Article 2 was submitted for publication in Journal of Knowledge Management Research &
Practice. The article is currently under review. This journal was chosen for two reasons:
First and foremost, it is the outlet for Andriessen’s much cited paper on knowledge as stuff or love. Further, the journal emphasizes the link between research and practice, which is also present in the research conducted in this project.
128
Chapter 14 Article 2
The Diversity of Metaphors for Knowledge – An Empirical Study
In July 2015 this article was submitted for publication in:
Knowledge Management Theory & Practice ISSN: 1477-8238
Manuscript number: KMRP-15-0133A
The Diversity of Metaphors for Knowledge:
An Empirical Study Linda Greve
Department of Business Communication and Center for Teaching and Learning, Aarhus University
Abstract
The study presented in this article shows that groups thinking aloud about the concept of knowledge will conceptualize knowledge metaphorically as both an object and as a personal resource—or as Stuff and Love, as pointed out by Andriessen (Andriessen, 2006, 2008). Know-ledge workers should, in other words, embrace the dichotomy between knowKnow-ledge as something measurable and thus manageable and knowledge as something personal and thus fragile to the manager. This provides the manager with a variety of ways to solve problems and imple-ment changes since diversity in metaphors will pave the way for diversity in solutions and new approaches. The results of the present study show that both managers and employees hold a diverse metaphorical concept for knowledge, implying that knowledge is not either a commodity or a personal experience, but both, depending on the context. The real task is to address the relevant domain at the right time.
Keywords: meaning of knowledge, knowledge management tools, knowledge use/utilization, knowledge transfer, knowledge creation
Introduction
Metaphors reveal how managers and employees think about and conceptualize abstract phenom-ena (Andriessen, 2006; Johnson, 2013; Lakoff &
Johnson, 1980, 1999). In organizational change and strategy work this has been known to be true for a long time (Abel & Sementelli, 2005; Arga-man, 2008; Cornelissen, 2005; Hogler, Gross, Hartman, & Cunliffe, 2008; Morgan, 2006; Vince
& Broussine, 1996).
In knowledge management (KM), however, we know little about how knowledge is actually conceptualized in groups of knowledge workers.
Metaphors in knowledge management literature have been studied in depth (Andriessen, 2006;
Andriessen & Boom, 2007), but these studies only report on how the concept of knowledge is represented metaphorically in literature and how these metaphors are perceived by individ-ual managers and employees (Andriessen, 2007;
Andriessen, 2008). Thus a more inductive study of metaphors for knowledge is needed to shed light on how knowledge is perceived by groups of knowledge workers. The group perspective is
im-portant since metaphors are used to negotiate the meaning of knowledge and to reason with about knowledge. Understanding metaphors in group conversations will clarify if metaphors are spe-cific to individuals or spespe-cific to groups and how they are being used in conversation and meaning making.
This article presents the results of six semi-structured group conversations. The transcripts are analysed by the same standards as in the stud-ies mentioned above, and the results are com-pared to the results in KM literature. The con-clusions point to a concept of knowledge being diverse and thus calling for more nuanced ap-proaches for the knowledge manager. Diversity in metaphors implies diversity in how problems and challenges in sharing, containing, selling, and creating knowledge should be approached. The article ends by presenting a process suitable for revealing metaphors for knowledge.
Metaphors in Management
Metaphor in management is a widely accepted but also broadly understood phenomenon. The
pure rhetorical sense of the word metaphor as a trope (Pinder & Bourgeois, 1982), such as irony, rhetorical questions, and comparison, has in lin-guistics and cognitive science been abandoned some 35 years ago (Lakoff & Johnson, 1980;
Steen, 2011). The fundamental finding made by Lakoff and Johnson (1980, 1999) was that a metaphor is not just a comparison in language;
it is also a link in thought. The metaphor influ-ences how we think and reason about the world.
One of the fundamental metaphors found to be primary by Lakoff and Johnson has to do with up: happy is up, more is up, and control is up (Lakoff & Johnson, 1999, pp. 46–54). This is used in an experimental study by Casasanto and Dijkstra showing that when telling a story of pos-itive valence participants are more likely to make upward movements than downward (Casasanto
& Dijkstra, 2010). This in turn indicates that not only do we understand happy, more, and control in terms of up, but we also connect the abstract phenomena happy, more, and control with our bodily experience of up. Metaphors are more than language. They activate the bodily expe-rience with the concrete part of the metaphor (up) in understanding the abstract phenomenon (happy, more, control). As a sign of this, just think about metaphorical phrases like climbing the career ladder or consider how most organi-zational diagrams show the managers above the employees.
The aim of this article is to provide better un-derstanding of the ways in which the abstract phenomenon knowledge is understood.
Using a metaphor is “. . . generation and creation of new meaning beyond a previously existing similarity” (Cornelissen, 2005, p. 751). This is a much more complex and invasive meaning of metaphor than metaphors just being a compari-son of incomparable elements to increase under-standing. The definition put forward by Cornelis-sen emphasizes that metaphors generate and cre-ate meaning. Thus, when understanding know-ledge in terms of something else, this can be both the result of a previous and long-lasting schema and a completely new metaphorical understand-ing (Johnson, 2013). The data collection method was designed to embrace this dualism in how
metaphors come about. The process was embrac-ing new metaphors and the use of multiple modes would reveal patterns between modes. Further, the comparison between companies would reveal potential dead or old metaphors.
Data and Method
The fundamental questions for this analysis were:
(1) Which metaphors for knowledge occur in the dataset? (2) Who speaks the metaphors, and do the participants stick to one metaphorical con-cept? (3) Do managers and employees differ in their choice of metaphorical domain?
The data for this study consists of six group con-versations. The groups represent six different companies. The companies are all less then five years old and define themselves as knowledge companies and creative start-ups. That covers consultancy and marketing in the areas of cul-ture, entrepreneurship, online presence, and the like. Each group contained three to five partici-pants, and the group members are characterized by being in a work relationship. The compa-nies are numbered in the order the conversations were conducted, and participants are identified by their company (C#) and a number (P#) given from the order in which they sat clockwise from the interviewer. Table 1 provides an overview of the participants in the conversations:
Each interview was conducted using a conver-sation guide. The questions in the guide were centred around the concept of knowledge and consisted of questions like ”How would you de-scribe knowledge to a new employee?” “Where does knowledge come from?” and “How do you know that you know something?” In order not to prime the participants, only metaphors put for-ward by the participants were used, but the same topics were covered in all conversations. Each conversation lasted approximately one hour.
The conversations were video recorded as well as transcribed. The coding took place by use of both video and transcript, in order to fully capture the contextual meaning of words. The analysis was conducted by use of the following method, inspired by Andriessen (2006, p. 96) 132
Table 1: Overview of participants in the conversation
Companies Participants Age Gender Academic Degree
6 25 Average:
29.8 years old Min. 23 Max. 60
Male: 13 Female: 12
Bachelor: 13 Master: 11 None: 1
and the metaphor identification method (MIP) (Pragglejaz Group, 2007):
All transcripts were read and videos watched to establish understanding of con-text.
In all transcripts nouns and verbs related to knowledge were marked.
For each related verb or noun, the basic and contextual meaning was established using a dictionary.
If the basic meaning deviated from the con-textual meaning, it was marked as metaphor-ically used.
Subclasses of metaphors were established and compared between companies.
Using the MIP-procedure would normally in-clude marking all potential metaphors. As only metaphors for knowledge is relevant to this study, only these were marked as was done in the study by Andriessen.
All conversations were conducted in Danish.
Thus the following examples will show the au-thor’s direct translation. As spoken language is not always eloquent, especially when the topic is complex and a group is negotiating meaning, the transcriptions are not perfect Danish and will not be perfect English either. This approach is chosen in order to provide understanding of what participants actually said and to give focus to the metaphors rather than the language.
In the following, an analysis of the metaphors is unfolded in order to provide an insight into how the concept of knowledge is conceptualized metaphorically in group-conversations.
The diverse concept of knowledge
The first coding of the conversations focused on knowledge as an object or knowledge as some-thing related to persons. This was inspired by the work of Andriessen as well as derived from the di-chotomy between knowledge transfer and know-ledge communication and thus the distinction between knowledge as something transferable and as something related to individuals. Even though the conversations had the same duration and format, the number of metaphors in relation to knowledge is very different. Figure 1 shows the number of metaphors in the categories “know-ledge is an object” and “know“know-ledge is something personal.”
From Figure 1 it becomes evident that some com-panies use more metaphors for knowledge than others, even though the conversations were of a similar length and conducted by use of the same interview guide. This difference in count will be explored below. Note that these are not unique metaphors but an overall count, so the same metaphor can be counted more than once if it is repeated.
Figure 2 illustrates the relation between the two metaphorical concepts in percentage.
Figure 2 shows that even though C6 uses much fewer metaphors than C3, the distribution of “ob-ject” and “personal” metaphors is similar.
It is evident that the groups use both KNOW-LEDGE IS AN OBJECT and KNOWKNOW-LEDGE IS SOMETHING PERSONAL in the conversa-tions. Thus how the groups conceptualize know-ledge is not a matter of either/or. But it is also evident that the KNOWLEDGE IS AN OBJECT
Figure 1: The total count of metaphors for knowledge for each company. C# refers to companies and the numbers on the y-axis refer to the number of metaphors in the two categories KNOWLEDGE IS AN OBJECT and KNOWLEDGE IS PERSONAL.
Figure 2: The relation between the two concepts of knowledge in percentage. The relation between the two concepts is similar between the six companies: object-metaphors are more frequent than person-metaphors except in C3, where the two types of metaphors are almost equally distributed.
134
metaphor is more dominant in all six companies.
Object-metaphors occur in 65.3 percent of the cases, and personal-metaphors occur in 34.7 per-cent of the cases. This in turn would suggest that reasoning about knowledge might not come from one of the two conceptual domains but might be a mix of the two depending on the context, since the metaphors are not participant- or company-dependent.
Looking deeper into the data, two questions stand out:
Do specific participants choose one of the two metaphorical concepts? This would im-ply that participants would reason on the grounds of that metaphor leading to misun-derstandings and different grounds of rea-soning between participants.
Is there a difference between the metaphors used by managers and employees, as sug-gested by Andriessen (2007)? This would support the hypothesis that different par-ticipants use different metaphors to reason with.
The answers to these questions and a more thor-ough analysis of the metaphors in the dataset are presented below.
Attending to the first question, the distribution of the two concepts on participants is presented in Figure 3.
Figure 3 shows a distribution of the two concepts similar to what was presented in Figure 2: both concepts are being used by most participants. To explain why C2-P3, C5-P4, C6-P3, and C6-P4 do not follow the pattern, it is necessary to take a look at the count of the metaphors in Figure 4:
C2-P3 uses very few metaphors and only object metaphors. C5-P4 uses no metaphors for know-ledge at all and also hardly says anything during the conversation. C6-P3 and C6-P4 only utter one metaphor for knowledge each. C3-P1 uses metaphors for knowledge much more frequently than all of the other participants.
Some participants use more metaphors than oth-ers. If divided into managers/owners and
employ-ees, the general picture of the usage of metaphor looks as presented in Figure 5.
Managers and company owners (red bars) seem to use more metaphors for knowledge than em-ployees (blue bars). This is to some extend ex-plained by managers and owners talking more in the interviews in general. To follow up on An-driessen’s findings that managers tend to think of knowledge as an object, this seems not to be true in these six cases. The answer to this difference in findings might be due to the two segments not being isolated in this study; by speaking to the other segment the concept became more nuanced.
More likely, though, the reason for the difference lies in the splitting of the concept of knowledge; it is both an object transferable and sharable in pre-sentations, spreadsheets, etc., and something per-sonal and tacit. Thus the knowledge conversion put forward by Nonaka and later by Nonaka et al.
between tacit and explicit almost predicts this pat-tern (Nonaka, 1994; Nonaka & Takeuchi, 1995;
Nonaka & von Krogh, 2009). Also, the know-ledge hierarchy as presented by Ackoff dividing knowledge in data, information, knowledge, and wisdom holds the assumption that knowledge a diverse concept (Rowley, 2007).
From this descriptive view of the dataset it is time to move into the details of the metaphors for knowledge. The two main categories of “ob-ject” and “something personal” are inspired by Andriessen’s “stuff” and “love” as well as by Non-aka’s tacit and explicit knowledge. However, the two categories of course cover a variety of metaphors. Some are present between companies and some are company-specific. The metaphors for knowledge that are used by more compa-nies also seem to be the conventional ones, such as “knowledge sharing” and knowledge as some-thing you gain from education, books, or the Internet.
Table 2 shows the distribution of metaphors in the two categories among all companies. Some of these are only used once or twice and only by one participant. Others are more commonly used both inside and between companies.
The variety of metaphors for knowledge is big-ger for KNOWLEDGE IS AN OBJECT than
Figure 3: Showing how the two metaphorical concepts are distributed on participants from the six companies.
The results mirror the result on company level—most participants are using both metaphorical concepts.
Figure 4: The count of metaphors distributed on participants.
136
Table 2: The metaphorical categories distributed on companies.
Knowledge ... C1 C2 C3 C4 C5 C6 Total
... is an asset of a person 11 4 10 4 14 3 46
... comes from books/edu 3 3 9 2 8 2 26
... can be moved 4 5 14 2 2 4 23
.... is derived from exp. 2 1 10 4 3 2 22
... can be found 2 2 3 1 1 1 10
... is in people in the co. 2 11 18 1 5 0 37
... can be shared 11 3 5 6 0 1 26
... can be contained 0 5 4 2 2 8 21
... has levels 1 5 3 1 3 0 13
... has a foundation 0 4 0 1 0 2 7
... can be added 0 0 7 13 5 8 33
... develops 5 0 2 1 0 3 11
... is a commodity 4 4 0 5 0 1 14
... can be acquired 0 0 5 2 2 1 10
... is between people 0 7 0 4 1 2 14
... is inside a person 1 4 27 1 0 0 33
... is fragmented 0 4 0 1 0 2 7
... is something organic 3 0 5 8 0 0 16
... can be accessed 0 0 0 9 2 0 10
... can be brought into play 0 0 2 1 0 0 3
... can be built 2 0 1 0 0 0 3
... can be translated 0 2 0 1 0 0 3
... emerges 0 0 0 2 1 0 3
... is a hindrance 0 0 0 0 1 2 3
... is a light 0 0 1 0 0 1 2
... is a liquid 0 1 2 0 0 0 3
... can be hungered for 0 0 0 0 3 0 3
... is a driving force 0 0 0 0 1 0 1
... is a gift 0 0 2 0 0 0 1
... is a person 0 0 0 1 0 0 1
... is a process 0 0 0 0 2 0 2
... is a ship 0 0 0 0 1 0 1
... is a tower 2 0 0 0 0 0 2
... is fragile 3 0 0 0 0 0 3
... is transformed 0 0 0 1 0 0 1
Total 56 65 130 74 57 43 414
Figure 5: Frequency of metaphors for knowledge distributed on participants. Dark bars indicate managers or company owners. Light bars indicate employees.
for KNOWLEDGE IS PERSONAL. Neverthe-less, KNOWLEDGE IS AN ASSET OF A PER-SON is the most frequently used metaphor in the dataset.
Some of the metaphors present in the category of object-metaphors cover an information approach to knowledge, as something you can share, add, and acquire from education, books, IT, etc. Non-aka would describe this conceptualization as in-formation rather than knowledge—it concerns the content rather than the context (Nonaka, 1994, pp. 15–16). However, these concepts are abundant in the dataset as a way of conceptualiz-ing knowledge.
Concluding on these counts of metaphors, it is evident that the groups in this dataset conceptu-alize knowledge as a diverse phenomenon. They do not have one single metaphor on either per-son level or group level. This in turn points to a concept of knowledge among knowledge workers as holding multiple interpretations and grounds for reasoning. Whereas it might be eas-ier for managers as well as employees to embrace a more overall concept of knowledge, having a diverse approach leads to the possibility of more
solutions and more interpretations and thus a more diverse approach to knowledge and work-ing with knowledge. In the understandwork-ing of metaphor put forward by Cornelissen and cited above, metaphors can help to create meaning beyond the previously existing. For such co-creation to take place in groups, it is less sur-prising that different metaphorical concepts can coexist. Rather than encouraging a more uni-form metaphorical concept of knowledge, this diverse approach to knowledge might support a more knowledge-creating approach to knowledge.
From the two concepts, something personal and an object, it is evident that the reasoning and creation of meaning is very different. When knowledge is something personal, the groups talk about knowledge as “sucking in competencies”
and “finding your core areas and diving deeper into them” but also that “knowledge can disap-pear from the company” when someone leaves the company and that if someone is away from the office “such a knowledge or personal relation can be hard to follow up on” (quotes are from company C1). In C3 the group talks about know-ledge as something you need to incarnate and get under your skin.
138
Thus when knowledge is something personal, you can internalize it, but this in turn leads to the knowledge leaving with the person, as it is personal.
In C2, as in all the other companies, knowledge is both something personal and an object. They talk about having a “knowledge bank” and about
“categorizing knowledge” and “putting it into boxes.” This concept of knowledge is different from the examples above and leads to different solutions. The important conclusion is that for the participants both concepts can be true at the same time. Knowledge can be both something personal and an object. It will infer different ideas and solutions and lead to a more diverse approach to knowledge, given that the group is aware that there is a difference in concept. The real prob-lem is not a diverse concept of knowledge but rather being ignorant about this diversity. If the group talks about two or more completely dif-ferent concepts without realizing it, this is of course an immense challenge to the group’s com-munication. Of the six groups in this study, two groups evidently succeed in co-creating a com-mon metaphor for knowledge holding diversity but providing a shared framework for understand-ing. It supersedes the purpose of this article to present the details of this, but an important aspect of the process is using a common mode besides language (Greve, submitted).
Externalizing concepts for knowledge
The two groups succeeding in creating a common concept for knowledge encapsulating knowledge as both something personal and an object were both characterized by making use of a common mode besides language. The value of this is well documented (El Refaie, 2013; Heracleous & Ja-cobs, 2008; Vince & Broussine, 1996), but even though all six groups were given the same oppor-tunities for using common modes, only two took those opportunities. For one group (C1) the mon mode was gesture. They invented a com-mon gesture for the metaphor KNOWLEDGE IS A TOWER. This led to reasoning like know-ledge being fragile, since it was very employer-dependent (personal), but also that customers could grab whatever knowledge they needed
(ob-ject). This common gesture-gestalt of knowledge gave way for exactly the meaning making and negotiation, which Cornelissen writes about.
When managers or groups wish to co-create a con-cept for knowledge, a common mode is essential.
Gesture is of course always available, but not all groups would turn to this as a common mode.
Thus pen and paper (Vince & Broussine, 1996), LEGO bricks (Greve, submitted; Heracleous &
Jacobs, 2008), or other materials should be made available and, moreover, can be used to negotiate the concept rather than used for one participant to show the rest of the group a concept.
A process for creating a shared and diverse con-cept for knowledge could look like this:
The group is given five minutes to draw/build/create their concept of some-thing abstract (in this case, participants built the concept of “dream office”).
Purpose: to get familiar with the method.
Then draw/build/create something closely related to the concept (in this case, partici-pants built experience). Purpose: to prime toward the concept of interest.
Then draw/build/create knowledge (be aware that if only one participant is talking, drawing, or building, all participants should be encouraged to interfere here). Purpose:
to create a shared gestalt of knowledge.
After the last five minutes, let the group explain what they built and what elements of knowledge could not be included in the building.
Let the group reflect upon how this ap-proach to knowledge could affect the work in the group and how this metaphor is al-ready present as well as how it could benefit the work in the future.
By use of this approach groups create a shared meaning and understanding of knowledge rather than each holding their own one understanding inflicted from outside.