Project Homepage: http://www.umich.edu/~genreevo
The core of work done for the current phase of the Genre Evolution Project (GEP) is reading issues of science fiction magazines. The stories are read and coded according to a pre-specified set of criteria designed by the GEP team over the years. Other material in the magazines is noted for the purpose of determining context (to be entered in the Context Database). This other material would include cover art, interior art, letters from the readers, editorial, advertisements and regular departments.
Reading assignments come in the form of magazine title, month / year or sometimes week / month / year. The magazine title abbreviations come from a list by William Contento; the ones we most commonly use can be found on the Project Homepage, in the Data Entry FAQ under the information about the Context Database. Most magazines read are on microfilm, which usually means taking a trip to the U-M Graduate Library.
Step 1 MIRLYN SEARCH
Upon receiving your weekly reading assignment, go to the second floor Graduate Library (don't hesitate to ask an older member to accompany you as it will make things go much more smoothly). First look up the title of your magazine on Mirlyn and find a microfilm number (begins with X generally). If this confuses you, or you can’t find what you need on the computer or on the shelf, ask at the Information Desk. Those people are paid to be nice and helpful.
Step 2 FIND REEL
Once you have the catalogue number, go to the microfilm stacks on Second Floor South of the Graduate Library and find your reel. Be careful to note the month and year you need. You will discover, over time, that most everything you read for the GEP is in the same place. If your reel is not there, again, ask at the Information Desk. If it is truly lost, they will do a library-wide search on the item.
The person making assignments does his/her best not to assign items that the library doesn’t have or issues that don’t exist but sometimes there are snafus. This would be a good reason not to wait until the last minute to do your reading. You must give the person making assignments time to respond to the problem and to give you a new assignment for the week. Do not just read from a different reel, until instructed to do so. The assignment maker needs to keep track of what we have and have not read. Also, do not skip your reading for the week. The database needs to be constantly and consistently added to.
Step 3 MICROFILM
Reel in hand, go over to the microfilm machines. There is no way to describe here how to operate one of these monstrosities if you have never done so before. Some have directions right there for you. If that and intuition does not suffice, ask at the Information Desk. They are used to being asked for help concerning this.
Step 4 READ
First and foremost, bring your coding packet, which includes all definitions, trees, and coding sheets. You do not have to do a whole issue in one sitting, but you should plan your time in order to complete the entire reading assignment within two weeks. Reading assignments vary for each reading group.
Do not read stories that are more than 12,000 words (or 30 pages at 400 words per page). Magazines will list longer stories in their table of contents as novellas or novelets but sometimes stories listed that way are actually less than 12,000 words. Thus, you still need to do a word count for all stories in a magazine to determine what you need to read that week.
Make notes as you are reading. Be sure to write down the name of the editor of your issue or issues as it will come in handy while entering your codes into the database later. Also, jot down names of three- and two-dimensional characters as well as instances of one-dimensional characters. Make note of the setting and time and anything else you want to remember later, while coding. When you come across a reprinted story, write down whatever information you are given about the reprint, especially the year of the original publication date and where it was first published, if given. One convenient way to keep notes is to use the back of a printout of the coding sheet mentioned in the next step.
Step 5 CODE
Code each story directly after you read it. For one thing, the story will be fresh in your mind, making this the best time to pull it apart. Second, this is a chance to give your eyes a break from reading the grainy microfilm screens (or just reading in general).
Use Zach Wright's coding sheet, which can also be found the coding packet. When printing sheets you probably will want to print two pages to a side or print on both sides in order to save the paper on your umich account for endless term papers and lecture notes.
Coding packet--trees, defintions, coding sheets--for most effective story
Be sure to pay close attention to the coding packet and review it periodically throughout your stay on the project. It is an infinitely helpful resource to the newcomer as it explains how each field should be handled. BUT it is not a fixed document! A newcomer’s fresh eye can often spot things that are vague, confusing, or completely unexplained. Please bring up your questions or concerns at the weekly meetings. This is part of the purpose of the meetings. No question is too basic. Feel free to e-mail questions to the group.
Other things to keep in mind besides the story itself:
Note any biographical information given about the author. This will be helpful later (explained).
Use the coding sheet to jot down all the information you’ll later need—about the author, the magazine and the editor of the magazine—to make a Context Record while the magazine is right in front of you. For information on how to make sense of the Context Database fields and values, see the Data Entry FAQ under Context Database and these helpful pages Amoreena Gonzales-Ralya put together. You want to look for topical and historical notes of interest as well. This would include ideas, events, trends or phenomenon occurring during the period when the magazine was released. This information is commonly coded into the Context Database, but you could also include a note in the Story Notes field or Salient Features field when you enter stories into the database.
How to Make a Context Database Entry:
Step 6 MEET
When you have all of your reading and coding completed, you are ready to meet with the rest of your reading group to compare coding. Data from the coding can be entered directly into the appropriate databases as your group goes through the stories; otherwise you can divide the work up and do it on your own time. Setting up a regular time and day to meet every two weeks, if all your group members have set schedules, is a good way to make sure your coding gets done on time.
Step 7 ENTERING CODES INTO THE DATABASE
a. Go to the Project Homepage (www.umich.edu/~genreevo) when you are ready to enter your coding into the databases. There is a "Database" hyperlink at the top—click there. An older member of the GEP group should give you a login name and password.
b. Click on "Add Record-Text". Then click on Lookup or Add Context. Use the editor and magazine information you jotted down earlier to make this record. When the record has been completed, go back to "Search Context Database" and look for the record you just completed by name and date of publication. Write down the CONTEXT ID number; you will need it for every story from this issue that you enter into the Text Database. You will make only one Context record for each magazine read.
We use Contento’s abbreviations, mentioned earlier, when entering magazine titles. FSF is The Magazine of Fantasy and Science Fiction, AMZ is Amazing Stories, and so forth. Consult the Data Entry FAQ under Context Database for a more complete list of the abbreviations we use, or for the full list, see Contento’s Index to Science Fiction Anthologies and Collections, that can be found in the reference area of the University of Michigan Graduate Library.
You may not have complete information for your Context record and that’s forgivable. There is only so much microfilm—black, white, and grainy—can impart. You may not be able to tell if the cover was originally in color, for example. In this case you would send the Context ID number and any other relevant information to your project managers.
c. Choose one story from that issue to start with and click on Lookup or Add Author. (NOTE: In the Author Database, there should be separate AuthorIDs for an author publishing singly and for the same author publishing with any given collaborator or using a pseudonym. Make sure to check for the appropriate AuthorID. In confusing cases, say authors with very similar names, search in other fields such as the Biographical Data field to confirm whether or not the author in question is already recorded in our database. Except for minor variants, our text records should reflect an AuthorID for every separate publishing name or combination of names, whether real names or pseudonyms; that is, we should not have one AuthorID for "Arthur C. Clarke" and another for "Arthur Clarke" or for "Arthur C. Clark" but we should have another AuthorID for Clarke publishing with Gregory Benford and yet another AuthorID for Gregory Benford publishing on his own and yet another AuthorID for Gregory Benford publishing under a pseudonym. Sometimes, though people make mistakes and incorrectly enter two IDs, i.e. “Arthur C. Clarke and “Arthur Clarke” as two different people. Make sure you search very thoroughly and not necessarily with the exact name of the author as printed in your magazine. If you notice a problem such as this please tell your project manager.)
If you cannot find an author in the Author database after very careful and thorough searching, go back to the hyperlink Lookup or Add Author. Enter the publishing name (regardless of whether or not it is a real name or a pseudonym) and any other biographical information that you noted down earlier. When you complete the record and submit it, go back to "Search Author Database" and look up that author’s name again. Enter the AuthID from your newly created record. This is the AuthID that you should enter when making a new text record. Keep track of the new author records you create (name, AuthID, and date of issue published in) to send to the project manager after you’re done. That person will fill in the biographical information that you did not have. .
If you successfully find a publishing name in the Author database, note its appropriate AuthID number and enter it in the AuthID field. Do not fill in any more author data in the text record. The databases will do that automatically based on the AuthID you filled in.
d. Finally, enter the story data. Make sure you type carefully. Enter the exact title as it appeared in the magazine. Enter the appropriate Original Publication Date. Leave Text ID blank, this will be entered automatically after you complete and submit the record. Do not enter the Author’s Name or the name of the context. The database will do this automatically given the correct Author ID and Context ID numbers.
Unless noted here, the rest of the fields should be self-explanatory and definitely completed for every record. Theme B, Reader Thoughts, and Salient Features do not have to be entered if they do not apply but it is common to put anything of note concerning a story in the Reader Thoughts field. You must put a first and a second reader in each record. If there are more than two readers, include the other readers’ names in the Reader Thoughts field.
You are done. Click Submit and restart the process. Remember to make new Context IDs and Author IDs, where appropriate, when continuing your coding.
Step 8 CLEAN-UP
Occasionally mistakes are made. Records can be edited if later it is discovered that a, for example, Second Reader was left off or if a Sentence Style check box was left unmarked. This is a good reason to save your notes on the stories. The fewer mistakes we make, the fewer holes get into our data set and the less time we have to take away from reading and coding to fill them in. It is considered good form to go over your Context and Text Database after you enter them. For more intensive information on how to sort data for cleaning, see this page:
Step 9 GEP MEETINGS
The Genre Evolution Project research team—students, faculty, and any others currently working with us—meet every Monday for an hour and a half in the evening at Angell Hall on the U-M Ann Arbor campus (the actual time may alter slightly from semester to semester). These meetings are vital and never to be missed. It is then that we do group coding so as to remain cohesive and consistent in our individual coding; questions big and small, are brought to table; people working on special assignments report on their progress; results of data analysis are shared and discussed. Most importantly, we stay in touch with each other.
If you have further questions, anyone would be happy to help. Consult the Team Members page for individual email addresses and phone numbers. Or email the whole group.
Team Members page:
Entire GEP (past and present members): firstname.lastname@example.org
Only currently attending members: email@example.com
Even if you forget everything else we’ve taught you, if you hang onto these then you’ll be OK.
1. CODING IS #1. Keep upon your knowledge of our literary definitions. Resist the urge to use street definitions.
2. THE TREES ARE YOUR FRIENDS. They will help you with inter-coder reliability (ICR).
3. DON’T HESITATE TO ASK IF YOU DON’T KNOW. We tend to work fast, in some ways.
4. DON’T BE AFRAID TO SUGGEST CHANGES—to the coding packet or for new project ideas.
6. ALWAYS HAVE THE PACKET WITH YOU FOR CODING AND MEETINGS.
7. RE-CHECK ALL DATA ENTERED, ESPECIALLY DATES.
8. DON’T BE AFRAID TO DEBATE WITH YOUR PARTNER OR THE GROUP. Even older members can be wrong. Fresh minds teach old dogs to break out of their coding patterns.
With Help From:
& the Fall 99 Genre Evolution Project Team
& and revised by many hands thereafter , including
Winter 04 Stephanie Wentworth & Zach Wright