Sustainable Food Laboratory

¹ Because of the unusual research approach undertaken in this project, this section of the report offers a review of methods used to conduct the research, as well as the rationales for these approaches. In addition, this section includes a discussion of some special difficulties encountered in this process, and the learning they yield for future international communications research efforts.

French Elicitations
As a first step in identifying explanatory models for the French context, Cultural Logic undertook a series of twenty “elicitations” with French citizens – in-depth, one-on-one interviews, conducted and analyzed according to principles of psychological anthropology, ethnography and cognitive linguistics. The purpose of the elicitations was to provide the researchers with insights into French patterns of thinking not available from the earlier focus groups supported by the King Baudouin Foundation. Those focus group reports provided an important source of insights going into this round of research, but nonetheless, one-on-one interviews consistently lead to different views from those afforded by the group context. Moreover, the kind of cognitive analysis performed by Cultural Logic is, in any case, rather different in its focus from more traditional kinds of opinion research.
Elicitations conversations have the following characteristics:

• They are semi-structured, meaning that they are based on a predetermined set of questions, but are also loose enough to explore interesting and important perspectives that emerge.
• They are meant to approximate the flow of natural conversation.
• They approach a topic from many different angles, including some that are deliberately surprising or challenging.
• They create many opportunities for research subjects to think aloud – for example by putting the subject in a situation where s/he has to “teach” or “convince” the interviewer – rather than simply repeating familiar ideas.

The conversations are audio-taped for later analysis, at which stage the researchers are most interested in topics such as:

• Significant “gaps” in public understanding related to sustainable food systems
• Patterns of thinking (typically implicit or unconscious) that could interfere with learning and productive engagement
• Understandings that might be productively built upon

Cultural Logic created the discussion protocol, provided guidance regarding the elicitations method, and conducted analysis of transcripts. The elicitations were conducted by a French sociologist (Lucie Philippeau), who also provided analytical input into patterns observed in the interviews .

Elicitations subjects
Subjects were recruited through a process of ethnographic networking—the interviewer began with “seed contacts” in a set of target communities, and developed a pool of subjects from which a diverse range was selected for interviewing. 
Demographics for the subjects were the following:

• Location: 10 in Île de France, 3 in Lyon and surrounding area, 3 in Lille and surrounding area, 4 in Angers and surrounding area
• Metro vs. Rural: 9 residents of rural areas, 11 residents of metro areas
• Gender: 9 women, 11 men
• Age: 4 in 20s, 4 in 30s, 4 in 40s, 4 in 50s, 4 60+

All subjects were French citizens and native or native-like speakers of French.

Generating material for testing
Since this work was explicitly designed to build on the American effort described earlier, the initial material consisted of “translations” – in both a linguistic and cultural sense – of the most successful material emerging from the US research. As data was gathered, materials were refined, and new approaches were developed, in consultation with French communications professionals (including the staff of Patrick Mathieu Conseil, a Paris-based consultancy), French scholars (including Michel Achard, linguist, Rice University) and other native French speakers. In all cases, the goal was to find new ways of expressing key topics that had more capacity to teach, to be remembered, used and passed along.

TalkBack
The chief methodology of the project was what Cultural Logic calls “TalkBack” testing, in which subjects are presented with an explanatory message and then asked to perform various tasks, the most important of which is to pass along the message to another lay person – or to repeat it as though they were passing it along.
Subjects for TalkBack were selected either via the type of networking described above (see discussion of Elicitations) or were selected at random using online telephone directories.
Most of the 400-plus TalkBack subjects participated in telephone calls of 5-10 minutes. During these calls, the researcher first read a paragraph of roughly one hundred words, such as the following:

Les scientifiques qui étudient la production alimentaire parlent du « Test Structurale» pour mesurer l’impact de l’agriculture et de la peche sur les éléments porteurs de ce système alimentaire. Le Test Structural est très important parce que nos méthodes intensives d’agriculture et de pêche peuvent avoir un impact important – et souvent invisible – sur la stabilité du système alimentaire. En effet, les écosystèmes des sols, des rivières, et de la mer sont comme des murs porteurs: Souvent cachés, ces murs porteurs doivent rester sains et solides pour empecher un effondrement du système de production alimentaire. Le “Test Structural” est la seule façon d’assurer la stabilité des structures naturelles qui soutiennent la vie sur notre planète.

Subjects were then asked several questions, such as:

• Could you please repeat the paragraph as though you were playing the children’s game of “gossip” or “telephone”and doing your best to pass it along to someone else?
• Why is this information important?
• What is the most important idea in the paragraph?
• How do the ideas in the paragraph apply to the topic of … ? (e.g. overfishing, chemical fertilizers, etc.)

The TalkBack method is a surprisingly difficult test for any message. While there is never the expectation that subjects might recall the paragraph in close to its original form, it turns out that it is usually difficult or impossible for subjects to even recall central ideas or key terms. The topic area, and the target propositions, are so unfamiliar and unexpected – and are competing with such well-established ideas in people’s minds – that the most typical responses simply restate common and familiar ideas that are only tangential to, or even in conflict with the paragraph. (See Other Findings for further discussion of TalkBack results.)
While the test is difficult, however, it is extremely important. If lay people cannot remember and use key ideas and terms, these ideas have little chance of entering and changing the public conversation on the issues.

TalkBack “Chains”
A small number of the subjects (roughly thirty) participated in in-person TalkBack “Chains,” which are videotaped for further analysis. In this exercise, a subject hears a paragraph like the ones referred to above, and is then asked to pass it along to a new subject who is brought into the room. The second subject then passes it along to a third subject. This method, for obvious reasons, represents an even more difficult test for any potential simplifying model than the “single-generation” TalkBack conversations conducted by telephone.

Negotiating several levels of analysis
In considering this research and its results, it is critical to bear in mind a key three-way distinction:

• Expert Proposition: the idea, however simple or complex – that is to be communicated
• Explanatory Conceptual Model: the chosen analogy or other conceptual aid used to convey the expert proposition
• Verbal Expression: the exact words used to convey the explanation (Note that the verbal expression is often intended to evoke a particular image, but that this research did not include testing of actual visual representations.)

Each of these represents an important area for research:

• Expert Proposition

As discussed above, it is far from obvious exactly which expert propositions are most important for promoting productive reasoning; how generally or particularly these ideas should be treated; which ones are hardest or easiest for lay people to grasp. Should communications focus on the importance of diverse crops? On threats to the actual production of food in the future, as opposed to more general threats to ecological or environmental systems? On the role of corporations in creating current problems? On agriculture or fishing? And so forth. Important aspects of the research and analysis relate to questions like these.

• Explanatory Conceptual Model

For any given Expert Proposition, there are any number of ways it might be explained, on a conceptual level. For instance, is an unbridled food production system best imagined as an Out Of Control Train? As an Escalation of Destructive Methods? As an Animal That Has Escaped Its Leash? Etc.

• Verbal Expression

It is not clear in advance which words are best for evoking any particular explanatory model. There may be a variety of words and phrases that can convey a given analogy, for instance.

The larger context of this research effort – an attempt to forge a set of basic concepts that are viable across national and linguistic boundaries – dictated a focus on the first and second levels, rather than on the specific language that is most effective in the French communications setting. At the same time, it proves impossible in practice to identify Expert Propositions or Conceptual Models without identifying one or more effective ways of conveying them (i.e. Verbal Expressions.)