Influence of Narratives of Vision and Identity on Collective Behavior Change

Round 1

Reviewer 1 Report

Review of "Influence of Narratives of Vision and Identity on Collective Behaviour Change"

This is an interesting and potentially important study topic.

The literature review is adequate and well written. I was struck,however, by the fact that authors don't take an epistemological stance on narratives. There are interesting debates between the postmoderns and the Narrative Policy Framework on what narratives are and in particular whether they can be empirically measured. In terms of citations of the NPF, the authors should consider the NPF chapter in the most recent edition of Theories of the Policy Process. See the 2016 Jones and Radaelli article in Critical Policy Studies and most importantly the 2015 article by the same two authors (along with responses from the postmoderns in the issue). You will see that there are big debates over how to measure narrative and whether they can be measured at all. While the authors don't have to settle this debate, they do need to think about where they fall in the argument. 

 

The critique about literature then leads to the methods (or lack of). The reader is uncertain how these narratives were developed. Could another researcher replicate this study and find the same narratives? Does that matter? (this relates back to the NPF and postmodern argument over narrative). Perhaps, it does not matter or perhaps it does but the authors need to take a position on how they developed the narratives. I was also struck that the narratives don't necessarily include narrative elements. The NPF argues that a narrative needs at least one character and a moral to be a narrative. I think these are narratives but I don't see the traditional elements of a narrative. 

The narrative policy framework’s call for interpretivists (Jones and Radaelli) Article March 2016 Critical Policy Studies     The Narrative Policy Framework: child or monster? Article June 2015 Critical Policy Studies     Full-text requested

Author Response

The authors are grateful to reviewer 1 for very constructive questions. We appreciate the opportunity to respond specifically to the points raised in the review in the text below and with the uploaded version of the manuscript that has undergone major revision. A second version with track changes indicated is also uploaded.

A) What is our epistemological stance on narratives in regard to the literature

We added to and reorganized our working definition of narratives and literature of relevance in the section 2.1 on Narrative expressions of vision and identity on pages 3, 4 lines 73-168.

B) What do we measure in narratives?

If “measure” may be taken to mean to assess or evaluate, rather than only in a quantitative sense, then we took the measure of the influence of narrative expressions. We sought to understand the influence that narrative expressions had in the community in which the case was situated, i.e., qualitative evidence in case studies on the form, source(s), and types of influence of the narrative expressions.

C) Replicable and does it matter?

Based on our process of redundant independent groups examining cases and identifying narrative expressions in the case studies (section 2.4 Approach on page 6 lines 276-291), we have confidence that our method is replicable, which does indeed matter in our opinion. We have instituted only a limited test of this approach as yet, but we anticipate further studies in which the case studies can be designed a priori so that data will be richer and more specific.

D) Narrative elements missing

Key actors, scene, and moral were present and described in the three cases in section 3 Results, but primarily as historical figures and conditions, rather than being explicitly present and characterized as archetypes in the narrative expressions that circulated. Nonetheless, what we are calling the narrative in the TRCT in Micronesia lines (lines 445-446 page 10) or a CANE in “Lyannai” in Martinique (lines 347-348, page 8) evidently has or had influence in the community.

Reviewer 2 Report

Overall, the manuscript sets its sites on a very important and urgent issue, that of motivating collective action to address environmentally degradation and avert catastrophic damage to the ecosystem. The major theoretical components present in the study: 1) narratives (specifically CANEs), 2) affect, 3) identities, 4) networks, and 5) desirable futures seem to be reasonable components of an approach to the analysis of persuasive narratives, and the research cited to justify these components was appropriate.

 

The study also suggests a fascinating potential application for the outcome of a narrative analysis in the construction of agent-based models (ABMs) that could help to illuminate the outcomes of behaviors which were to change due to the impact of a narrative.

 

The four different cases were also quite interesting to read and added a great deal of useful background (context) that was helpful in situating the narratives.

 

There were, however, some significant issues that I believe would need to be addressed prior to my feeling comfortable with the publication of the manuscript. The majority of these issues, I think, are the result of the authors trying to do much, too much in this short manuscript, and the number of concepts employed are so numerous as to resist a coherent integration, and the result is a manuscript which is a bit overwhelming to read. In order to address these issues, I have included some suggestions below:

 

First, I felt that the inclusion of condensed affective narrative expressions (CANEs) was one of the most enticing aspects of the manuscript and it was given priority placement in the abstract. However, it was not developed throughout the manuscript. Given that the authors asserted that the “focus in this paper is on how narratives and particularly how concise affective narrative expressions (CANEs) can be used as a window into the dynamics of societal movements directed toward sustainable future” (p. 2 l. 56-8), this was puzzling (and slightly disappointing, I was really looking forward to seeing that concept more deeply explained, explored, and exploited to cast light on the narratives under study. This should either be dropped as a concept or clearly developed and explicitly tied to the ABM approach suggested in the final sections of the manuscript. CANEs need to be defined concretely and this definition should be presented early in the manuscript so that the audience can carry the concept through the cases studies, discussion, and conclusion. On page 6 (line 258) the authors refer to “the role of narratives and CANEs” this suggests that they are not equivalent and therefore I was unclear if CANEs are a subtype of narratives or if they are a separate construct that draws upon embedded cultural narratives (or, perhaps, some other relationship between the two). The authors use the example of the phrase “I have a dream” from Martin Luther King Jr.’s speech as a example of a CANE. However, it is not clear in the case studies what the CANE for each larger narrative actually was. In any case, greater clarity on the difference is key. Second, I felt that the treatment of affect was also insufficiently foregrounded. If affect is indeed of import, which the authors suggest it is, why is it not investigated in the cases studies? Additionally, the analytic categories developed by KLASICA during the symposium seemed to emerge out of nowhere. The categories were not tied to the literature which was reviewed in the Concepts and Approach section, and so there appearance and centrality in the case studies was especially jarring. The categories also didn’t seem to fit cleanly into the structure of the case study narrative reporting, which presented a Context section and then a Vision and Identity section. Where does Associative plausibility (and the other categories) link with the identity, affect, and desirable futures? Looking at Table 1, it seems like the Motivational incentives and better futures for the creole community, could be identical, but I’m unclear if they are intended to be. If so, why use two different names? If not, then what is the relationship of the concept to the analytic category? I have a slight quibble with the use of the phrase “test our hypotheses on the role of narratives” (line 258) simply because the paper did not put forth any hypothetical statement or even propositions (at least none that were explicitly stated). As such, I would resist the urge to use this type of terminology and employ different language. There is also the issue that the authors indicate that their next steps will be to collect more data and “use the resulting data in constructing ABMs to text [sic] our hypothesis…” (Lines 569-70). This suggests that the current paper, did not, in fact, text the hypothesis. The identities piece seems to be most consistently deployed throughout the manuscript to me, as it is tied into the conceptual section, and contained in the case study descriptions, as well. That said, it doesn’t seem to do much work in explaining what is happening in the narrative. For instance, the authors write about the creole garden as being “coupled with a strong cultural identity of the creole community” (lines 325-26) but it is unclear what that identity is, especially if we compare it to the identity invoked by the members of the Tamil community, which is “emphasized in reference to traditional stewardship” (lines 397-398). Are these examples of group identities? Is the group identity of a creole community self-referential? There is more work that needs to be done here theoretically and conceptually to be clear about what the identity is and how it works. Relatedly, characters feature prominently in the Concepts and Approach section, but disappear in the cases. Do the different characters deployed impact the identity that is constructed? Just very unclear about the role of identity other than that it is something that is being constructed/invoked in these narratives. The incorporation of networks in this piece is also a bit confusing to me. The authors spend time talking about the importance of networks to meaning creation, the study does little to explain how the network is practically impactful on narrative efficacy or efficient dissemination. It is, of course, understandable how networks would be an enlightening perspective to take when thinking about transitioning to ABMs, but the concept doesn’t seem to contribute too much to the current study. The “so what” question is in need of significant attention. After I read the manuscript, I was excited about the idea of using narrative-derived information to define the operations and parameters of ABMs, but I didn’t have a great “takeaway” from this study itself. I think that this may be due to the over-abundance of concepts, theories, and factors that have found their way into the manuscript. For example, the authors use some of the work in the Narrative Policy Framework to discuss the nature of narratives as “social constructions of realities and empirically measurable” (lines 147-48), but they don’t seem to take the second part of this to heart (at least in this study). The incorporation of narrative-derived data into ABMs, would be an approach that would share an affinity the NPF regarding its empirical-nature, (how else to code the agents in the model without quantifying the narrative data?), but in this study, the attention is focused on the context and the content of the narratives, which are clearly viewed by the NPF as NOT transferrable (in that they are categorized as narrative content and not part of the narrative form, which is structural and generalizable).

Author Response

The authors want to thank reviewer 2 for many very insightful questions that have helped us develop the ideas further. We appreciate the opportunity to respond specifically to the points raised in the review in the text below and with the uploaded version of the manuscript that has undergone major revision. A second version with track changes indicated is also uploaded.

A) CANEs: This should either be dropped as a concept or clearly developed and explicitly tied to the ABM approach suggested in the final sections of the manuscript. CANEs need to be defined concretely and this definition should be presented early in the manuscript so that the audience can carry the concept through the cases studies, discussion, and conclusion.

On page 6 (line 258) the authors refer to “the role of narratives and CANEs” this suggests that they are not equivalent and therefore I was unclear if CANEs are a subtype of narratives or if they are a separate construct that draws upon embedded cultural narratives (or, perhaps, some other relationship between the two). The authors use the example of the phrase “I have a dream” from Martin Luther King Jr.’s speech as a example of a CANE. However, it is not clear in the case studies what the CANE for each larger narrative actually was. In any case, greater clarity on the difference is key.

We have substantially added to the discussion of CANEs in 2.1 Narrative expressions of vision and identity. As the reviewer suggested, this is now presented early in the paper starting on line 122 page 3, where we also expanded the discussion to include more examples. We also link the concept and more general examples to the case studies by noting relevant CANEs in case study 1 (Caribbean), lines 347-348 on page 8 and case study 2 (Micronesia) on page 10 line 444 to illustrate instances in our cases. In these two cases, the sources of the CANEs as shorthand for a more complex and extended narratives embedded in specific cultural and historical contexts is noted.

In answer to the reviewer’s question, CANEs are a subset of narratives in that they are a condensed affective form which in many instances contains or represents the core or kernel of an extended narrative discourse. In other instances, the CANE emerges as a deliberately designed or an emergent slogan, image, song or other form. Either way, the CANE becomes the currency of acknowledging identity that circulates in a community and often into wider social spheres.

B) Affect: I felt that the treatment of affect was also insufficiently foregrounded. If affect is indeed of import, which the authors suggest it is, why is it not investigated in the cases studies?

We agree that affect is very important and should figure more prominently. However, it was not easily teased out explicitly in the case studies when only seen a posteriori as in the work described herein. Nonetheless, affect is an element of one of the analytical categories - emotional identification - in the three cases described. It figures as well in the examples of CANEs on page 3 line 122 to page 4 line 151.

Nostalgia, anger, fear, love or affection, for example, are affective responses that CANEs and sometimes more extended narratives specifically play on. Affect thereby influences social dynamics in the contexts of globalization, polarization, and xenophobia and is leading one of the authors (IC) into a study of this with colleagues. 

C) Analytical categories: Additionally, the analytic categories developed by KLASICA during the symposium seemed to emerge out of nowhere. The categories were not tied to the literature which was reviewed in the Concepts and Approach section, and so there appearance and centrality in the case studies was especially jarring. The categories also didn’t seem to fit cleanly into the structure of the case study narrative reporting, which presented a Context section and then a Vision and Identity section. Where does Associative plausibility (and the other categories) link with the identity, affect, and desirable futures?

We agree that in the previous version of the manuscript, the categories seemed to appear out of nowhere. We have rewritten this to describe the process of developing provisional categories in the KLASICA Taipei symposium. This is now done in Section 2.4 Approach on pages 6 and 7.

Looking at Table 1, it seems like the Motivational incentives and better futures for the creole community, could be identical, but I’m unclear if they are intended to be. If so, why use two different names? If not, then what is the relationship of the concept to the analytic category?

The point we are making is that the category “motivational incentives” - in this case for the Creole community - are contained in narrative expressions of the aspiration and stimuli for actions for a better future with environmental and social justice.

D) hypotheses: I have a slight quibble with the use of the phrase “test our hypotheses on the role of narratives” (line 258) simply because the paper did not put forth any hypothetical statement or even propositions (at least none that were explicitly stated). As such, I would resist the urge to use this type of terminology and employ different language. There is also the issue that the authors indicate that their next steps will be to collect more data and “use the resulting data in constructing ABMs to text [sic] our hypothesis...” (Lines 569-70). This suggests that the current paper, did not, in fact, text the hypothesis.

We agree and removed the term “hypothesis” from the text, using instead other language to describe how we challenge and improve our understanding thus far from case studies – e.g., page 12 lines 562-573.

E) Identities: The identities piece seems to be most consistently deployed throughout the manuscript to me, as it is tied into the conceptual section, and contained in the case study descriptions, as well. That said, it doesn’t seem to do much work in explaining what is happening in the narrative. For instance, the authors write about the creole garden as being “coupled with a strong cultural identity of the creole community” (lines 325-26) but it is unclear what that identity is, especially if we compare it to the identity invoked by the members of the Tamil community, which is “emphasized in reference to traditional stewardship” (lines 397-398). Are these examples of group identities? Is the group identity of a creole community self-referential?

Yes, these are examples of social identities tied to the community to which individuals feel connected. Therefore, in these and many other cases, they are self-referential in the sense that the group self-characterization emerges from the intra- and inter-group interactions and purpose. It may evolve as the group or purpose does.

There is more work that needs to be done here theoretically and conceptually to be clear about what the identity is and how it works. Relatedly, characters feature prominently in the Concepts and Approach section, but disappear in the cases. Do the different characters deployed impact the identity that is constructed? Just very unclear about the role of identity other than that it is something that is being constructed/invoked in these narratives.

We agree that more work needs to be done on identity through design of research studies that explore the community/group self-images and group portrayals. In the cases described in this manuscript, characters figure (e.g., Édouard Glissant in Martinique or Chief Makanjira in Malawi) as sources of narrative expressions and as such do indeed impact the identities both constructed or modified by the narrative in the community. In some cases, this may be a self-selecting process – i.e., effective on those for whom the narrative resonates and thus can draw the individual into the group.

F) networks: The incorporation of networks in this piece is also a bit confusing to me. The authors spend time talking about the importance of networks to meaning creation, the study does little to explain how the network is practically impactful on narrative efficacy or efficient dissemination. It is, of course, understandable how networks would be an enlightening perspective to take when thinking about transitioning to ABMs, but the concept doesn’t seem to contribute too much to the current study

We agree that the concept of network was not used meaningfully in this paper, so we have removed it.

G) Punchline: The “so what” question is in need of significant attention. After I read the manuscript, I was excited about the idea of using narrative-derived information to define the operations and parameters of ABMs, but I didn’t have a great “takeaway” from this study itself.

Hopefully the conclusion now contributes a clearer takeaway and “so what” message.

H) NPF and quantitative coding: the authors use some of the work in the Narrative Policy Framework to discuss the nature of narratives as “social constructions of realities and empirically measurable” (lines 147-48), but they don’t seem to take the second part of this to heart (at least in this study). The incorporation of narrative- derived data into ABMs, would be an approach that would share an affinity the NPF regarding its empirical-nature, (how else to code the agents in the model without quantifying the narrative data?), but in this study, the attention is focused on the context and the content of the narratives, which are clearly viewed by the NPF as NOT transferrable (in that they are categorized as narrative content and not part of the narrative form, which is structural and generalizable).

We gladly acknowledge the NPF, which provides an interesting and valuable perspective and approach. We also see from our work that there are “legs” to the qualitative data from narrative expressions and CANEs without necessarily referencing archetypal figures and quantitative assessments. The qualitative data provides insights and may prove useful for models of social dynamics. We think that the content of narratives or more precisely the core message in the narrative expression is indeed transferable, if by that one means that it can have strong impact in diverse communities. That is why the artistic expression as dance, music, and visual art, as well as spoken/written forms is so important. It carries the core idea across some boundaries of different contexts and cultures, though not necessarily all.

Round 2

Reviewer 1 Report

I find the author(s) response to my review to be strong. I now accept the paper for publication.

Reviewer 2 Report

The changes made by the authors to the original manuscript have addressed my concerns. What remains to be done is, in my opinion, minor copy-editing work which should not preclude its acceptance and publication.

Page 3, Line 118: Should be “separating their own from the foreign” The explication of CANEs is much clearer and explicit in this version. Page 7, Line 317: approx. should be “approximately”