Table 5.1: Distribution of Source Domains for Knowledge (Andriessen, 2006)
Nonaka and Takeuchi
Davenport and Prusak
Knowledge is something physical 29% 59%
Knowledge is a living organism 7% 25%
Knowledge is thoughts and feelings 31% 2%
Knowledge is a process 23% 2%
Knowledge is a structure 1% 1%
Unknown 5% 5%
2007), pointing out that managers seem to prefer the objectified knowledge as opposed to employees who prefer the personalized knowledge which is no surprise since managers prefer what is manageable and employees prefer what makes them unique. Steen (2011) points out that the error is the inappropriate conflation of linguistic and conceptual metaphors. It is not immediately possible to conclude on the grounds of a linguistic metaphor analysis that a metaphor in thought or a conceptual metaphor is present (see 3.2.7 above). As Steen points out (2011, p. 187), the linguistic metaphors might be traces of old, conventional metaphorical mappings and thus not perceived as metaphors by the reader. This is the trap of metaphors: it seems intriguing that metaphors in language cover metaphors in thought, but that is not necessarily so. Metaphors as a management tool is presented and discussed at length below in chapter 10 “Using Metaphors as a Management Tool”. To return to the dichotomy in metaphors for knowledge as either stuff or love, an object or personal, there is no empirical evidence in my data set that these two domains are mutually exclusive.
Chapter 14 contains an article which is a comment on Andriessen’s claim that knowledge is either or. The data material suggests that it might be both/and as would also be inferred by the theory of knowledge orders presented in chapter 4.3.
In regards to metaphor theory, I place my research in the bottom-up tradition (Krennmayr, 2013), analyzing without a clear hypothesis and not inferring from metaphors in language to metaphors in thought (Steen, 2008). As I analyze conversations, I make use of the conversation analysis approach focusing on multiple modes and metaphoricity as well as metaphors rather than a pure textual analysis (Cameron et al., 2009; Jensen & Cuffari, 2014).
Last but not least, I place my research in the third generation of knowledge communication, focusing more on the complexity of knowledge than on the levels or nature of knowledge.
This in turn is a natural consequence of the embodied and embedded approach to cognition and metaphoricity.
In other words, cognition is distributed, metaphors are multimodal, and knowledge is based on the observations and experience of the knower. These elements infer theoretical expectations with regards to metaphors for knowledge:
Knowledge is created, modified and negotiated by the group Metaphors for knowledge are present in more than one mode
The concept of knowledge is based upon experience and a high degree of complexity.
With this summary of the essential theoretical elements, I will now turn to the methodologi-cal assumption and the concrete research design of the data collection and data analysis.
Part III
METHODOLOGY AND RESEARCH
DESIGN
Chapter 6
Methodology
The ontological and epistemological assumptions founding the dissertation are described and discussed above in chapter 2.1. From this it is important to note that my scientific approach is pragmatism with a specific focus on embodied realism. The methodology concerns how the research question can be approached (Grix, 2002; Kuada, 2012). In chapter 7, the research design is explained, but this chapter aims at describing the underlying assumptions of the research design.
Both Kuada and Grix illustrate research design as a ladder, indicating that a step in one direction will lead to a certain path deeper and deeper into the specificities of the research.
Thus, in figure 6.1, the ladder of this research project is presented.
My study is a mixed method study based on interdisciplinarity. As my own theoretical and academic background is somewhat interdisciplinary and stands in a tradition always applying a mix of methods, this is a very natural choice. My masters degree is in Theology with supplementary elements from Journalism and Rhetoric. Theology as such is very interdisciplinary, combining studies of history, languages, philosophy, ethics and exegesis.
Looking at a specific phenomenon from different angels and with different methods and combining the angles into a more nuanced analysis and understanding are thus a fundamental part of how I was trained and shaped as an academic.
Below I will present the methodology of my research design.
Guided by the research question, the study of metaphors for knowledge in knowledge-intensive groups is a mixed method study. Fundamentally, it is an inductive, qualitative study of how and which metaphors for knowledge emerge in semi-facilitated conversations.
The setup for data collection, which is described in detail below, is a mix of methods. In the traditional sense of mixed methods being a mix of qualitative and quantitative methods, this is not a mixed method design. However, the study is a mix of methods with regards to design and analysis. The mix could be classified as method triangulation, both with regards to data collection and analysis. The data collection method is a mix between an experimental setup and semi-structured conversation. As the emergence of metaphors as
Ontology Epistemology Methodology Methods Sources
Pragmatism
Embodied realism
Mixed Methods Triangulation Action Research
Semi-structured conversations Metaphoricity in:
- Language - Gesture
- Joint epistemic action
Six creative start-ups + 1 group from food production
Figure 6.1: The ladder of the research project. The levels indicate the specificity from the overall ontology to the sources of the project. Adapted from Grix (2002, p. 180) and Kuada (2012, p. 58).
joint epistemic action is essential to answering the research question, the initial part of the conversations was inspired by a building task performed in a different research project.
Using this same method provides a taxonomy to analyze the building processes by. In order for the conversation to stay on the topic of knowledge, the researcher followed a conversation guide. This in turn ensures that the six groups in the dataset touch upon the same aspects of knowledge.
Triangulation is defined by Denzin as combining methods in the study of the same phe-nomenon (Johnson, Onwuegbuzie and Turner, 2007, p. 114). The phephe-nomenon in the context of this dissertation is obviously the nature and emergence of metaphors for know-ledge. However the study is made by combining methods from experimental disciplines, lingustic analysis, gesture analysis and discourse analysis. Method triangulation (Mathison, 1988, p. 14) instead of data triangulation or investigator triangulation is chosen in order to investigate from different angles how and which metaphors for knowledge emerge. The symbiosis of how and which enrich the analysis, as will be unfolded in the articles as well as in the data presentation in chapter 7.4.
With regards to method triangulation, figure 6.2 shows the different modes and contexts in which metaphors might emerge. This complexity is reflected in the research method-ology. Constituting the frame for the whole study is the conversation. About half of the conversation consists of the three building tasks and the explanation of the building of knowledge (see figure 7.1 for an overview of the flow of the conversations). Specific to the building part of the conversation is metaphoricity in the building. Crossing the lines between conversation and building is metaphors in language and gesture, and specific to the 54
conversation is the potential metonymic reference to the building after it has been removed.
The purpose of figure 6.2 is to illustrate how the levels of analysis interfere with each other. Metaphors in gesture are often also present in language, and metaphors in the building as well as metonymic reference to the building would also co-occur with language or gesture metaphors. Separating the different elements from each other would disregard the importance of understanding the emergence as well as the multimodality of metaphors for knowledge in group conversations. Thus, the methodology has to be multi-modal and make use of multiple methods.
In sum, the methodology reflects the ontological and epistemological choices. On this basis, the research design was made and the sources chosen. Moving down the ladder in figure 6.1 the following chapter presents the methods and sources.
THE CONVERSATION
BUILDING TASK/
JOINT EPISTEMIC ACTION METAPHORICITY
IN GESTURE
METONYMIC REFERENCE TO BUILDING
METAPHORICITY IN BUILDING
METAPHORICITY IN LANGUAGE
Figure 6.2: The levels of analysis in the conversations. The building approach is analyzed, and so are the metaphoricity in the building and potential metonymic reference to the building after it has been removed. Further, the metaphoricity in language and gesture is analyzed.
Chapter 7
Research Design
In this chapter, I present my cases, my data collection method, my methods of analysis and the findings in the six cases. A seventh case presented itself in the form of a group from a Danish food production company requesting a presentation of knowledge sharing strategies.
I used the developed method on the group, and through an action research process, they changed knowledge sharing strategy completely. This study is presented in article 3 and will not be presented further in this chapter. The analyses presented below are not as deep as the ones in the articles. Here they only serve to provide the reader with an overview of the data collected and to motivate why some cases were chosen for the articles. However, there are many interesting analyses to be made of the less salient cases as well. That is why I chose to present them here.
Based on all the above theoretical and methodological considerations, I designed a study to reveal how metaphors for knowledge are co-created in groups and which metaphors for knowledge occur in their conversations. Throughout the research period, a new approach to analyzing metaphors developed. As outlined above in chapter 3.5, I moved from a MIP-inspired approach to metaphor analysis, combining it with gesture analysis and an analysis of joint epistemic action towards a more holistic approach, looking at the conversations as ecological systems more in line with both 4E cognition and distributed cognition. This shift is evident in the articles presented below. Article 1 represents a merger-of-methods-approach, and especially article 3 represents a distributed conversation analysis approach.
In other words, my method developed and — to some extent — improved over time. Thus, below I will present all six cases as timelines of conversations. This is not presented in any of the articles and represents a dynamic conversations approach. As not all six companies are presented in the articles, I chose to present my data in this chapter in order to outline differences and similarities in the six cases with regards to co-creation of a concept of knowledge as well as metaphors for knowledge in the six conversations.
Table 7.1: The demographics of the participants
Gender Educational Level Age
Male: 13 Female: 12 Population: 25
Bachelor degree: 13 Master’s degree: 11 No finished degree: 1
Average: 29.8 y Min: 23 y Max: 60 y
Table 7.2: The demographics of the companies
C# Est. Number of
people in C#
Number of participants (owners)
Growth
(Y/N) Parameters of growth
C1 2009 25 5 (0) Y
Employees, Products, Services, Competencies
C2 2011 2 full time
3 half time 5 (1) Y Clients, Employees,
Offices in GE
C3 2013 4 3 (2) Y
Cash flow, Employees, Competencies, Prod-ucts
C4 2011 2 full time
2 part time 3 (2) Y
Assignments,
Ambition, Relations, Cash flow
C5 2010 7 4 (3) Y
Employees,
Assignments, Cash flow
C6 2013 7 5 (2) Y Cash flow,
Employees
7.1 The Dataset: Six Creative Startups
The dataset consists of six conversations conducted on the basis of the same template. The six groups were sampled to be as similar as possible. They all categorize themselves as creative startups and as knowledge companies. The demographics of the groups are presented in tables 7.1 and 7.2. All companies were promised anonymity and will only be discussed as C1-C6. Participants will be discussed as C#-P#. The participants were numbered as they were placed around the table clockwise from where I sat.
As becomes evident from the tables 7.1 and 7.2, the companies are similar with regards to being relatively young companies. Apart from one participant who never finished his education, all participants hold at least a bachelor’s degree. 11 out of the 25 participants 58
hold a master’s degree. Thus, the educational level is high and supports their self-image of being knowledge companies.
They all express that they experience growth, though the definition of growth is diverse.
The question was phrased as an open question, and the answers range from traditional growth signs like income and customers to growth in ambition and relations. The groups in the conversations consisted of 3-5 persons. In five out of six cases, one or more owners or partners were present.
The companies differ with regards to gender (nearly a 50-50 split on male and female participants), age (from 23 to 60 years old) and number of workers (the biggest has 25 workers, and two have four workers in all). However, the sampling has been made on the company profiles rather than the individuals.
The groups participating in the conversations are first and foremost working together on a daily basis on complex assignments demanding cooperation and a high degree of specialized knowledge. Conversations took place in their home environment in order to create a comfortable atmosphere and to take up as little of their time as possible.