strategic management

Metaphorical images of organization: How organizational researchers develop and select organizational metaphors Joep P. Cornelissen, Mario Kafouros and Andrew R. Lock

A B S T R AC T The article examines how metaphors are developed and selected

within organizational theorizing and research.The issue addressed is

not whether metaphors exist and play a part in organizational

theorizing – as this is now widely accepted – but to draw out how

metaphors are actually used and are of conceptual value, particularly

as such insights may aid organizational researchers in a better use of

them.Working from this position, the article reviews the extant theor-

etical literature on metaphor, surveys the organizational literature to

document past and contemporary metaphors-in-use (1993–2003),

and identifies the heuristics (i.e. judgmental rules) that have been used

by organizational researchers in developing and selecting these

metaphors. The identified heuristics are the integration, relational,

connection, availability, distance and concreteness heuristics. On the

basis of these identified heuristics, and the biases and errors associ-

ated with them, the article also posits a number of governing rules

that can guide organizational researchers in their continued develop-

ment and selection of metaphors in the organizational field.

K E Y WO R D S heuristics � metaphor � organization theory � tropes

This article reflects at least three trends that have become increasingly apparent over the past two decades within the field of organization theory: the development of interest in the paradigms, schemes and concepts that over

1 5 4 5

Human Relations

DOI: 10.1177/0018726705061317

Volume 58(12): 1545–1578

Copyright © 2005

The Tavistock Institute ®

SAGE Publications

London,Thousand Oaks CA,

New Delhi

www.sagepublications.com

time define and characterize the field (e.g. Barley & Kunda, 1992; Morgan, 1980); the development of interest in the process of theorizing, particularly in those cognitive processes underlying it (Weick, 1989); and the develop- ment of interest in the nature of language for (re)presenting organizational life, particularly the incidence and function of metaphor (Daft & Wiginton, 1979; Oswick et al., 2002). Although these trends overlap in various regards, the article tightens their association by explicating the role of metaphor in the thinking and behavior of organizational theorists and researchers. More specifically, the article clarifies how metaphor, and the imagination that follows from it, is used within organizational theorizing and what this implies for its continued use within organization theory and research.

Specifically, the purpose of our analysis is 1) to clarify how organiz- ational researchers circumscribe and understand the world of organizations through the use of metaphor, 2) to document the heuristics of metaphor that they use in doing so, and 3) to suggest, on the basis of this documentation, how metaphor can be used to its fullest effect. The latter suggestion is evidently more prescriptive in nature and moves beyond the descriptive data presented to discuss missed opportunities and potentialities in the heuristic inferences used by researchers in order to probe more deeply and generate new insights into the world of organizations. This is not, however, the central thrust of our analysis. Rather, our major emphasis is on providing a histori- cal and empirical overview of the past and contemporary metaphors in use over the period 1993–2003, and to delineate the heuristics (i.e. the judg- mental rules in producing and selecting metaphors) that have guided their development, selection and use. We focus particularly on the dominant metaphors within organizational theorizing and research and by looking at their heuristics attempt to explain their prevalence and continued use.

In what follows, we will first contextualize the role of metaphor in organizational theorizing, before moving on to a more specific and detailed discussion of prior work that has speculated on the heuristics of metaphor. From this discussion the article then proceeds with a survey of the use of metaphor in organizational theorizing and research over the past period (1993–2003) in order to infer and document empirically the heuristics-in-use in organizational theorizing and research. The results of the empirical survey and the uncovered heuristics-in-use are discussed and are also used to formu- late a number of governing rules for the selection, adoption and continued use of metaphor in organizational theorizing and research. We conclude with a discussion of theory and research implications, positioning the suggested governing rules for the use of metaphor within the wider realm of organiz- ation theory and suggesting research applications.

Human Relations 58(12)1 5 4 6

Metaphor in organization theory

The trends indicated at the beginning of this article suggest a marked increase of interest in recent years in the paradigms, schemes and metaphors that organizational theorists and researchers work from in their theorizing and research endeavors (Bacharach, 1989; Morgan, 1980; Weick, 1989). Although this interest comes in various forms (see Gioia & Pitre, 1990) and reflects wider meta-theoretical issues around theorizing and research, our concern in this regard is with the specific use of metaphor within the process of organizational theorizing. This concern is given in by previous work (Morgan, 1980, 1983; Weick, 1989) which has suggested that metaphors play a crucial role within theorizing, that theorists cannot really surpass them and that theorists and researchers therefore need to be more mindful of their use and the images that they evoke in such a way that they become ‘more deliberate in the formation of these images and more respectful of represen- tations and efforts to improve them’ (Weick, 1989: 529). This view stands in sharp contrast to an earlier view of metaphor as a derivative issue of only secondary importance. That is, metaphor was thought to be either a deviant form of expression or a nonessential literary figure of speech (e.g. Pinder & Bourgeois, 1982). In either case, it was generally not regarded as cognitively fundamental. This denial of any serious cognitive role for metaphor was prin- cipally the result of the longstanding popularity of strict ‘objectivist’ assump- tions about language and meaning. The objectivist view suggests that the world has its structure, and that our concepts and propositions, to be correct, must correspond to that structure. Metaphors, then, may exist as cognitive processes of our understanding, but their meaning must be reducible to some set of literal concepts and propositions (Bourgeois & Pinder, 1983; Pinder & Bourgeois, 1982).

In marked contrast with this ‘objectivist’ view, Morgan (1980, 1983) forcefully demonstrated that metaphors involve a cognitively fundamental way of structuring our understanding of organizations as a new meaning is created through the creative juxtaposition of concepts (e.g. ‘organization’ and ‘machine’) that previously were not interrelated. Ever since, a whole range of theories and frameworks have been proposed (e.g. the ‘transfor- mational’ model, Tsoukas, 1991, and the ‘domains-interaction’ model, Cornelissen, 2004, 2005) that have both advanced and challenged Morgan’s characterization of metaphor as proceeding ‘through assertions that subject A is, or is like B, the processes of comparison, substitution and interaction between the images of A and B acting as generators of new meaning’ (Morgan, 1980: 610). Tsoukas (1991, 1993), for example, suggests that a metaphor, as a figurative play of words, can be used in a creative manner to

Cornelissen et al. Metaphorical images of organization 1 5 4 7

reveal ‘literal’ structural similarities between concepts that were not salient before, and may as such provide for ‘enriching’ and ‘insightful’ new under- standings of organizations. Cornelissen (2004, 2005) argued that metaphor- ical language sets up a creative and novel correlation of two concepts which forces us to make semantic leaps to create an understanding of the infor- mation that comes off it. The notion of semantic leaps, then, points to certain ‘non-compositional’ processes that are at work in metaphor, that evoke the imaginative capacities of meaning construction, and that eventually lead to the production of a new, emergent meaning (see also Fauconnier & Turner, 1998; Tourangeau & Rips, 1991). Accordingly, in Cornelissen’s view, metaphors are cognitively fundamental in themselves – a metaphor creates new, emergent meaning that is not compositional; instead, there is new meaning constituted in and through the metaphor (e.g. ‘an organization having certain identity traits in its strategies, values and practices that give it its specificity, stability and coherence’ in case of the ‘organizational identity’ metaphor) that is not a composition of meanings that can be found in either the target or source concepts per se.

Beyond this discussion of how metaphors ‘work’, the organizational literature on metaphors has also drawn attention to further analytical distinctions; primarily between ‘live’ and ‘dead’ metaphors, and between ‘root’ metaphorical schemata versus specific ‘surface’ metaphorical language and concepts (Alvesson, 1993; Morgan, 1980; Oswick et al., 2002). Tsoukas (1991), for example, pointed to the difference between ‘novel’ or what are sometimes understood as ‘live’ metaphorical word combinations (e.g. ‘organizational identity’) versus ‘conventionalized’ or ‘dead’ metaphors (e.g. ‘organizational structure’); language and concepts that have become so familiar and so habituated in theoretical vocabulary that scholars have often ceased to be aware of the metaphorical underpin- nings (see also Hunt & Menon, 1995; Inns, 2002; Sandelands & Srivatsan, 1993). Alvesson (1993) and Morgan (1980) have drawn a distinction between ‘root’ or ‘second-order’ metaphorical schemata as schools of thought that filter and structure a researcher’s perceptions of the subject of study (e.g. ‘social phenomena as information processing systems’) (e.g. Daft & Weick, 1984) which then pre-structure and give rise to more specific ‘first- order’ metaphorical concepts (e.g. ‘organizational memory’) (e.g. Walsh & Ungson, 1991) with the latter serving as more concrete frameworks for scholarship and analysis. Inns (2002), finally, in her review of writings on metaphor within organization theory, suggested that many authors not only explore and use metaphors differently (for example as a qualitative research tool, as a generative tool for creative thinking, as a pedagogical or communicative tool) but also differ in terms of whether they critically

Human Relations 58(12)1 5 4 8

engage with them. At the level of organization theory, then, Inns’ review suggests that organizational researchers primarily appear to use metaphors in their theory building as ways of ‘making the unfamiliar familiar’ (akin to Inns’s view of metaphor as an explicatory teaching or communicative tool) or as a means of generating novel understandings that push the boundaries of the body of knowledge on organizations (cf. Inns’s view of metaphor as a generative tool for creative thinking) (see also Oswick et al., 2002; Schön, 1993, for a similar discussion). The latter generative capacity of metaphor to create new ways of seeing, conceptualizing and understanding organiz- ational phenomena is indeed widely acknowledged within the scholarly organizational community (Alvesson, 1993; Chia, 1996; Cornelissen, 2004, 2005; Grant & Oswick, 1996; Inns, 2002; Morgan, 1996; Tsoukas, 1991, 1993).

In the present study, we accommodate the aforementioned analytical distinctions (between ‘live’ and ‘dead’ metaphors, between ‘root’ metaphor- ical schemata and ‘surface’ metaphorical language, and between ‘explicatory’ and ‘generative’ uses of metaphor) and define metaphor as a linguistic utter- ance in which the combination of words is literally deviant in the sense that terms that have originally or conventionally been employed in relation to a different concept or domain are applied and connected to a target term or concept within organization theory (cf. Cameron, 1999; Gibbs, 1996; Steen, 1999). We also assume that metaphors as linguistic utterances reflect and intimate cognitively fundamental meanings about organizations and organiz- ational life; and that these meanings can be traced and inferred through a cognitive linguistic analysis. In other words, we consider metaphor as ‘a salient and pervasive cognitive process that links conceptualization and language’ (Fauconnier, 1997: 168). Within our study, we focus as mentioned on the extent to which metaphorical language features within organization theory and on the concepts and referent domains (e.g. economics, sports, machines, systems) that have cognitively been connected within organiz- ational metaphors. Our underlying aim here is to reveal whether the develop- ment of metaphorical language within organizational theory reflects certain heuristics by which metaphors in organization theory are developed, selected and accommodated within the practice of organizational theorizing and research.

Focusing on the heuristics of metaphor

The little, if any, research so far into the heuristics of metaphor is particu- larly problematic for, as Weick (1989) already suggested, an understanding

Cornelissen et al. Metaphorical images of organization 1 5 4 9

of the judgmental rules or heuristics in relation to metaphor would aid theor- ists and researchers in their selection and use of them. Armed with such heuristics organizational researchers would be able to select those metaphors that not only guide them towards plausible paths to follow (and away from implausible paths), but also generally break new ground (rather than scout old ground for neglected gems) and lead to conceptual advances that were inconceivable before.

Heuristics are compiled hindsight: they are nuggets of wisdom which, if only we’d had them sooner, would have led us to our present state much faster. This means that some of the blind alleys we pursued would have been avoided, and some of the powerful discoveries would have been made sooner.

(Lenat, 1982: 223)

The word ‘heuristic’ is often used in two senses: as a cognitive judg- mental or inferential process, and as a cognitive effect whereby it refers to a conceptual advance or improved decision-making (cf. Kahneman, 2002). In our usage here, the word ‘heuristic’ refers to the cognitive judgmental process that researchers engage in when they conjoin concepts in metaphor, judging them as fitting and as potentially revelatory of the organizational subject under investigation. The purpose of our survey of the organizational litera- ture (in the following section) is to elicit the heuristics that have so far been used by researchers in developing and selecting metaphors, and in doing so we attempt to answer, through our metaphorical lens, the fundamental question of how and on what grounds organizational researchers choose to represent and circumscribe the world of organizations.

In other words, uncovering these heuristics may give us an insight into why certain past and contemporary metaphors as ‘organizational identity’, ‘organization as theatre’ and ‘organization as machine’ have found their way into organizational theory, and have sparked off further inquiry, whereas other metaphors have not (e.g. ‘organization as chocolate bar’ or ‘organiz- ation as soap bubble’) (Cornelissen, 2002; Tsoukas, 1991) or have lost appeal after initial popularity (e.g. ‘organizational decision-making as garbage can’). Previous work on metaphor in the organizational literature has only paid scant attention to these questions.

In the stream of literature where metaphor is conceptualized as a comparison – that is, where metaphor is seen as a comparison in which the first term A (i.e. the target) is asserted to bear a partial resemblance (i.e. the ground) to the second term B (i.e. the source) (Alvesson, 1993; Oswick et al., 2002; Tsoukas, 1991) – it is speculated that the heuristic used by

Human Relations 58(12)1 5 5 0

researchers is to search for two concepts that bear an exact and literal simi- larity or sameness (that is implicit in the metaphor), and can then be compared. In this comparison model, metaphor interpretation is assumed to involve a comparison of concepts to determine, or rather extract, what discrete properties or relations applying to one concept can also apply to the other concept in the same or a similar sense, and accordingly the suggested heuristic within such an account is to judge the aptness of a metaphor on the basis of the similarity of the concepts conjoined within it. Alvesson (1993: 116) articulated this heuristic by saying that:

a good metaphor means the right mix of similarity and difference between the transferred word [i.e. the source concept] and the focal one [i.e. the target concept]. Too much or too little similarity means that the point may not be understood and no successful metaphor will have been created.

An alternative stream of literature (Cornelissen, 2004, 2005; Morgan, 1980, 1983) suggests that metaphor does not work by comparing or likening the target to the source as the comparison model assumes. Rather, metaphor is seen to involve the generation or creation of new meaning through an inter- active process of ‘seeing-as’ or ‘conceiving-as’, effectively moving beyond an antecedently existing similarity between the concepts conjoined within it (Table 1). Metaphor, in this so-called domains-interaction view (Cornelissen, 2005), involves the conjunction of whole semantic domains in which a corre- spondence between terms or concepts is constructed rather than extracted or deciphered, and the resulting image and meaning that comes off it is creative with the features of importance being emergent. The heuristics in this model for selecting metaphors and for judging them as apt follow from this position that the distinction between higher-order semantic domains and lower-level instance specific information of the target and source concepts is central to metaphor production and comprehension. Cornelissen (2004, 2005), Morgan (1980) and Tourangeau and Sternberg (1982) have proposed in this respect that metaphors are more apt and fitting and create strong and mean- ingful imagery when they relate concepts from more diverse or distant semantic domains (between-domains similarity) and when the correspon- dence between the target and the source concepts is conceived as more exact (within-domains similarity).

The suggested heuristics from both the comparison and domains-inter- action camps have so far been only speculative, as more broad-based empiri- cal research into metaphors-in-use in organization theory and their antecedent heuristics has been non-existent. The present article, as

Cornelissen et al. Metaphorical images of organization 1 5 5 1

Human Relations 58(12)1 5 5 2

Ta bl

e 1

T he

c om

pa ri

so n

an d

do m

ai ns

-in te

ra ct

io n

m od

el s

of m

et ap

ho r

Th eo

re tic

al a

ss um

pt io

ns Pr

in cip

le s

of m

et ap

ho r

Su gg

es te

d he

ur ist

ic

C om

pa ri

so n

m od

el

D om

ai ns

-in te

ra ct

io n

m od

el

M et

ap ho

r w

or ks

b y

lik en

in g

th e

so ur

ce to

t he

t ar

ge t

co nc

ep t.

T he

c om

pa ri

so n

pa ra

di gm

s ug

ge st

s th

at a

n an

al og

y or

si m

ile n

ot o

nl y

ex is

ts a

s a

ne ce

ss ar

y co

nd iti

on fo

r m

et ap

ho r

bu t

al so

p ro

vi de

s th

e gr

ou nd

fo r

its c

om pr

eh en

si on

( af

te r

a m

et ap

ho r

ha s

be en

r ec

og ni

ze d

as fa

ls e

an d

pa ra

ph ra

se d

in to

a c

om pa

ri so

n st

at em

en t)

.

M et

ap ho

r cr

ea te

s an

e m

er ge

nt m

ea ni

ng st

ru ct

ur e

th at

c an

no t

be r

ed uc

ed o

r ex

pl ai

ne d

by r

ef er

ri ng

t o

its c

on st

itu en

t pa

rt s

(i. e.

th e

ta rg

et a

nd s

ou rc

e co

nc ep

ts ).

A d

is tin

ct io

n be

tw ee

n hi

gh er

-o rd

er c

on ce

pt ua

l d om

ai ns

a nd

lo w

er -le

ve l,

in st

an ce

-s pe

ci fic

in fo

rm at

io n

(a t

th e

le ve

l o f t

he t

ar ge

t an

d so

ur ce

co nc

ep ts

) is

im po

rt an

t to

m et

ap ho

r, as

do m

ai n-

le ve

l k no

w le

dg e

is a

ut om

at ic

al ly

en ga

ge d

on e

nc ou

nt er

in g

a m

et ap

ho r

an d

gu id

es fu

rt he

r pr

oc es

si ng

.

Ex tr

ac tio

n: M

et ap

ho r

co m

pr eh

en si

on in

vo lv

es a

c om

pa ri

so n

of (

ta rg

et a

nd so

ur ce

) te

rm s

or c

on ce

pt s

to de

te rm

in e

(e xt

ra ct

) w

ha t

di sc

re te

pr op

er tie

s or

r el

at io

ns a

pp ly

in g

to o

ne te

rm c

an a

ls o

ap pl

y to

t he

o th

er t

er m

in t

he s

am e

or a

s im

ila r

se ns

e.

C on

st ru

ct io

n: M

et ap

ho r

co m

pr eh

en si

on in

vo lv

es c

on si

de ri

ng (

co ns

tr uc

tin g)

in w

ha t

st ru

ct ur

al s

en se

t w

o co

nc ep

ts a

re al

ik e

(a t

th e

do m

ai n

le ve

l), w

hi ch

pr ov

id es

a fr

am e

fo r

th e

fu rt

he r

bl en

di ng

o f i

m pl

ic at

io ns

a nd

t ha

t le

ad s

to a

n em

er ge

nt m

ea ni

ng .

A pt

a nd

in si

gh tf

ul m

et ap

ho rs

a re

t ho

se m

et ap

ho rs

o f w

hi ch

t he

c on

ce pt

s ar

e ju

dg ed

a s

di ss

im ila

r in

s om

e re

sp ec

ts ,w

hi le

s im

ila r

in th

eo re

tic al

ly im

po rt

an t

re sp

ec ts

.

A pt

a nd

in si

gh tf

ul m

et ap

ho rs

a re

t ho

se m

et ap

ho rs

t ha

t ar

e ju

dg ed

t o

re la

te c

on ce

pt s

1) b

et w

ee n

w hi

ch a

c or

re sp

on de

nc e

(in a

n st

ru ct

ur al

s en

se )

ca n

be c

on st

ru ct

ed ,a

nd 2

) th

at ar

e dr

aw n

fr om

d om

ai ns

t ha

t ar

e in

t he

fi rs

t in

st an

ce s

ee n

as d

is ta

nt .

mentioned, addresses this void through a survey of past and contemporary metaphors in organization theory.

Method

Data collection

In order to answer the main research questions, we set out to collect data that, first, inform us of the range of metaphors-in-use and, second, allow us to provide a motivated explanation of the heuristics used in their develop- ment and selection. We selected the Social Sciences Citation Index (SSCI) database and the subsection of ‘management’ journals within it as the source for our data. The subsection of ‘management’ journals includes the major theoretical and empirical journals within management and organization studies in general, and as such incorporates work in organization theory.1

We specifically selected these journals on the ‘management’ list within the SSCI database as the source for our data for a number of reasons. First, we wished to have a database of published articles within management and organization studies that is in a sufficiently formal form through coded field tags (e.g. keywords, title, source) and allowed us to perform searches and retrieve data. Second, we wanted to have a database that spans a consider- able period of time (the SSCI database covers ‘management’ journals since 1956), so that we could retrieve data on the spread and use of different metaphors over time. Third, we required a database containing publications of journals that would be largely representative of the body of theoretical and empirical work within organization theory. The ‘management’ list within the SSCI database provides for such a representative sample, although, of course, it is by no means exhaustive of all theoretical and empirical work that is carried out within management and organization studies.

The data that we collected were sourced from the academic publi- cations within the journals listed under ‘management’ in the SSCI database over the period 1993–2003. The period 1993–2003 was chosen for two reasons. First, this period provided us with a sufficient time series for data collection and analysis. Second, there are substantial gaps within the SSCI records for many ‘management’ journals prior to 1993 and only from 1993 onwards are the records for the Topic (title, abstract and keywords) complete.

Data were obtained through a number of steps. As a first step, we screened the list of journals under ‘management’ within the SSCI database, and selected those journals with 1) a weighted impact factor of more than 0.5 for the period 2001–3, and 2) a weighted number of citations above 500

Cornelissen et al. Metaphorical images of organization 1 5 5 3

for the period 2001–3. The motive for doing so was that by using these criteria we would arrive at a set of journals (Table 2) that are reflective of the general body of knowledge within organization theory and its advance- ment over time. As a second step, we performed searches for the publications of each of these journals, entering the adjective ‘organizational’ and the noun ‘organization’ as target terms into the Topic (title, abstract or keywords) field tag. Here, we set out to identify metaphorical word combinations involving a source term being compared to the target terms ‘organizational’ and ‘organization’. The two mentioned terms were chosen as target terms as they constitute part of word combinations, including metaphorical ones, within the organizational field. Each of the searches then produced the co- occurrence of words related to these two terms. We printed the full abstracts

Human Relations 58(12)1 5 5 4

Table 2 Journal publications selected for the database

Journal title Three-year Three-year average average impact cites factor (2001–3) (2001–3)

1 Academy of Management Review 3.757 4460 2 Administrative Science Quarterly 3.110 4630 3 Academy of Management Journal 2.906 5115 4 Strategic Management Journal 2.699 4683 5 MIS Quarterly 2.493 1662 6 Sloan Management Review 2.367 1197 7 Organization Science 2.012 2166 8 Harvard Business Review 1.955 3826 9 Human Resource Management 1.631 776

10 Journal of Management 1.567 1713 11 Research Policy 1.440 1597 12 Management Science 1.440 6576 13 Organizational Behavior and Human Decision Processes 1.417 2993 14 Information & Management 1.414 743 15 Organization Studies 1.253 844 16 Journal of International Business Studies 1.240 1563 17 California Management Review 1.192 1057 18 Journal of Product Innovation Management 1.182 788 19 Human Relations 0.926 1887 20 Journal of Management Studies 0.865 1025 21 Decision Sciences 0.731 1105 22 International Journal of Operations and Production Management 0.550 739 23 Interfaces 0.539 681

for all of these word combinations, and excluded the book reviews from our database and further analysis. As a third step, the words that we retrieved were entered into a database as source terms together with the mentioned adjective and noun (‘organizational’ and ‘organization’) as target terms and we recorded the occurrence of each combination of words over the time period (1993–2003) surveyed.

Data analysis

The data that we collected included 969 different word combinations involv- ing the target term ‘organizational’ and 262 word combinations including the term ‘organization(s)’ that were mentioned at least twice over the period (1993–2003) surveyed. For our data analysis, we used a definition of concep- tual metaphor as a linguistic utterance in which the combination of words is literally deviant in the sense that terms that have originally or conven- tionally been employed in relation to a different concept or domain are applied and connected to a target term or concept within organization theory (cf. Cameron, 1999; Gibbs, 1996; Steen, 1999). This definition is intention- ally broad so that it includes both ‘novel’ or what are sometimes understood as ‘live’ metaphorical word combinations (e.g. ‘organizational identity’), as well as ‘conventionalized’ or ‘dead’ metaphors (e.g. ‘organizational struc- ture’); word combinations that have become so familiar and so habituated in theoretical vocabulary that scholars have often ceased to be aware of their metaphorical precepts. The definition is also sufficiently formal in specifying metaphor as a conceptual combination involving the composition of features of the target and source concepts or terms compared, with the source concept coming from a domain that is distant to the subject of organizations and organizational behavior within organization theory. Hence, it enabled us to identify and map metaphorical word combinations and to distinguish them from other word combinations that involve any of the specified target terms but are not metaphorical such as ‘organization analyst’ (i.e. fails the compo- sition criterion) or ‘organization work’ (i.e. fails the distance criterion).

Using this definition, two of the authors acted as coders and indepen- dently screened the entire database to identify and code the word combina- tions that qualified as conceptual metaphors in this sense and to exclude non-metaphorical word combinations. A total of 861 metaphorical word combinations remained, including 786 ‘organizational’ metaphors and 75 ‘organization’ metaphors. These metaphorical word combinations were then further analyzed by each of the two authors independently. Both authors read the abstracts involved and coded the identified source term for the conceptual metaphor, as well as the larger source domain (e.g. economics, biological

Cornelissen et al. Metaphorical images of organization 1 5 5 5

evolution) that the source term was sourced from, and then ordered and named the coded material into categories. We grouped these source terms according to the domains (i.e. the organization of knowledge, such as our knowledge of ‘clockworks’) that these are sourced from, and then further grouped these into meaningful larger root metaphorical categories (i.e. the underlying root metaphor, such as ‘machine’). Where more than one root category might apply, we used the central features of the conceptual metaphor to select among alternatives. If two categories seemed equally applicable, we assigned the construct in question to the most specific of them. This issue arose, for example, in respect of the ‘organizational monitoring strategies’ compound, which incorporates features from both the ‘image’ (monitoring) and ‘warfare’ (strategies) root metaphors. Once all of the abstracts were read, interpreted and roughly ordered, each of us took time to integrate, refine and arrange categories so that these began to come together as a more conceptual whole. Here, each of us compared and contrasted interpretations of conceptual metaphors into coherent and signifi- cant categories. The individually identified categories were then laid next to one another, compared, discussed and integrated. These categories of metaphors-in-use that we identified by our own reading were informed and further refined by existing lists of root and conceptual metaphors (Morgan, 1980, 1986, 1996; Putnam et al., 1996) in order to increase the reliability and validity of our final categorization. The final product involves a categorization of metaphorical theoretical constructs central to the field of organization theory, classified according to the root metaphorical schemes upon which they are each formulated and understood. We then elaborated upon the different ‘root metaphorical schemes’ and the set of ‘organizational’ and ‘organization’ conceptual metaphors that according to the first stage of our analysis are prominent within organization theory; that is, metaphors that are frequently mentioned and used, and on that basis occupy a central place within organization theory. Within this elaboration, we aimed to retrace and reconstruct the heuristics that organizational researchers have used in developing and selecting certain metaphors in their theorizing and research.

The specific analyses and methods that we employed to answer our research questions are a direct outcome of our cognitive linguistic perspec- tive on conceptual metaphor as a cognitive process that links cognitive conceptualization with the specific use of language. In line with much other work in cognitive linguistics (see Coulson, 2001; Gibbs, 1996; Lakoff, 1993), our methodological approach is therefore aimed at inferring conceptual knowledge based on the analysis of systematic patterns of linguistic struc- ture. Such analyses of systematic patterns in language suggest a variety of

Human Relations 58(12)1 5 5 6

conceptual structures including metaphorical mappings and noun–noun compounds (see Gibbs, 1998). This emphasis on the content of what people know and the linguistic behavior that corresponds with it is quite different from the major focus in cognitive science on the general architectural form of human thought and language (Murphy, 1996, 1997). One result of this difference in emphasis is that a cognitive linguistic analysis is by definition post hoc; the focus lies on conceptual metaphors as existing linguistic utter- ances that reflect certain conceptualizations and patterns of thinking. Cogni- tive psychologists instead focus on individuals’ conceptual knowledge and aim to predict how that influences the existence of different linguistic behavior, not that an individual’s linguistic behavior can be explained post hoc by inferring conceptual knowledge, including metaphorical mappings.

Past and contemporary metaphors in use

The data that we collected included 969 different word combinations involv- ing the target term ‘organizational’ and 262 different word combinations with the term ‘organization(s)’; 786 of the 969 ‘organizational’ word combi- nations qualified as metaphorical. Only 75 of the 262 ‘organization’ word combinations were identified as metaphorical. One explanation for this difference between the number of metaphorical ‘organizational’ and ‘organization’ word combinations is that a metaphor is more directly cued or evoked with ‘organizational’ combinations; in these combinations an organization is seen to have certain features or characteristics which presup- poses a metaphorical lens of what an organization is conceived to be. Nouns like ‘organization’, on the other hand, function primarily referentially (Cameron, 1999; Hopper & Thompson, 1984). To illustrate this point, 96 different word combinations involving an adjective with the nouns ‘organiz- ation’ and ‘organizations’ were identified, with only 18 of these combina- tions qualifying as metaphorical. The large majority of these adjective–noun combinations are of a simple predicate-subject form with the adjective predi- cating the nouns that they modify. For example, ‘Japanese organization’ simply predicates the location of the organization involved. Such predicat- ing adjectives specify one of the predicated object’s attributes. The large majority of these predicating adjectives appear to have a referential purpose in that they specify the location (e.g. ‘Japanese organization’), size (e.g. ‘large organization’, ‘small organization’) or the nature (e.g. ‘economic organiz- ation’, ‘industrial organization’, ‘multinational organization’) of the organiz- ation involved. In other words, word combinations of this kind have a referential rather than metaphorical or indeed generative function.

Cornelissen et al. Metaphorical images of organization 1 5 5 7

Tables 3 and 4 outline the different categories of ‘root metaphors’ that we identified for both the ‘organizational’ and ‘organization’ metaphorical word combinations. Here, as mentioned, we categorized each conceptual metaphor (a metaphorical word combination involving either ‘organiz- ational’ or ‘organization’) according to the larger root metaphorical scheme upon which they are each formulated and understood. Twenty-five signifi- cant categories for the ‘organizational’ conceptual metaphors and 10

Human Relations 58(12)1 5 5 8

Table 3 Root metaphorical categories of ‘organizational’ conceptual metaphors

Root Conceptual metaphors (examples) Frequency metaphorical (counts) category 1993–2003

Animate being Creativity, character, behavior, identity, reputation, learning 1457 Machine Control, restructuring, resources, performance, design, 851

capacity Flow-change Change, renewal, development, transformation 419 Evolution Growth, functions, fit, decline, demography, life 249 Architecture Level, form, model, architecture 242 System Environment, contingencies, barriers, inertia, responses 207 Culture Culture, subculture, cultural climate 203 Geographical space Domain, world, setting, landscape 195 Economics Innovation, returns, efficiency, assets, dividends 97 Image Vision, image, perspective, focus, myopia, outlook 85 Institutionalized Codes, rules, guidelines, styles, accountability 78 norms Language Communication, names, discourse, rhetoric, message 78 Interpersonal Alliance, partners, cooperation 73 relationships Society Citizens, community, prosperity, citizenship, social cohesion 71 Warfare Strategy, plans, ranks, mission, competitive tactics 69 Politics Governance, imperialism, power, political regimes 63 Law Liability, (in)justice, (il)legality 59 Mathematics Factors, covariate, correlate 51 Space Niches, dimension, span, container, buffer 51 Environment/ecology Sustainability, greening 46 Symbolism Brand, symbols, myths, dramas 23 Linkage Network, link 18 Time Future, trend, momentum, schedule, time 15 Chaos Complexity, emergence, bifurcation 11 Sports Arenas, athletes, team, tournament 8

Note: Root metaphorical categories with five or fewer counts are treated as inconsequential and are not displayed in the table.

categories for ‘organization’ conceptual metaphors were identified. As Tables 3 and 4 reveal, the dominant root metaphorical categories for ‘organiz- ational’ and ‘organization’ conceptual metaphors are ‘animate being’, ‘machine’, ‘flow-change’, ‘evolution’, ‘architecture’, ‘culture’, ‘system’, ‘warfare’ and ‘geographical space’. The root category of machine metaphors likens organizations to mechanical systems and suggests an integrated picture of corporations as comprised of a series of mechanically structured inter- connected parts and resources. This category of metaphors lends its promi- nence to its roots in early organization theory, including the contributions of Max Weber and Frederick Taylor, the themes and subjects (e.g. ‘organiz- ational structure’, ‘capacity’, ‘control’, ‘design’) that emerged on the back of it, and the concrete mechanical concepts with which the subject of an ‘organization’ is compared. Animate being metaphors liken aspects of ‘organizations’ to living organisms, specifically humans. These are sometimes direct descriptions of organizations as acting beings, as when organizations are seen as carrying out certain behaviors or as trying to impress groups within their environments, but in other cases specific human properties such as ‘learning’, ‘creativity’, ‘character’, and ‘involvement’ are employed to conceptualize and explain organizations and the behavior and events involved.

Besides the observation that the ‘animate being’ and ‘machine’ categories assume the dominant position in both Tables 3 and 4, the

Cornelissen et al. Metaphorical images of organization 1 5 5 9

Table 4 Root metaphorical categories of ‘organization’ conceptual metaphors

Root Conceptual metaphors (examples) Frequency metaphorical (counts) category 1993–2003

Machine Design, structure, control, size 117 Animate being Learning, self, success, ability, knowledge, behavior 117 Culture Culture 32 Warfare Strategy 22 System Environment, fit 19 Linkage Network, networked 17 Family Parent 7 Symbolism Storytelling 7 Space Context 7 Architecture Level 7

Note: Root metaphorical categories with five or fewer counts are treated as inconsequential and are not displayed in the table.

categorization in Tables 3 and 4 also points to additional dominant root metaphorical categories. One of these additional root metaphorical categories, the category of flow-change, likens organizations to processes and forces that are constantly in flux and transformation. Conceptual metaphors in this category such as ‘organizational change’ are typically seen as an ‘ongoing process, a stream of interactions, and a flow of situated initiatives, as opposed to a set of episodic events’ (Tsoukas & Chia, 2002: 569) or other fixed reference points of ‘stability’ and ‘routines’ within organizations. The root category of evolution metaphors is modeled upon Darwinian thought and likens organizations and their activities to evolutionary patterns and processes. There is an extensive literature discussing the different ‘species’ of organizations and their development around concepts as ‘demography’, ‘fit’ and ‘decline’ and also evolutionary principles such as ‘variety–selection– retention’ have made their way into much organizational theorizing. The root category of architecture metaphors likens organizations to general physical or morphological structures, often in terms of ‘models’, ‘levels’, or ‘forms’. A common feat of these metaphors is that they are spatial and structural in form, and their use leads to a view of organizations as erected structures, frame- works or edifices. Within the category of culture metaphors, organizations are seen as cultural phenomena and as manifestations of cultural morals, values, rituals and symbols. The root category of systems metaphors likens organiz- ations to simple systems interacting and adapting to an ‘environment’. An organization is seen to consist of various interdependent subsystems and its behavior is said to be emitted and then selected by environmental ‘contin- gencies’ in much the same way that variations among individuals have been selected by the environment during the course of evolution. This notion of simple systems and environmental selection is thus itself metaphorical although may not always strike us as highly metaphorical as it is based on the now familiar Darwinian metaphor of natural selection. The systems metaphor entered organizational thought at around 1955 (Barley & Kunda, 1992), and has since led to such dominant and well-known conceptual metaphors as ‘contingency’, ‘open systems’, ‘environment’, and ‘loose coupling’. In recent years, systems thinking has been connected to the root categories of animate beings and evolution more generally and also adapta- tions and extensions of systems thinking including ‘chaos’ and ‘complex adaptive systems’ have entered the theoretical vocabulary of organizational researchers (see, for example, Thiétart & Forgues, 1995). Warfare metaphors liken organizations and related behaviors to war-like activities and events including ‘strategies’, ‘competitive tactics’, and ‘competitive wars’. Geographical space metaphors, finally, have as their source domain the distri- bution of objects or places in space (Lakoff & Johnson, 1980), particularly

Human Relations 58(12)1 5 6 0

geographical locations and places. This root metaphorical category includes such metaphors as ‘domain’, ‘world’, ‘setting’ and ‘landscape’, whereby organizations are represented in terms of geographical spaces and locations. The image of organization as an ‘organizational domain’, for example, repre- sents the scope and nature of organizational activities as confined to an enclosed space (see, for example, Stapel & Koomen, 1998).

Together, these dominant root metaphorical categories subsume much of the theorizing and research in relation to organizations within the field of organization theory. As such, there is value in trying to explain the domi- nance of these root categories in terms of the heuristics underlying their development, selection and continued used by organizational researchers. Previous explanations of the development of organization theory have tended to approach the topic from primarily a sociological perspective; emphasiz- ing the uptake of a certain theoretical concept or larger school of thought as the result of sociological and political factors (e.g. whether a particular concept resonates with the preoccupations and interests of the stakeholders in organizational research at a particular point in time; pressures to focus on certain theories and concepts) (see, for example, Barley & Kunda, 1992; McKinley et al., 1999; Tsoukas & Knudsen, 2003). Here, we suggest an alternative and potentially complementary explanation for the root metaphorical categories that we identified; one that is rooted in our cogni- tive linguistic method of metaphorical analysis and in a view of theorizing as ‘disciplined imagination’ (Weick, 1989, 1995). Specifically, we argue that organizational researchers theorize by designing, conducting and interpret- ing imaginary experiments where they rely upon metaphors to provide them with vocabularies and images to theoretically represent and express organiz- ational phenomena (cf. Weick, 1989). In this process, organizational researchers use certain heuristics in selecting and retaining a metaphorical combination of concepts, either for reasons of making the unfamiliar familiar or of generating new insights that were inconceivable before. The range of metaphors-in-use points to particular dominant ways of thinking by researchers about the world of organizations, and to certain heuristics that they use in doing so. For example, a particularly striking observation in relation to Tables 3 and 4 is that very little attention has been given to time metaphors, a point also raised by Hassard (2002), while these metaphors have found their way into virtually all other social scientific disciplines including economics (McCloskey, 1995) and psychology (Leary, 1990). In other words, the question that this evokes is why certain metaphors have been imagined and chosen whereas others have not, or at least to a lesser degree? And what have been the thought processes, or rather heuristics, that have guided this imagination and choice?

Cornelissen et al. Metaphorical images of organization 1 5 6 1

Heuristics-in-use

Informed by the overview of metaphors-in-use presented in Tables 3 and 4, we subsequently reconstructed the heuristics used by organizational researchers in developing and selecting these metaphors. In this reconstruc- tion exercise, we have dwelled upon cognitive scientific and cognitive linguis- tic research into the principles of metaphor production and comprehension (e.g. Fauconnier & Turner, 1998; Gentner & Markman, 1997; Gibbs, 1996; Katz, 1992; Tourangeau & Sternberg, 1982). One of the main ideas that has emerged out of this body of research is a constructive conception of metaphor comprehension and judgment. According to this view, the meaning of a metaphor is constructed through the creative juxtaposition of words or concepts – and not merely revealed (as the comparison model suggests) – and, as such, judgment of its aptness is also based on this construction. This conception is entailed by findings that judgments of the aptness, ‘fit’ and revelatory value of a metaphor are based on a relational conjoining or blending of the target and source concepts into a newly constructed metaphor- ical image (e.g. Fauconnier & Turner, 1998; Gentner & Markman, 1997; Gibbs, 1996; Tourangeau & Rips, 1991). Furthermore, such judgments of the value and use of a metaphor rest in the context of organizational theorizing and research, as in the world at large, on a limited number of simplifying heuristics rather than extensive algorithmic processing (Gilovich & Griffin, 2002). These judgmental heuristics can be seen as automatic, often intuitive, and sensible estimation procedures of the aptness and potential of the metaphorical image construed and feature as a response to uncertainty; when the full measure of a metaphor’s value for organizational theorizing and research is not yet known. Another important point is that such judgments relate to estimations of the ‘aptness’, ‘goodness of fit’ or (potential) ‘revela- tory value’ of a metaphor, and not to criteria specifying validity or truth conditions. As our remarks in earlier sections of this article already indicated, a metaphor is judged through and in the construed image that it evokes, not on the basis of a certain correspondence to reality as is the case with more formal models that may however be derived from them (see Beyer, 1992; Tsoukas, 1991). In a recent article, Von Ghyczy (2003) remarks to this effect:

Like the model, the metaphor bridges two domains of reality. For it to be effective, those domains must clearly share some key and compelling traits. But this correspondence differs from the direct mapping of a model. Rather than laying claim to verifiable validity, as the model must do, the metaphor must renounce such certainty, lest it become a

Human Relations 58(12)1 5 6 2

failed model. Metaphors can be good or bad, brilliantly or poorly conceived, imaginative or dreary – but they cannot be ‘true’.

(Von Ghyczy, 2003: 90)

This means that the judgment of a metaphor’s aptness, and thus also the heuristics underlying it, hinges upon the components of the metaphor in question, and the metaphorical image thus construed, rather than external appraisal through correspondence to a material or lived reality as envisaged by organizational researchers. In other words, the heuristics used center around, and are indeed confined to, perceptions of the relation or match of the target and source concepts of a metaphor and the larger semantic domains that they are drawn from. In all, we identified the following six heuristics on the basis of our survey. Together, they provide a motivated explanation for how predominantly metaphors are developed and selected within organization theory.

1) The integration heuristic; that representations in the metaphorical image can be manipulated as a single unit;

2) The relational heuristic; that relations in the metaphorical image should match the relations of their counterparts in other semantic domains;

3) The connection heuristic; that the representation in the metaphorical image should maintain a relationship to the input target and source concepts;

4) The availability heuristic; that, given a metaphorical image, the inter- preter should be able to infer the structure in relation to other subjects and applications;

5) The distance heuristic; that the target and source concepts need to come from semantically distant semantic domains;

6) The concreteness heuristic; that the source concept compared to the target is sufficiently concrete (rather than abstract) to be understood and manipulated.

Despite their poetic names, most of these heuristics are derived from standard pressures obtained in all mapping situations, including metaphori- cal mappings (see Hofstadter, 1995, for a review). The integration heuristic, first of all, refers to the pressure to bring partial structure from different concepts and domains together in such a way that it produces a fully inte- grated image with an easily manipulable representation (Fauconnier & Turner, 1998, 2002). In research on metaphorical mappings, the integration

Cornelissen et al. Metaphorical images of organization 1 5 6 3

principle is embodied in the observation that metaphors are more apt and fitting when they relate target and source concepts that are more exact of one another (e.g. Katz, 1992; Tourangeau & Sternberg, 1981, 1982). Morgan (1980), in his earlier writings, equally emphasized that the most powerful use of metaphor arises where concepts are correlated that carry significant differences between them (i.e. the distance principle discussed below) but are nonetheless representative of one another, and can therefore be connected and integrated in a meaningful way. Most of the conceptual metaphors within the dominant root metaphorical categories of ‘animate being’, ‘machine’, ‘culture’, ‘system’, and ‘evolution’ in Tables 3 and 4 are examples of the application of the ‘integration principle’. For each of these conceptual metaphors, a frame for a source concept (e.g. ‘learning’ within ‘animate being’, see Table 3) has been mapped onto the target concept of ‘organization’. Such mappings are normally guided by perceived relation- ships of identity, similarity or analogy between the target and source (i.e. the integration heuristic), where these perceived commonalities provide the semantic rationale for the metaphorical correlation of the concepts involved (Gentner et al., 2001; Oakley, 1998).

The relational heuristic exerts normative pressure to construct and maintain metaphorical mappings in such a way as to preserve relational structure (Coulson & Oakley, 2000; Fauconnier & Turner, 1998, 2002). In research on metaphorical mappings, this pressure has been referred to as the invariance hypothesis; the observation that the underlying mappings in metaphoric expressions are almost always based on shared image-schematic structure (see Lakoff, 1990; Turner, 1987). Gentner and Clement (1988) have found in this respect that relational metaphors (i.e. those whose interpret- ation is based on relational properties) are also judged more apt than attribu- tive metaphors (i.e. those metaphors whose interpretation is based on non-relational properties, namely common object attributes, and are there- fore mere-appearance matches) (see also Gentner et al., 2001; Tsoukas, 1993, for this point). The following two examples illustrate this distinction between relational and attributive metaphors: ‘managerial cognition is captured in cognitive maps’ and ‘managerial cognition includes blind spots’. Clearly, a relational metaphor is expressed with the first example since its interpret- ation has to do with a relational property (e.g. ‘managers operating on environments under the guidance of a map that structures cognition and provides managers with the ground for hypotheses of action that may then be confirmed or disconfirmed when acted upon’, which, according to Gentner and Clement’s notation, is a relation between entities in the relevant domain: managers, their thoughts, and the environments that they act upon) (e.g. Reger & Palmer, 1996). In contrast, an attributive metaphor is

Human Relations 58(12)1 5 6 4

expressed in the second example since its interpretation involves non- relational properties, namely, a ‘missing element’ and/or ‘unnoticed event’. Many of the identified conceptual metaphors in Tables 3 and 4 satisfy the relational heuristic including those in the ‘animate being’ (e.g. ‘organizational identity’), ‘flow-change’ (e.g. ‘organizational change’) and ‘evolution’ (e.g. ‘organizational decline’) categories. However, attributive metaphors are also being developed and selected including the machine metaphor of ‘organiz- ational structure’ and the ‘animate being’ metaphor of ‘organizational memory’. The metaphor of ‘organizational memory’, for example, provides a now well-established lens for examining the distributional aspects of organizational cognition. The metaphor projects above all the attribute of ‘knowledge repositories’ or ‘storage bins’ onto organizational cognition, and in doing so it has provided a framework of ‘storage bins’ for researching how knowledge is conserved and retrieved by the socialization and control systems that constitute organizational cognition (i.e. routines, rules, appren- ticeships) (Walsh & Ungson, 1991).

The connection heuristic suggests that the representation in the metaphorical image should maintain its links to the input target and source concepts. Satisfaction of the connection heuristic is what allows one to access elements in the metaphorical image with names and descriptions from the input concepts, as well as what allows the projection of structure from the image to other applications and subjects, including the input target and source concepts. This heuristic is at work in many conceptual metaphors in the root metaphorical categories of ‘animate being’, ‘machine’, ‘system’, and ‘evolution’; the conceptual metaphors in these categories tap into and are connected to a rich body of knowledge on animate being and human behavior, machine structures and operations, and so forth. The dominant conceptual metaphor of ‘organizational learning’, as one example, likens organizations to animate beings with thinking and learning capacities of their own. More specifically, it likens the thinking capacities of organizations to the behavioral responses of organisms and makes further connections with the body of work on behavioral evolution. The metaphor can be traced back to Skinner’s (1935, 1938) operant conditioning, where the behavior of an individual is said to be emitted, then selected by environmental contingen- cies in much the same way that variations among individuals have been selected by the environment during the course of evolution. Learning theory, and its embodiment in ‘organizational learning’ is an obvious example of this metaphorical line of thinking tracing back to Skinner; learning is conceptu- alized as the acquisition of discriminating responses to an environment, with the environment posing as the stimulus. ‘Organization learning’, then, is completed and elaborated to an image of organizations as organisms that

Cornelissen et al. Metaphorical images of organization 1 5 6 5

can think and learn by acquiring discriminating responses to an environment, with the environment posing as the stimulus. The thinking or learning capaci- ties of organizations are framed as behaviors that are adaptive to the environ- ment and inferred, corrected and made routine on the basis of past experience (Crossan et al., 1999).

The availability heuristic refers to the ease with which general domains or classes of ideas can be brought to mind (e.g. Gilovich & Savitsky, 2002; Tversky & Kahneman, 1974) and are then metaphorically combined with the concept and domain of ‘organization’. In other words, the heuristic ‘refers to a tendency to form a judgment on the basis of what is readily brought to mind’ (Medin & Ross, 1997: 522) which, over the period 1993–2002 includes such general ideas and semantic domains as ‘evolution’ (e.g. Betton & Dess, 1985), ‘animate being’ or ‘actor’ (e.g. DiMaggio & Powell, 1983), ‘culture’ (e.g. Hofstede, 2002) and ‘machine’ (e.g. Morgan, 1996) that are salient for this period and are the source for specific concep- tual metaphors. Barley and Kunda’s (1992) historical study of the surges of managerial and organizational rhetorics provides further support for the availability heuristic in that it suggests certain historical patterns in the uptake of language, including metaphorical language, for the study of organizations. One particularly interesting observation made by Barley and Kunda (1992) is that the ‘systems’ root metaphor entered organizational thought and became rather dominant at around 1955 (and effectively triggered such dominant conceptual metaphors as ‘contingency’) as ‘systems’ thinking as a whole was salient at that time and pervaded intellectual and social scientific thinking. The availability heuristic can also be thought of as a pressure to use conventional mapping schemas that facilitate comprehen- sion. Thus construed, the heuristic applies pressure to use common and well- known root and conceptual metaphors, such as the link between seeing and knowing (e.g. ‘managerial scanning’), organizational development and evolution (e.g. ‘population ecology’), or between organizational perfor- mances and theater (e.g. ‘organizational theater’).

The distance heuristic is rooted in findings from empirical research which clearly suggest that for a metaphor to be apt and effective, the conjoined target and source concepts need to come from distant domains in our semantic memory. Tourangeau and Sternberg (1981, 1982) conceptual- ized this pressure as the search for ‘between-domains distance’, which must be fairly large for the metaphor to be effective because close distances provide little interaction or surprise (see also Blasko & Connine, 1993; Katz, 1989, 1992; McGlone & Manfredi, 2001; Trick & Katz, 1986). The premise in this regard is that when the distance between the semantic domains is high; it shocks researchers into conceiving of a subject in a completely new way.

Human Relations 58(12)1 5 6 6

This happens as the semantic anomaly which is evident in the tension or surprise that is triggered when distant domains are related needs to be overcome by the organizational researcher and ‘nudges us into noting some- thing’ that was not there before (Davidson, 1984: 253). Cornelissen (2004, 2005) has recently pointed to the importance of the distance principle in relation to organizational metaphors, arguing that, on average, effective metaphors are indeed characterized by distance in the semantic domains that the target and source concepts are drawn from. The ‘organizational identity’ metaphor, one of his examples from the root category of ‘animate being’, consists of the correlation of a target and source concept (i.e. ‘organization’ and ‘identity’) from distant referent domains (i.e. the social world of organiz- ations versus the psychological world of personality and identity formation), and has as such forced organizational researchers to create resemblances between these concepts and their respective domains that did not seem particularly related beforehand. Many of the conceptual metaphors within Tables 3 and 4 embody the distance heuristic in that the referent domains that they are sourced from (for example, ‘animate being’ and ‘human behavior’, ‘evolution’, ‘culture’, ‘flow and change’, ‘system’) are sufficiently distinct and distant from the world of organizations, at least when these metaphors were first suggested.

The concreteness heuristic, finally, refers to pressure to select concrete rather than abstract source concepts for metaphorical blending with a target concept. Katz (1989, 1992) produced empirical evidence suggesting that the aptness and effectiveness of a metaphor is higher when a concrete, rather than abstract, source concept is metaphorically compared to the target. The rather widespread ‘root’ metaphorical image of ‘organization’ as a kind of ‘machine’ illustrates this heuristic well. The ‘root’ machine metaphor suggests an integrated picture of organization as comprised of a series of mechanically structured interconnected parts; and is based upon a rather specific and concrete conception of ‘machines’ (e.g. Morgan, 1996; Tsoukas, 1993; Zohar & Morgan, 1996). Similarly, the computational ‘organizational memory’ metaphor draws upon an explicit and concrete source domain; the computer metaphor being familiar to most and allowing for rather precise inference and explication (Holyoak & Thagard, 1995). The ‘organizational learning’ metaphor equally draws upon explicit and well-understood prin- ciples of operant conditioning. What this suggests is that concrete source concepts have a currency for organizational researcher in that they are more easily mapped, used and understood. As such, the concreteness of the source concept is an important determinant of the development and selection of a metaphor by organizational researchers.

Cornelissen et al. Metaphorical images of organization 1 5 6 7

Discussion: Heuristics, metaphors and theory construction

Weick (1989) noted that organizational researchers, like scientists in other social scientific fields, not only direct themselves the imagination process but also subsequently select the theoretical representation(s) for the target subject under consideration. In one sense, this artificial selection process, to para- phrase Weick (1989), is reflected in the huge variety of ways in which the subject of organization itself has been thought of and represented. Organiz- ational researchers have over the years likened organizations to, for example, anarchies, seesaws, space stations, garbage cans, orchestras, savage tribes, octopoids, market places, data processing systems, athletic teams, organic systems, theaters, human beings, and machines, to name but a few (e.g. Cornelissen, 2004, 2005; Morgan, 1980, 1996; Oswick et al., 2002; Putnam et al., 1996; Tsoukas, 1991, 1993; Weick, 1979). The artificial, and there- fore in part subjective, nature of the imagination process has been interpreted by some commentators (e.g. Morgan, 1980, 1996) as suggesting that a continuous process of ‘imaginization’ – fully free and creative metaphorical thinking – is satisfactory enough to be a substitute for ‘organization’ (see also Tsoukas, 1993). We have shown, however, that this is a flawed inference, particularly when one considers the select range of animate being, systems, evolutionary, warfare, culture and machine metaphors that prevail in organization theory (see also Baum & Rowley, 2002). Thus, it appears that there must be certain heuristics at work which, ceteris paribus (e.g. political pressures), suggest which organizational metaphors are developed and selected; that is, are deemed most effective.

In other words, metaphorical imagination processes are not uncon- strained, and the six identified heuristics embody the rules and constraints by which metaphors are developed and selected. We suggest therefore that these heuristics are important determinants of the aptness of a metaphor (in the judgment of organizational scholars), and, as a corollary, of whether a metaphorical image resonates with organizational researchers and is subse- quently used within theorizing and research. We also discussed metaphors- in-use which embodied one or more of these principles, although, it needs to be mentioned, satisfaction of these heuristics is selective, and satisfying one heuristic may be inconsistent with satisfying another. To illustrate this point, the metaphorical image of ‘organizational mind’ – the idea that behaviors of organizational members are connected in such a way that they are in them- selves ‘mental’ in the sense of being capable or reflective of intelligent and creative thought (e.g. Sandelands & Stablein, 1987; Weick & Roberts, 1993) – fulfills the distance heuristic as it likens connected behaviors within

Human Relations 58(12)1 5 6 8

organization to neurological patterns in the brain (that produce an emergent effect and is then seen as ‘intentional’, ‘heedful’ or ‘mindful’). However, it is at odds with the concreteness heuristic as it is unclear what kind of neuron- like relationships from the notion of ‘mind’ are projected onto organizational behaviors. This is primarily due to the ongoing disagreement and debate on the workings of the mind in the neuropsychological source domain; in particular between those championing a computational connectionist or associative model of the mind (see, for example, Rumelhart & McClelland, 1986) as opposed to a neuropsychological view that considers the mind as a combinatorial architecture (see, for example, Dupuy, 2000). This confusion, in turn, has led to difficulties for organizational researchers in understanding and manipulating the metaphor for theorizing and research purposes. ‘Organizational mind’ is therefore often only referred to in a cursory way in academic writings (see, for instance, Orlikowski, 2002), and hardly figures directly in empirical research, if at all.

We argue that metaphorical images are often selective in the heuristics that they embody, and that the most apt and effective metaphors are the ones that satisfy multiple heuristics rather than a single one. We also suggest, following Weick (1989), that the creative use of metaphors is facilitated and/or constrained by practical factors and considerations before they are worked out into theoretical representations. Metaphorical thinking, in organization studies as elsewhere, can hardly be treated as some sort of disembodied or radically free play of the mind, limited (if at all) only by the past experiences, cognitive habits, and biases of individual researchers (e.g. Chia, 1996; Cornelissen, in press; Weick, 1989). That such treatments are sometimes proposed has been sufficient reason for Knorr-Cetina’s (1981) well-known and repeated insistence that metaphorical theories of theory construction and scientific innovation are incomplete. It is certainly true, as she argues, that scientists must ‘work out’ or ‘realize’ metaphorical concepts in the tangible, nitty-gritty process of ‘knowledge production’ that takes place in the field before any truly consequential innovations can be brought about (see also Beyer, 1992). Consequently, ‘the process of research produc- tion and reproduction is more complex than the equation of metaphor and innovation suggests’ (Knorr-Cetina, 1981: 66), which suggests that rather than viewing metaphorical imagination as the imposition of static images it actually involves a more evolving process or activity in which metaphorical images (as organizing structures) partially order and form a research community’s perspective and are modified by their embodiment in concrete experiences of research and further experimentation.

Cornelissen et al. Metaphorical images of organization 1 5 6 9

Governing rules

Given this important distinction between metaphorical images on the one hand and more fully formed theoretical representations (and their embodi- ment within the practice of organizational theorizing and research) on the other, we believe that the identified heuristics are important within ‘disci- plined imagination’ in providing heuristics at the level of the metaphorical image for considering whether an image is apt and effective (i.e. a fully inte- grated, meaningful and sufficiently rich image). Each of these heuristics has been used in isolation or in combination and to varying degrees. On this account, each of them has been found useful in simplifying and guiding the task of constructing metaphors and models for research, but, importantly, they may also lead to characteristic errors or biases (cf. Kahneman & Tversky, 1996). For example, use of the availability heuristic may lead to bias whenever memory retrieval is a biased cue to actual frequency because of the broader world’s tendency to call attention to examples of a particular restricted type. Similarly, a mere focus on the exactness heuristic as is the case with the ‘organization’ as ‘theater’ metaphor, instead of combining this heuristic with the distance heuristic, provides for ‘apt’ yet cognitively limited metaphors. That is, when concepts are related from not too distant domains a metaphor often fails to break any new ground or generate new insights that were inconceivable before (cf. Cornelissen, 2004; Oswick et al., 2002). Thus, organizational researchers need to be more conscious and aware of the heuristics that they use in developing and selecting metaphors, to a greater extent than before (Weick, 1989), and of the possible biases and errors involved. This is an important point, as even ‘experienced researchers are also prone to the same biases – when they think intuitively’ (Tversky & Kahneman, 1974: 1130).

While each of the mentioned heuristics has characteristic biases and errors attached to it (as indicating departures from a normative rational theory or probability accounts) (Gilovich & Savitsky, 2002), particularly important from the vantage point of organizational theorizing and research is that these heuristics and the metaphors that they produce lead to concep- tual advances and breakthroughs. This means that the choice and use of new metaphors (beyond existing ones), and thus the heuristics underlying them, needs to transcend the mere illustrative-cum-rhetoric level (at which metaphors have a communicative or rhetorical effect, lead to mere clarifi- cation, or scout out old ground) and rather provides for creative and cogni- tively profound new insights and conceptual advances that were inconceivable before. Against this criterion, two of the abovementioned heuristics are particularly important: the relational and distance heuristics.

Human Relations 58(12)1 5 7 0

Both these heuristics, on their own account and in combination, have the potential to produce conceptual advances, thus going beyond mere similarity matches between the target and source concepts of a metaphor. Two govern- ing rules in relation to these two heuristics are therefore proposed; to aid organizational researchers in their search for novel categorizations and deeper insights through metaphor.

Governing rule #1: relational metaphors are preferred over attributive metaphors. This first governing rule, related to the relational heuristic, is based upon the evidence (see Gentner & Clement, 1988; Gentner et al., 2001) that there is a marked difference between the effect of an attributive metaphor (which may provide conceptual clarification and frameworks through perceived common object-attributes between the target and source concepts) and a relational metaphor. The relational metaphor, through its projection and mapping of interconnected relations between previously unrelated concepts, has the potential to produce novel cognitive categorizations and new frames for research- ing the world of organizations.

Governing rule #2: it is preferred in metaphor to conjoin concepts from semantic domains that are in the first instance seen as distant from one another. This second governing rule, related to the distance heuristic, implies that relating concepts from distant semantic domains has potentially a greater cognitive effect than juxtaposing concepts from semantic domains that are perceived as close. The guiding premise here is as mentioned that distance provides for strong and cognitively profound metaphorical imagery and that without perceived (sufficient) distance a metaphor fails to shock organizational researchers into conceiving of a subject in a completely new way. That is, in such cases, a metaphor may have conjoined two concepts that are conceived of as in some sense alike (in either an attributive or relational sense), but because the semantic domains are seen as not too distant it has little further cognitive effect. In effect, such a metaphor may then just lead into a re-labeling of the targeted subject with concepts and terms from the source domain, but at a rather superficial, nominal level and without offering any new and truly profound insights.

Together, these governing rules encourage organizational researchers to search for creative and new ways of conceptualizing organizations. This is particularly important when considering that organizational researchers may at times have been conservative in selecting mere similarity matches in

Cornelissen et al. Metaphorical images of organization 1 5 7 1

and through metaphor (Oswick et al., 2002). Oswick et al. (2002) have recently argued in this respect that many organizational researchers remain in the ‘cognitive comfort zone’ when they develop and select metaphors, primarily focusing on the similarities or overlapping ground between closely related concepts, and not on the dissimilarities or ‘tension’ that may exist when comparing more distant concepts and semantic domains. In these instances, metaphors are best seen as a means of elaborating and explicating already existing knowledge, as in their focus on similarities and resemblances between closely related concepts they merely make ‘the familiar more familiar’ (Oswick et al., 2002: 295). The two governing rules cater instead for a more progressive and advanced use of metaphor with metaphor being used to reveal deeper and more profound insights into the world of organiz- ations. When used in such a way, we believe that metaphors can prove enor- mously productive of further theoretical advances and empirical observations within organization studies; by sparking off inquiry and directing researchers to explore links that would otherwise remain obscure.

Limitations

Our study of metaphors-in-use within organization theory depends on a specific cognitive linguistic conception of metaphor production, comprehen- sion and use. This theoretical conception, as we have suggested, is particu- larly effective in providing an account of how metaphorical reasoning underlies, in a cognitive sense, much organizational theorizing and research. That is, our empirical study was grounded in a theoretical framework that attempts to explain how conceptual structure is invoked in metaphor use. No single theory, however, provides a comprehensive account of how people understand all kinds of metaphorical language, given all the temporal moments of understanding that are discussed by metaphor scholars (compre- hension, recognition, interpretation, appreciation, use). Theories based in cognitive linguistics best explain metaphor comprehension, interpretation and use, whereas other theories such as speech act theory and rhetorical theory may be better at explaining metaphor recognition and appreciation.

As is the case with all research methods, there are also limitations to the strategy of trying to infer something about conceptual structure from a systematic analysis of linguistic structure and behavior. The primary limi- tation is one shared by most linguistic research, namely, the problem of making conclusions about phenomena based on the analysts’ own moti- vated explanations. Psychologists have often argued that there is some circularity in how cognitive linguists argue for the psychological reality of

Human Relations 58(12)1 5 7 2

conceptual metaphors (Murphy, 1996, 1997). Thus, trying to infer aspects of conceptual knowledge from an analysis of systematic patterns of linguis- tic structure results in theories that seem post hoc. For instance, the claim that ‘every large corporation needs to have a corporate strategy’ is due to the presence of an independent, preexisting conceptual metaphor (i.e. the warfare strategy metaphor) provides only a motivated explanation for linguistic behavior. What psychologists seek is empirical, objective evidence that people’s conceptual knowledge somehow predicts the existence of different linguistic behavior, not that people’s linguistic behavior can be explained post hoc by positing conceptual metaphors (Gibbs, 1996, 1998). Within our research, we indeed assumed that metaphors as linguistic utter- ances reflect and intimate cognitively fundamental meanings about organiz- ations and organizational life; and that these meanings can be traced and inferred through a cognitive linguistic analysis. This assumption was guided by our research objectives of mapping and documenting the past and contemporary metaphors-in-use within organization theory, and of provid- ing a motivated explanation for the heuristics underlying their development, selection and use. Our research design was therefore justifiably post hoc. We inferred the ‘disciplined imagination’ and heuristics used by organiz- ational researchers from the identified patterns of linguistic data. Further experimental psychological research is worthwhile to corroborate our research findings and in particular for testing and refining this motivated explanation of the heuristics governing the development and selection of metaphors by organizational researchers.

A further limitation is that we only focused on word combinations involving the adjective ‘organizational’ and the noun ‘organization’ as target terms within a select set of management journals. This was essential to focus the research study. Nonetheless, there is scope for further research that focuses on other adjectives like ‘corporate’ or nouns such as ‘business’ as target terms and explores the extent to which such adjective–noun and noun–noun combinations (e.g. ‘corporate strategy’, ‘business domain’) are metaphorical. A final limitation is that we inferred the conceptual metaphors-in-use by mapping the direct reference of a combination of words (e.g. ‘organizational structure’) within the abstracts of publications within the SSCI database, rather than tracing the genealogy of certain word combinations and their appropriation in different schools of thought and within different communities of researchers (cf. Cornelissen, in press; Danziger, 1990) although, of course, this is also a subject warranting further research.

Cornelissen et al. Metaphorical images of organization 1 5 7 3

Note

1 Organization theory is defined here as the academic enterprise concerned with the study of organizational phenomena; and as such includes studies at both the micro (e.g. research into organizational behavior) and macro levels of analysis (e.g. research into organizational populations and organizational fields) (cf. Tsoukas & Knudsen, 2003) and specialist research areas such as strategic management, human resources, operations management and international business.

References

Alvesson, M. The play of metaphors. In J. Hassard & M. Parker (Eds), Postmodernism and organizations. London: Sage, 1993, pp. 114–31.

Bacharach, S.B. Organizational theories: Some criteria for evaluation. Academy of Manage- ment Review, 1989, 14, 496–515.

Barley, S.R. & Kunda, G. Design and devotion: Surges of rational and normative ideolo- gies of control in managerial discourse. Administrative Science Quarterly, 1992, 37, 363–400.

Baum, J.A.C. & Rowley, T.J. Organizations: An introduction. In J.A.C. Baum (Ed.), Companion to organizations. Oxford: Blackwell, 2002, pp. 8–34.

Betton, J. & Dess, G.G. The application of population ecology models to the study of organizations. Academy of Management Review, 1985, 10, 750–7.

Beyer, J.M. Metaphors, misunderstanding and mischief: A commentary. Organization Science, 1992, 3, 467–75.

Blasko, D. & Connine, C. Effects of familiarity and aptness on metaphor processing. Journal of Experimental Psychology, 1993, 19, 295–308.

Bourgeois, V.W. & Pinder, C. Contrasting philosophical perspectives in administrative science: A reply to Morgan. Administrative Science Quarterly, 1983, 28, 608–13.

Cameron, L. Identifying and describing metaphor in spoken discourse data. In L. Cameron & G. Low (Eds), Researching and applying metaphor. Cambridge: Cambridge University Press, 1999, pp. 105–32.

Chia, R. Metaphors and metaphorization in organizational analysis: Thinking beyond the thinkable. In D. Grant & C. Oswick (Eds), Metaphor and organizations. London: Sage, 1996, pp. 127–45.

Cornelissen, J.P. The merit and mischief of metaphor: A reply to Gioia, Schultz and Corley. British Journal of Management, 2002, 13, 277–9.

Cornelissen, J.P. What are we playing at? Theatre, organization and the use of metaphor. Organization Studies, 2004, 25, 705–26.

Cornelissen, J.P. Beyond compare: Metaphor in organization theory. Academy of Manage- ment Review, 2005, 30, 751–64.

Cornelissen, J.P. Metaphor and the dynamics of knowledge in organization theory: A case study of the organizational identity metaphor. Journal of Management Studies, in press.

Coulson, S. Semantic leaps: Frame-shifting and conceptual blending in meaning construc- tion. Cambridge: Cambridge University Press, 2001.

Coulson, S. & Oakley, T. Blending basics. Cognitive Linguistics, 2000, 11, 175–96. Crossan, M., Lane, H.W. & White, R.E. An organizational learning framework: From

intuition to institution. Academy of Management Review, 1999, 24, 522–37. Daft, R.L. & Weick, K.E. Toward a model of organizations as interpretation systems.

Academy of Management Review, 1984, 9, 284–95. Daft, R.L. & Wiginton, J.C. Language and organization. Academy of Management Review,

1979, 4, 179–91.

Human Relations 58(12)1 5 7 4

Danziger, K. Generative metaphor and the history of psychological discourse. In D.E. Leary (Ed.), Metaphor in the history of psychology. Cambridge: Cambridge University Press, 1990, pp. 331–56.

Davidson, D. Inquiries into truth and interpretation. Oxford: Clarendon Press, 1984. DiMaggio, P.J. & Powell, W.W. The iron cage revisited: Institutional isomorphism and collec-

tive rationality in institutional fields. American Sociological Review, 1983, 48, 147–60. Dupuy, J.-P. The mechanization of the mind: On the origins of cognitive science. Prince-

ton, NJ: Princeton University Press, 2000. Fauconnier, G. Mappings in thought and language. Cambridge: Cambridge University

Press, 1997. Fauconnier, G. & Turner, M. Conceptual integration networks. Cognitive Science, 1998,

22, 133–87. Fauconnier, G. & Turner, M. The way we think. New York: Basic Books, 2002. Gentner, D., Bowdle, B., Wolff, P. & Boronat, C. Metaphor is like analogy. In D. Gentner,

K.J. Holyoak & B.N. Kokinov (Eds), The analogical mind: Perspectives from cognitive science. Cambridge, MA: MIT Press, 2001, pp. 199–253.

Gentner, D. & Clement, C. Evidence for relational selectivity in interpreting analogy and metaphor. In G.H. Bower (Ed.), The psychology of learning and motivation. New York: Academic Press, 1988, pp. 307–58.

Gentner, D. & Markman, A.B. Structure mapping in analogy and similarity. American Psychologist, 1997, 52, 45–56.

Gibbs, R.W., Jr. Why many concepts are metaphorical. Cognition, 1996, 61, 309–19. Gibbs, R.W., Jr. The fight over metaphor in thought and language. In A.N. Katz (Ed.),

Figurative language and thought. Oxford: Oxford University Press, 1998, pp. 88–118. Gilovich, T. & Griffin, D. Heuristics and biases: Then and now. In T. Gilovich, D. Griffin

& D. Kahneman (Eds), Heuristics and biases: The psychology of intuitive judgment. Cambridge: Cambridge University Press, 2002, pp. 1–18.

Gilovich, T. & Savitsky, K. Like goes with like: The role of representativeness in erroneous and pseudo-scientific beliefs. In T. Gilovich, D. Griffin & D. Kahneman (Eds), Heuris- tics and biases: The psychology of intuitive judgment. Cambridge: Cambridge University Press, 2002, pp. 617–24.

Gioia, D.A. & Pitre, E. Multiparadigm perspectives on theory building. Academy of Management Review, 1990, 15, 584–602.

Grant, D. & Oswick, C. (Eds) Metaphor and organizations. London: Sage, 1996. Hassard, J. Essai: Organizational time: Modern, symbolic and postmodern reflections.

Organization Studies, 2002, 23, 885–93. Hofstadter, D. and the Fluid Analogies Research Group. Fluid concepts and creative analo-

gies. New York: Basic Books, 1995. Hofstede, G. Culture’s consequences: Comparing values, behaviors, institutions, and

organizations across nations. London: Sage, 2002. Holyoak, K. & Thagard, P. Mental leaps: Analogy in creative thought. Cambridge, MA:

MIT Press, 1995. Hopper, P.J. & Thompson, S.A. The discourse basis for lexical categories in Universal

Grammar. Language, 1984, 60, 703–52. Hunt, S.D. & Menon, A. Metaphors and competitive advantage: Evaluating the use of

metaphors in theories of competitive strategy. Journal of Business Research, 1995, 33, 81–90.

Inns, D. Metaphor in the literature of organizational analysis: A preliminary taxonomy and a glimpse at a humanities-based perspective. Organization, 2002, 9, 305–30.

Kahneman, D. Maps of bounded rationality: A perspective on intuitive judgment. Nobel Prize lecture, Stockholm, Sweden, 8 December 2002.

Kahneman, D. & Tversky, A. On the reality of cognitive illusions: A reply to Gigerenzer’s critique. Psychological Review, 1996, 103, 582–91.

Cornelissen et al. Metaphorical images of organization 1 5 7 5

Katz, A.N. On choosing the vehicles of metaphors: Referential concreteness, semantic distances, and individual differences. Journal of Memory and Language, 1989, 28, 486–99.

Katz, A.N. Psychological studies in metaphor processing: Extensions to the placement of terms in semantic space. Poetics Today, 1992, 13, 607–32.

Knorr-Cetina, K.D. The manufacture of knowledge: An essay on the constructivist and contextual nature of science. Oxford: Pergamon Press, 1981.

Lakoff, G. The invariance hypothesis: Is abstract reason based on image-schemas? Cogni- tive Linguistics, 1990, 1, 39–74.

Lakoff, G. The contemporary theory of metaphor. In A. Ortony (Ed.), Metaphor and thought, 2nd edn. Cambridge: Cambridge University Press, 1993, pp. 202–51.

Lakoff, G. & Johnson, M. Metaphors we live by. Chicago, IL: University of Chicago Press, 1980.

Leary, D.E. Psyche’s muse: The role of metaphor in the history of psychology. In D.E. Leary (Ed.), Metaphor in the history of psychology. Cambridge, MA: Cambridge University Press, 1990, pp. 1–78.

Lenat, D.B. The nature of heuristics. Artificial Intelligence, 1982, 19, 189–249. McCloskey, D. Metaphors economists live by. Social Research, 1995, 62, 215–37. McGlone, M.S. & Manfredi, D.A. Topic-vehicle interaction in metaphor comprehension.

Memory and Cognition, 2001, 29, 1209–19. McKinley, W., Mone, M.A. & Moon, G. Determinants and development of schools in

organization theory. Academy of Management Review, 1999, 24, 634–48. Medin, D.L. & Ross, B.H. Cognitive psychology. Orlando, FL: Harcourt Brace Jovanovich,

1997. Morgan, G. Paradigms, metaphors and puzzle solving in organizational theory. Adminis-

trative Science Quarterly, 1980, 25, 605–22. Morgan, G. More on metaphor: Why we cannot control tropes in administrative science.

Administrative Science Quarterly, 1983, 28, 601–7. Morgan, G. Images of organization. Beverly Hills, CA: Sage, 1986. Morgan, G. Is there anything more to be said about metaphor? In D. Grant & C. Oswick

(Eds) Metaphor and organizations. London: Sage, 1996, pp. 227–40. Murphy, G.L. On metaphoric representation. Cognition, 1996, 60, 173–204. Murphy, G.L. Reasons to doubt the present evidence for metaphoric representation. Cogni-

tion, 1997, 62, 99–108. Oakley, T. Conceptual blending, narrative discourse, and rhetoric. Cognitive Linguistics,

1998, 9, 320–60. Orlikowski, W.J. Knowing in practice: Enacting a collective capability in distributed

organizing. Organization Science, 2002, 12, 249–74. Oswick, C., Keenoy, T. & Grant, D. Metaphor and analogical reasoning in organization

theory: Beyond orthodoxy. Academy of Management Review, 2002, 27, 294–303. Pinder, C. & Bourgeois, V.W. Controlling tropes in administrative science. Administrative

Science Quarterly, 1982, 27, 641–52. Putnam, L., Phillips, N. & Chapman, P. Metaphors of communication and organization.

In S. Clegg, C. Hardy & W. Nord (Eds), Handbook of organizational studies. London: Sage, 1996, pp. 375–408.

Reger, R. & Palmer, T.B. Managerial categorization of competitors: Using old maps to navigate new environments. Organization Science, 1996, 7, 22–39.

Rumelhart, D.E. & McClelland, J.L. Parallel distributed processing: Explorations in the microstructure of cognition, Vols 1 and 2. Cambridge, MA: MIT Press, 1986.

Sandelands, L.E. & Srivatsan, V. The problem of experience in the study of organizations. Organization Studies, 1993, 14, 1–25.

Sandelands, L.E. & Stablein, R.E. The concept of organization mind. In S. Bacharach &

Human Relations 58(12)1 5 7 6

N. DiTomaso (Eds), Research in the sociology of organizations. Greenwich, CT: JAI Press, 1987, pp. 135–61.

Schön, D.E. Generative metaphor: A perspective on problem setting in social policy. In A. Ortony (Ed.), Metaphor and thought, 2nd edn. Cambridge: Cambridge University Press, 1993, pp. 137–63.

Skinner, B.F. The generic nature of the concepts of stimulus and response. Journal of General Psychology, 1935, 12, 40–65.

Skinner, B.F. The behavior of organisms: An experimental analysis. New York: Appleton Century, 1938.

Stapel, D.A. & Koomen, M. Interpretation versus reference framing: Assimilation and contrast effects in the organizational domain. Organizational Behavior and Human Decision Processes, 1998, 76, 132–48.

Steen, G. From linguistic to conceptual metaphor in five steps. In R.W. Gibbs Jr, & G. Steen (Eds), Metaphor in cognitive linguistics. Amsterdam: John Benjamins, 1999, pp. 57–77.

Thiétart, R. & Forgues, B. Chaos theory and organizations. Organization Science, 1995, 6, 19–42.

Tourangeau, R. & Rips, L. Interpreting and evaluating metaphors. Journal of Memory and Language, 1991, 30, 452–72.

Tourangeau, R. & Sternberg, R. Aptness in metaphor. Cognitive Psychology, 1981, 13, 27–55.

Tourangeau, R. & Sternberg, R. Understanding and appreciating metaphors. Cognition, 1982, 11, 203–44.

Trick, L. & Katz, A.N. The domain interaction approach to metaphor processing: Relating individual differences and metaphor characteristics. Metaphor and Symbolic Activity, 1986, 1, 185–213.

Tsoukas, H. The missing link: A transformational view of metaphors in organizational science. Academy of Management Review, 1991, 16, 566–85.

Tsoukas, H. Analogical reasoning and knowledge generation in organization theory. Organization Studies, 1993, 14, 323–46.

Tsoukas, H. & Chia, R. On organizational becoming: Rethinking organizational change. Organization Science, 2002, 13, 567–82.

Tsoukas, H. & Knudsen, C. Introduction: The need for meta-theoretical reflection in organization theory. In H. Tsoukas & C. Knudsen (Eds), The Oxford handbook of organization theory. Oxford: Oxford University Press, 2003, pp. 1–36.

Turner, M. Death is the mother of beauty: Mind, metaphor, criticism. Chicago, IL: University of Chicago Press, 1987.

Tversky, A. & Kahneman, D. Judgement under uncertainty: Heuristics and biases. Science, 1974, 185, 1124–31.

Von Ghyczy, T. The fruitful flaws of strategy metaphors. Harvard Business Review, 2003, 81, 86–94.

Walsh, J.P. & Ungson, G.R. Organizational memory. Academy of Management Review, 1991, 16, 57–91.

Weick, K.E. The social psychology of organizing, 2nd edn. Reading, MA: Addison-Wesley, 1979.

Weick, K.E. Theory construction as disciplined imagination. Academy of Management Review, 1989, 14, 516–31.

Weick, K.E. What theory is not, theorizing is. Administrative Science Quarterly, 1995, 40, 385–90.

Weick, K.E. & Roberts, K.H. Collective mind in organizations: Heedful interrelating on flight decks. Administrative Science Quarterly, 1993, 38, 357–82.

Zohar, A. & Morgan, G. Refining our understanding of hypercompetition and hyper- tubulence. Organization Science, 1996, 7, 460–5.

Cornelissen et al. Metaphorical images of organization 1 5 7 7

Human Relations 58(12)1 5 7 8

Joep P. Cornelissen is Senior Lecturer in Corporate Communications at Leeds University Business School.He previously worked at the Amster- dam School of Communications Research, University of Amsterdam. His research interests include the management of corporate communications and the use of metaphor in management and organization theory and practice. He is author of Corporate communications: Theory and practice (Sage). He teaches mainly on MBA and MA programs in the areas of corporate communications.He is currently principal investigator on a UK Economic and Social Research Council project on the use of metaphors in organization theory. His research articles on metaphor have appeared in Academy of Management Review, Organization Studies, British Journal of Management, Psychology and Marketing and the Journal of Management Studies. [E-mail: [email protected]]

Mario Kafouros is currently an ESRC Research Fellow at Leeds University Business School. He is an electronic engineer by first degree, and also holds a degree in economics and a PhD in management. He has extensive industrial and academic experience in the field of innovation. His publications include articles in journals of management but also in journals of economics.His research interests include the effects of knowl- edge spillovers, the impact of innovation on productivity performance, and the role of metaphors in organization theory. [E-mail: [email protected]]

Andrew R. Lock is Professor of Marketing and Business Administration and Dean of the Business School at the University of Leeds. He holds a Masters and PhD from London Business School. He was previously Pro- Vice-Chancellor and Dean of Faculty at Manchester Metropolitan University and has also held lecturing posts at the University of British Columbia and Kingston Polytechnic (now Kingston University). He is a past Chair of the Association of Business Schools. His work has been published in the Journal of Marketing Management, the European Journal of Marketing, the Journal of Advertising Research, the Journal of the OR Society, the Journal of Public Affairs and Management Learning. [E-mail: [email protected]]