Developing Genre Ecology Models
Clay Spinuzzi, University of Texas at Austin
At the meso level, actions are
- Tactical
- Medium to short term (on the order of minutes or hours)
- Goal-oriented
As we saw in the ASD, these actions are mediated by different tools and resources, especially texts. The ASD lists many of these tools at a general level. But it leaves us with several questions:
- How do people actually use these tools, in what combinations?
- Which tools are established parts of the work, and which are workarounds or innovations?
- How do the tools work together?
- How do people use different tools to accomplish the same things?
Genre Ecology Models (GEMs) are a way to answer those questions. GEMs let you map out the tools, especially the texts, that people are using at your site.
Genres are relatively stable responses to recurrent situations (see Spinuzzi, Tracing Genres through Organizations). Usually these are instantiated in texts or speech. Since they're regular and recurrent, we can recognize them, share them, and use them in predictable ways. They're a way to mediate our own work and the work of others. Consequently, examining genres tells us a lot about how people understand their work.
For instance, when we talk about a shopping list, we have a basic idea of what it might look like, based on what it's supposed to accomplish and on how similar texts have looked in the past. Shopping lists tend to be very different, as you'll see if you watch people shop at a grocery store, but they are similar and predictable enough that strangers could swap lists and probably do a pretty good job of filling each other's orders.
More specialized genres might be harder to understand and might take a bit more time to learn. Think about learning a new software interface or doing your taxes or filling out a travel request.
As you analyze how people accomplish things at your site, you'll look for the genres they use. Think in terms of
- text types that participants identify as types
- text types that are repeatedly used at the same site to accomplish the same things
To identify genres, look for any sort of text or exchange that participants used, composed, or placed: interfaces, forms, documents they wrote, calendars they consulted, notes they wrote on scraps of paper, checkmarks they made on a printout, even greeting cards that they have posted in their cubicles.
- Observations: Note any texts you see, especially texts that you see them read, compose, or annotate. You might use highlighters to highlight these in your field notes.
- Interviews: Ask participants about those texts. When do they use them? Why? Where did they get them? Don't be shy about asking about apparent mistakes or issues. And see if they make the same distinctions about texts as you do - if they refer to both handwritten lists and printed lists simply as "lists," that might tell you that they don't draw a distinction between the two.
- Artifacts: Examine the texts. Get copies of the texts your participants worked on or with. Get photos of their workspaces so you can see how texts are placed.
Now we have a good idea of the genres involved. Let's go to the next step: figuring out how they relate.
2. Identifying Relationships among Genres
As you look over your field notes and interviews, you'll notice that these genres are rarely - if ever - used on their own. For instance,
- One participant might use a printout of customers as a checklist when calling them, checking off the people they've called, then entering customer data into a spreadsheet.
- Another participant might consult documentation to understand how to fill out an unfamiliar form, and at the same time look through their email to get the content they fill in the form.
- A third might pull information from several databases in order to write a report for customers, using Instant Messaging to ask the account manager questions and a calendar to estimate how long she has before the report is due.
All of these scenarios are based on what I've seen in my own research - and they're typical. If anything, they're a little simple. Knowledge work involves a lot of texts, and we tend to use these texts in complex combinations. In fact, a set of texts lets us do things that individual texts won't let us do.
- The combination of a printout of customers plus a checklist plus a spreadsheet yields a system for tracking overdue accounts.
- The combination of documentation plus a form plus email yields a fairly accurate report.
- The combination of databases plus a report plus Instant Messaging plus a calendar yields information that the customer needs, delivered in time and with the right level of detail.
That is, when it comes to genres, the whole is greater than the sum of its parts. So we need to track the mediational relationships among genres: the ways that they work together to guide or enable people's work.
- Observations: Note how the texts interact, especially in terms of
- juxtaposition (two texts attached to or overlapping each other)
- placing (two texts placed side by side, in a stack, or in regular places)
- annotation (writing or altering a text)
- transfer (using one text as source for filling in another)
- modeling (using one text as a model for another)
- reference (using one text to interpret or operate another)
- Interviews: Ask participants about those texts' relationships. Why do they use these texts together? Why do they stack these sheets in this way? Where did they learn to make these sorts of marks? Where did they get the idea to use this printout as a checklist? Do other people do the same thing?
- Artifacts: Examine how texts are juxtaposed, placed, annotated, transferred, modeled, or referenced. Again, get copies of the texts in their relationship - and if you can't get these, at least get photos.
At this point, you have enough information that you can describe the genre ecology, the interrelated set of genres that are brought to bear at the research site.
3. Developing a Genre Ecology Model
Now that you have documented the genres at work and the relationships among them, you can depict these in a Genre Ecology Model (GEM). In a GEM, you map these texts, drawing lines between the ones that are used together -- the ones that co-mediate each other. Once you have this map of texts, analytical possibilities open up:
- At the individual level, you can compare how a participant uses texts in their work. At this level, you can examine how people marshal different genres to get things done, and especially how they bring their own innovations to their tasks. You can also map out the most central genres, the ones that link most densely to other genres; these tend to be heavily used and important for coordinating work.
- At the group level, you can compare how people in the same group, or in different groups, use texts. At this level, you can compare their ecologies to see how different people or groups tackle similar tasks with different sets of genres. These comparisons can help you detect different innovations, especially when different groups respond to the same issue with different solutions.
To construct a genre ecology model, start with a single set of field notes, artifacts, and interview.
- Get a single sheet of paper. In one corner, write the participant's name and the observation date so you can remember which participant you're describing.
- In the center of the paper, write the name of the most used genre. Circle it.
- Starting from there, write names of other genres that are associated with it (e.g., through juxtaposition, placing, annotation, transfer, modeling, or reference). Circle them too, and draw lines to connect each genre with the first genre. Make sure you can point to at least one piece of evidence for each link.
- Repeat the process for each of these genres.
At the end of the process, you should have a GEM that looks something like Figure 2.
You can repeat this process across a participant's observations, across the group, or across the organization. Consider generating separate GEMs whenever you see significant differences. These might include
- One participant who does things significantly differently from others.
- Two different roles (say, account managers and web developers in the same unit).
- Two different sections (say, two different units).
- Two different offices or branches (say, the uptown and downtown branches).
At this point, you should be able to show how people coordinate their texts and how that coordination does or does not differ. Now let's look at the discoordinations you may have detected.
4. Detecting Discoordinations
Often, genre ecologies suffer from action-level disturbances called discoordinations: points at which two or more different genres just don't "fit."
For instance, in Tracing Genres through Organizations, I examined how traffic safety workers used a database of traffic accidents and a set of 3'x3' maps to find patterns in traffic accidents. Unfortunately, the workers had a lot of trouble coordinating the two genres. The database could only search for intersections designated by six-digit "node numbers," which were printed on the maps, and "links" between those nodes. To find information about a particular intersection, the worker would have to look up the intersection or stretch of road on a map, find the appropriate node number(s), and type them into the database. They encountered all sorts of problems: they had a hard time finding the node numbers, they would transpose the numbers when typing into the database, and they would even have a hard time finding a large enough flat surface to spread out the map. And since they ran these searches only a couple of times a year, they weren't able to familiarize themselves with the system through repetition.
But some people didn't have as much difficulty. One police officer, for instance, didn't even bother with the node map. Instead, she simply took out a folder, opened it, and used sticky notes with node numbers written on them. She had to look at accidents for the same intersections each year, so years ago she had simply written the node numbers on the sticky notes. Now she didn't have to even glance at the map: she should put the sticky note next to the dialog box, juxtaposing the two genres and typing the node numbers in much more accurately.
In this case, workers encountered a discoordination between the two genres: they represented the roadway in entirely different ways (maps vs. nodes), with different logics and conventions, and workers had a hard time getting them coordinated. The police officer's innovation routed around this discoordination.
When you look for discoordinations, look for these characteristics in the data:
- Observations: Look for points at which people seem to have trouble relating two texts - or points at which an unexpected, informal text (like sticky notes) is substituted for another text.
- Interviews: Ask participants about these texts. Why do they think they're having trouble relating the texts? Why do they go "off-script," using an innovation or workaround rather than the official texts?
- Artifacts: Examine texts that are involved in discoordinations. What are their characteristics? Where did they come from? Do they represent different logics or conventions?
When you detect a discoordination, mark it in a different color of ink - I suggest red - on your GEM.