re-enter data or use the "Not Applicable" or "N.A." option in the offending field.

Q: Are there other situations in which I should enter "Not Applicable" or "N.A."?

A: "Not Applicable" or "N.A." is offered as a choice for many of the fields. This should be used sparingly, typically as a last resort, and always with special care. In fields with value lists, N.A. is most likely to signify that the value needed is missing. In that case, the coder should enter in the reader's notes a suggestion as to what that missing value should be. If the total of N.A.s in a given field becomes significant, we will consider adjusting the value list. In fields that need not be filled in (e.g., Theme B), N.A. should NOT be used to indicate that no data are being entered; simply don't enter the data. If there is some other use for N.A. in your estimation, always include an explanation of your use in the reader's notes and report that item by record i.d. and field name to gecontextdb@umich.edu.

Q: How do I enter countries in the databases?

A: Use the standard Internet two-letter country code (available at http://www.ics.uci.edu/pub/websoft/wwwstat/country-codes.txt). For example, do use US; do NOT use U.S., U.S.A., United States, United States of America, or America.

top

Text Database

Warning: When doing a RECODE, search only for records where Reprint = No. Updating reprints will be handled separately.

Q: What do I record if the story title appears differently in a magazine table of contents and on the title page of the story?

A: Record the exact title of the story as it appears on the title page.

Q: What do I do if I'm not sure what a field or value means?

A: Go to the Text Fields and Values page for definitions and examples.

Q: How do we handle the Theme fields?

A: That depends on whether the theme is stable, evolving, or a surprise.

+ If Theme is Stable, try to settle on one theme, enter it under ThemeA and put n.a. under ThemeB. However, if it is impossible to settle on one theme, enter the more dominant as ThemeA and the other as ThemeB.

+ If Theme is Evolving, there must be both a ThemeA and a ThemeB. ThemeA is the original theme and ThemeB is the theme the story evolved into.

 + If Theme is surprise, there must be both a ThemeA and a ThemeB. ThemeA is the first theme and ThemeB is the one that surprises you.

Q: How do I code main characters?

A: Choose ZERO as the number of main characters if there is no main character. Choose MANY if there are more than three. If the number isn't zero, always focus on one character for Main Character Dimension. In the main character sex field, if there are characters of more than one sex or both no males and no females, enter "other." It is possible to code some Main Character fields for more than one character (i.e., select more than one TYPE or AGE) because those fields have checkboxes to allow recording cases in which the main character changes or there is more than one main character. However, always choose ONE Main Character to code if you can because that aids in reading the data. If there was a close second, you can note this in the READER'S NOTES field. See the Data Entry FAQ under Text Database, for more on this.

Q: How do I fill out the data on the main character when I find that there are no main characters?

A: In MainCharNum, put a zero. For the MainCharType, select the ground type of character in the story (is this an epistolary story on earth with humans as the participants, or on Mars with Martians?). Do not put a MainCharDim, MainCharSex, or MainCharAge.

Q: What should I enter if there is more than one main character?

A: Enter only one main character. If this presents problems (e.g., there seem to be two genuinely equally important main characters), enter only one and explain your difficulty in the ReadersThoughts field.

Q: Why is there more than one text record for some stories?

A: A story has more than one context if it has been reprinted. We use a record for each publication of a text, thus automatically generating multiples Text IDs for that story and requiring associating each of that story's Text IDs with the Context ID that represents its place of publication and/or reprinting.

Q: What publication date should I enter for a story?

A: In the Text Database, always enter the original publication date for a story and an appropriate Context ID. If the story is not a reprint, the original publication date is the date of the magazine or book in which it originally appeared; if the story is a reprint, the original publication date is still the date of the magazine or book in which it first appeared. However, originals and reprints of the same story will have different Context IDs because they appeared in different contexts. In the Context Database, find or create a record for the context in which the story was published (e.g., the March, 1930, Amazing Stories) or reprinted (e.g., some anthology) and enter the appropriate ContextID in the Text Database, that is, the Context ID for the original place of publication when the Text Database record for the story has Reprint No, the ContextID for the place of reprinting when the Text Database record for the story has Reprint Yes.

Q: How do I know if a story is a reprint?

A: You cannot be sure if a story is a reprint if the magazine or anthology does not tell you. Check the table of contents and the first and last page of the story. Look in the editor's preface to the story (if there is one). Look for footnotes. Look for a list of Permissions or Acknowledgments. If there is no evidence, say NO. If there is, say YES and note the date in the ORIGPUBDATE field, above. Also note the magazine or book title (you could put this info in READER'S NOTES) to be of help to the person managing reprints for the GEP.

Q: Do I have to count every word in a story to get a word count?

A: You can use sampling to estimate word count. One method is to select a sample page of text at random from the story you are reading. Count the number of lines on the page. Select seven or more adjacent, visually representative (typical use of white space) lines and count the number of words in those lines. Divide that number by the number of lines you selected to count from. That number is the average number of words per line. Now multiply by the number of lines on the page and you have the number of words per page. Then multiply by the number of pages in the story. Talk to other members in the group if you want to try another method. Chances are your reading partner(s) and you will not have the exact same number. You can average these together or come to some other common agreement. In cases of serious disagrement (more than 10% discrepancy) , you can scan a story and use a word processor's word count feature.

Q: How do I count three- , two- , and one-dimensional characters?

A: Three-dimensional (well rounded) characters and two-dimensional (stereotyped characters) should be counted according to how many of them are in a given story. The main character, for example, is likely to be two- or even three-dimensional and would count once only for the whole story. One-dimensional characters, on the other hand, are counted not by the number of characters but by the number of scenes that use one-dimensional characters regardless of the number or identity of those characters. Thus, a scene with one policeman directing snarled traffic and many angry drivers counts as one instance of a scene with one-dimensional characters, even if a two- or three-dimensional character also happens to be in that scene. If the same policeman occurs in a later scene on a now empty street, that scene also counts as another one instance. But if, say through varying uses in several scenes, that policeman becomes sufficiently well defined to be a two-dimensional character, the policeman adds to the count of two-dimensional characters. In that instance, the first scene with the traffic still counts as one instance of the use of one-dimensional characters because of the drivers; however, the second scene, with our now more substantial policeman standing alone, would not. As you read, write down the names of all characters and list the instances of one-dimensional characters. When you are done reading, you should decide which characters, if any, are two-dimensional, which three-dimensional. Then count the number in each category and the number of one-dimensional instances. Record those three numbers in their appropriate places in the text database.

Q: How do I know when a one-dimensional character is on the scene?

A: Although this question may raise difficult matters of interpretation in some instances, as a rule of thumb the scene counts as using a one-dimensional character if the character in question is to be visualized but not if the character is merely referred to. Thus, if one character tells another a story about her parents and within that story the parents are visualized for the reader, they count. On the other hand, if one character says to another, "My brother always used to say exaclty what you just said" but otherwise the story makes no reference to the brother, we do not visualize the brother, the brother does not function as a one-dimensional character, and so the scene in which he is mentioned does not count.

Q: What is meant by "Narrative Person"?

A: If you don't remember your first-, second- and third-person narrative voices from high school English, do some research at the library or online, or feel free to ask another team member for clarification.

Q: What is meant by "Narrative Tense"?

A: Again, recall your high school English, do some research, or ask another team member.

Q: What are "Recurring Elements"?

A: Authors may write a series of short stories utilizing the same character, setting, invention, etc. In this field, which you will encounter at the end of the Text Database (it is not on Jon's Text Database fields and values worksheet), you will note if there are recurring elements in this story and what those elements are. The way to know this is from personal experience of reading a story beforehand with the recurring element or from context information, such as might be found in the editor's preface to the story.

Q: What should I do about entering information about reprints or, if I’m reading a reprint, the original publication of a story?

A: This is a crucial, straightforward, but sometimes complicated task. Our goal in entering information about reprints as well as original publication of each story read is to enable study of the characteristics of a story that increase its likelihood of being reprinted. In Darwinian terms, since reprinting is a clear form of reproduction, this allows for the determination of what makes a story “fit” to survive.

Initial Text Record Entry: Begin by entering the information for the story just read into the text database. Since the text entry asks for both the author id and context id, it is recommended that these be tackled first.

I. Enter author id for the story just read. To do this, search for the author in the author database. If necessary, create a new author record, save the record, reopen the record, and find its automatically generated author id Report any new author id created to gecontextdb@umich.edu.

II. Enter context id for the story just read. To do this, search for the context in the context database. If necessary, create a new context record, save the record, reopen the record, and find its automatically generated context id Report any new context id created to gecontextdb@umich.edu.

III. Enter the rest of the text record information. The text id will be generated automatically once you save the record.

For more information on any of the three previous steps, please read the appropriate FAQ’s and the Newcomers’ Guide.

Subsequent Text Record Entries: For information on any given text to be complete, all the reprints of that text must be entered into the text database.

  1. Search Contento’s Index for the story title as reprinted through 1983:
    http://users.ev1.net/~homeville/isfac/0start.htm
  2. On finding other contexts, search the context database for them.
  3. If necessary enter the new context into the database. (Note: if the new context has an editor or editorial team not already present in the author database, the creation of the new context record will require first the creation of a new author record in order to have an author id for the editor or editorial team. Create that author record, save it, and retrieve it to find its automatically generated author id. Use that author id in the author id field of the new context record and to notify gecontextdb@umich.edu about the creation of any new author record.)
  4. Report all context entries that are incomplete to gecontextdb@umich.edu.
  5. On the text database, call up the original record you entered for this story and use the Duplicate Record button to create a duplicate record
  6. Retrieve this duplicate. (You will be able to distinguish the duplicate from your original record by the duplicate’s automatically generated new text id)
  7. Edit the duplicate record to change the context id and, if necessary, the Reprint Yes/No selection, and then resave this newly edited entry.
  8. Repeat steps 2-7 for each new context found.
  9. Search The Locus Index for the story title as reprinted 1984-1998.
    http://www.locusmag.com/index/0start.html
  10. Repeat steps 2-7 for each new context found.
  11. Search The Locus Index for 1999, 2000, 2001. There are separate links on the main Locus web page at the bottom.
  12. Repeat steps 2-7 for each new context found.
  13. One last check of the Internet Speculative Fiction DataBase for reprints that the other sources may have missed.
    http://www.sfsite.com/isfdb/
  14. Repeat steps 2-7 for each new context found if necessary.
  15. You can combine all reports of new authors, new contexts, and new texts entered in any given posting session on one message to gecontextdb@umich.edu. This message should indicate clearly what was created, its id, and who created it.

top

Author Database

Q: What do I do when there are two authors for a story?

A: First, search the text database for each author. If, for either author, there is a record with the AuthMultiple checked, then check to see if the other author listed in that record is the second author in the story you are attempting to code. If it is, then use that Author ID for the story. If no Author ID currently exists that represents both of the authors in the story you are attempting to code, make a new author record. For the AuthFirstName and AuthLastName, enter the information for the author whose name comes first alphabetically. In Biographical Notes, enter the other authors name(s...if more than one).

Q: What information should I enter about a multiple author?

A: Enter only the name of one of the authors in the team under AuthFirstName and AuthLastName and the other member(s) of the team in the Biographical Notes. Do not enter biographical data about the members of the team in the MultAuth Yes record; those data go with the individual author records.

Q: What do I enter if the author's name is a pseudonym or a housename?

A: Enter the pseudonym or house name in the AuthName fields and select Pseud Y. The real (legal, mundane, etc) name of the author or authors is entered in the AuthRealName field. (Thus, when searching for stories by a certain author who may have occasionally published under a pseudonym or house name, one needs to search both the AuthName fields and the AuthRealName field to make sure one is getting all records for that author.)

Q: How do I enter an author’s date of birth if I do not have complete information (the day, for example).

A: We have agreed to use 01/01/yyyy as the birthdate for an author for whom we have only a year of birth and mm/01/yyyy if we have only month and year.

Q: What should I enter for an author's name if that name varies for publication to publication or contains titles or other features, such a E. E. "Doc" Smith, Ph.D?

A: Use the AuthLastName field for last name only and AuthFirstName field for all other name data, as indiciated here:

+ Check the Author database by author last name for an existing record. If you find one, follow its conventions.

+ If no record exists, create one putting the last name in the AuthLastName field and all other data, starting with the most commonly used first name, in the AuthFirstName field, e.g., AuthLastName = Smith; AuthFirstName = E. E. "Doc", Ph.D. Although Smith was named Edward Elmer, he did not publish under his given names but under his initials, so you should use those.

Q: What do I do if I'm in doubt about the name an author usually used for publishing?

A: E-mail your problem to the group (genreevo@umich.edu) so that the person responsible for the Author Database can supply the correct answer.

Q: What country should I list if the author has not lived and worked his/her whole life in one country?

A: Enter the country with which the author is most closely associated, usually the country of residence while writing but not always. Isaac Asimov, born in the Soviet Union and brought to the US at age three, ought to be entered as US. If in doubt, e-mail your problem to the group (genreevo@umich.edu) so that the person responsible for the Author Database can help.

Q: What should I enter if an author has more than one working name (as opposed to pseudonym), for example, Theodore Sturgeon and Ted Sturgeon?

A: Don't worry. In the Text Database, the Author will be identified by AuthID and found by looking up the author's last name in the Author Database. In entering the data in the Author Database, include the most common working name (if you know it) in the name field and the variants in the biographical notes.

top

Context Database

Q: What does it mean if there are two different Context IDs for a story?

A: A story has more than one context if it has been reprinted. A context includes a date and a medium. If the work is read in an anthology, the Context ID should be for that anthology. Then, if we want to have an entry for the work as originally published in, say, AMZ, we should have a second text record for the story, that record ought to have its own Text ID, and everything else ought to be the same in that text record except for the Context ID, which would refer to the appropriate issue of AMZ.. If a needed Context ID doesn't exist, the researcher ought to create it.

Q: What initialisms should I use for SF magazines?

A: Use the standard abbreviations found in such reference works as Contento's Index to Science Fiction Anthologies and Collections.

Q: To what do the field names in the Context database refer?

A: All fields in the Context database refer to the work in which a story appears, not to a story itself. Therefore:
ContextID is generated automatically when you first save a new context record for a magazine or anthology.
ContextDate is the date of publication of the magazine or anthology.
AuthorID is the AuthorID of the editor of the magazine or anthology.
Work fields (...Title, ...Type, ...Format, ...PubPlace) hold data for the magazine or anthology.
ModalAud fields (...Age, ...SES [SocioEconomic Status], ...Sex) reflect your inferences from the binding, format, paper type, illustrations, ads, editorials, letters to the editor, typical included story, or anything else you notice that could indicate the Modal Audience for whom the work was published. "Modal" is a terms from statistics. It is contrasted with "Mean" and "Median." Let us assume that our research group is a population for analysis and that of fifteen student researchers 10 are 19 years of age and 5 are 21 while 5 faculty are each 50. The mean of a population is the numerical average (which would 27.25 years in this case). The median of a population is the calculated central value above which and below which we have the same number of members of the population (which would be 20 years in this case). The mode of a population is a calculated value representing the most frequently occurring value (which would be 19 years in this case). The following quote is found in the Oxford English Dictionary: 1895 K. Pearson in Phil. Trans. R. Soc. A. CLXXXVI. 345, "I have found it convenient to use the term mode for the abscissa corresponding to the ordinate of maximum frequency. Thus the `mean', the `mode', and the `median' have all distinct characters." Our database aims to capture information about modes.

Q: Who should be listed as the "author" of an author anthology?

A: If an editor other than the author of the pieces is explicitly identified, that editor's AuthID should be entered in the Context Record. Also, if that AuthorID does not already record the person's having functioned as an editor, check Yes in the Editor field in the Author Record. If no editor is explicitly identified, enter the AuthID of the author of the pieces in the author anthology in the appropriate place in the anthology's Context Record. However, self-editing does not by itself warrant our recording that an author also worked as an editor. Therefore, if no editor other than the author of the pieces is explicitly identified, do not change the contents of the Editor field in the Author Record.