Close-Text Analysis: Using Lists to Expand Meaning

In this article, Eric Haas introduces a set of tools commonly used in frame analysis, giving particular emphasis to the study of written text and the power of lists for shaping how a political issue is understood.  Lists are a common and powerful way to promote a worldview.  This is the first in a series of monthly discussions about tools and techniques used by Cognitive Policy Works staff to inform political and social change processes.

Tools for Frame Analysis

There are a number of tools for analyzing the frames we and others use automatically in thinking and communicating about issues that are important to us.  They can be used in situations that concern politics, such as health care, education and the role of government, or in our private lives, like how we feel our should children behave.

The broad categories of analytic tool types include:

  • Direct participant response analysis where people are engaged directly to study their responses to different frames. Examples include opinion polls and focus groups in the context of an advocacy campaign, and;
  • Text analysis where written text is analyzed to deconstruct meanings that appear in print media.  Examples are given below.

Text analysis can be broken down into two domains:

  1. Large-scale analysis targets a large data set to find recurring patterns in public discourse.  Examples include looking for specific words or phrases across thousands of newspaper articles, TV shows or a year’s total of your organization’s press releases to identify competing narratives (in the press or internally within your organization) and measure how often they appear in the press or your publications over time.
  2. Close-text analysis dissects the language use in one or a small number of important documents to identify how a key actor is framing an issue.  Examples include deconstruction of frames in the President’s State of Union address or prominent articles on an organization’s website.

Cognitive Policy Works (CPW) uses all these tools in its work, depending on an organization’s goals. For example, organizations can use an overview of the prominent understandings or different “common senses” around a specific political issue by society at large or among a target group, such as their state legislators, to determine the difference between these groups and their own understanding of the issue. Once they know this, then they can develop a strategy to effectively promote their positions to a large audience. To make this determination, CPW can do a large-scale text analysis of newspaper articles on the topic or analyze the websites and speeches of their state legislators. We might also combine one of these large-scale text analyses with a poll to compare the news or legislator results with the understandings of people in general.

Organizations are also interested in knowing the subtle structures of their own thinking and communications. This knowledge enables organizations to hone their thinking and more clearly present their ideas and positions to constituents. CPW can do a close-text analysis of a selection of an organization’s press releases or website content, to illuminate the implicit messages that their word choices reveal. We might then follow up on all these analyses with trainings to help the organization incorporate key insights into their everyday work, which often entails learning new techniques and – in some instances – subtle shifts in how the organization goes about its business.

We will describe more about the methods we use to serve clients in the coming months.  For now, let’s look at one tool used in close-text analysis to reveal a hidden agenda.

How Lists Expand A Particular Meaning

In our inaugural tool description, we want to focus on close-text analysis and one specific language device, the list. The list is both commonly used and powerful for its ability to emphasize a particular perspective while marginalizing others.  What is a list?  Here’s a working definition:

A list is the sequencing of words together so that they brain automatically places all of them into the same category, with a presumed set of shared features that make all of them similar to each other.

Here is an example about education, taken from an article in the Boston Globe:

Education policy in the United States treats Americans as too incompetent to provide for their children’s schooling. Unlike food or clothing or health care — where the market generates lots of options and parents are free to choose among them — education is mostly supplied on the Soviet model: Schooling is “free,” but the schools are owned and operated by the state. A small fraction of parents pay to educate their children privately, but the great majority simply take what the state supplies.

The public education system is essentially a monopoly, and like most monopolies, it wastes money, performs indifferently, and doesn’t much care if its customers — American mothers and fathers — are satisfied. But give those mothers and fathers the same freedom of choice when it comes to their kids’ education that they have when it comes to their kids’ shoes or dinner, and all of that would change. (Jacoby, 2004, p. D11, bold added)

Proponents of market policies for education often use the list. Why? Because a list triggers your brain to connect the dots between the objects placed together by the characteristic suggested by the communicator (in this case, the article writer). Proponents of market policies want to promote interactive, semi-social goods (like education and health care) as if they are nothing more than commodities that follow the laws of consumer transactions.  They seek to lump these public goods in the same category as shoes, televisions, and salad dressings (all of which are pretty interchangeable and it really makes no difference to the general prosperity which one is the rage).

Even more subtly, pro-market proponents want to present education as a purely individual decision: it’s not what choices society provides that matters, it’s what you choose to buy. The blame is on the parents, with no societal responsibility, for making bad parenting decisions. Following this logic further, the government would have little or no obligation to ensure that good schools were available to everyone, just like it has no obligation to ensure that good quality supermarkets are present in every neighborhood.

Of course, a long discussion like this is not enjoyable to read. It’s also unnecessary. Our brains do this automatically for us by reading the list: shoes + clothing + food +dinner + education.

So, what would a list look like that promotes strong government support for public schools in all neighborhoods?  What if we listed schools with roads, electric power lines, the internet, mass transit, fire fighters, and national parks: all things that we need for us to prosper (electricity sure comes in handy); things that benefit us when they help our neighbors (my drive to work is better when more people take buses and the subway); and each requires large scale cooperation to be effective (what if fire fighters returned to being privately contracted to some but not all houses?).

Which list you put public schools in says a lot about your thinking. Your list also promotes a subtle, but strong, difference about what you say public schools are about and how they should be supported.

References

Haas, E. (2006). Civil right, noble cause, and Trojan Horse: News portrayals of vouchers and urban education, pp. 439 – 450. In J. Kincheloe, K. Hayes, K. Rose, & P. M. Anderson (Eds.), The Praeger Handbook of Urban Education, Volume II. Westport, CT: Greenwood Press.

Jacoby, J. (2004, May 30). Vouchers and equal education.  Boston Globe, D11. Available on the web at http://www.boston.com/news/globe/editorial_opinion/oped/articles/2004/05/30/vouchers_and_equal_education/.

——–

Enjoy this article?  Perhaps you’d like to sign up for our newsletter.

——–

Cognitive Policy Works specializes in providing organizations and individuals with frame analysis, policy briefs, strategic advising, and training.