Contribution
It takes Albert and Whetten’s (1985) definition of organizational identity, “central, enduring, and distinctive”, and argues the “enduring” pillar away. Identity is constructed in interaction with outsiders, image in its several forms feeds back on it, and the result is an identity that changes in meaning over short periods while keeping the same labels. The authors argue that this instability is useful to the organization, not a threat to it.
Key insight
The durability is in the labels, not in their meaning. “We stand for service!” or the “H-P Way” persist while what they mean is reinterpreted, so an identity can have continuity without being enduring. The paper separates the two: “An identity with a sense of continuity … is one that shifts in its interpretation and meaning while retaining labels for ‘core’ beliefs and values that extend over time and context” (p. 64).
Assumptions
- Identity is “an organization’s members’ collective understanding of the features presumed to be central and relatively permanent” (p. 64), socially constructed, at the organizational level by the same processes as at the individual (Cooley, Gergen, Weick).
- Members are also outsiders (customers, audiences of the media), so they are sensitive to how others see the organization.
- The first, process-model account rests on realist assumptions (identity has some substantive basis and changes incrementally); the revisionist-history and postmodern accounts that follow are nominalist (p. 63–64).
Key results
The results are conceptual.
- Forms of image (Table 1). Construed external image (members’ view of how outsiders see the organization), projected image, desired future image, corporate identity, transient impression and reputation, “relatively stable, long-term, collective judgments by outsiders”.
- Process model (Figure 1, pp. 67–70). A triggering external feedback or event prompts members to compare identity with construed external image (“Who do we think we are?” against “Who do they think we are?”). No discrepancy affirms identity; a discrepancy may be ignored, or acted on by changing identity (helped by a projected desired future image) or by changing outsiders’ impressions (communicating identity more clearly, impression management, or false images). Projected images return in “refracted” form and the comparison recurs. Shell’s Brent Spar (1991–1999) runs through every branch.
- Revisionist history (pp. 71–72). Members reinterpret the past to fit present and desired images: “The ‘facts’ of the past might not be in doubt, but their meaning always is.” This “makes identity appear stable to perceivers, even as it changes.”
- Postmodern lens and its limits (pp. 72–74). Baudrillard’s “everything ends up as image” is set against three limits: the environment’s demand for recognizable, stable partners; evidence that members keep identity apart from image even when image is flattering (Elsbach and Kramer’s business-school rankings); and multiple identities that do not shift together and so damp wholesale change. Identity “is inevitably influenced by image but does not necessarily become image”.
- Adaptive instability (pp. 74–75). Instability “allows better adaptation”, reactively (IBM’s response to its image of a ponderous mainframe maker) or proactively (top managers introducing “ambiguity-by-design” and a desired future image, in Gioia and Chittipeddi’s and Gioia and Thomas’s university).
- For research (pp. 75–77). Cross-sectional studies may treat identity as stable; longitudinal ones may not. The authors ask how to account for ephemeral organizations, “hollow corporations”, and “the Florida Marlins baseball team, which won the 1997 World Series and then was decimated in the space of months by the trading of key players”.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Organizational identity is better seen as continuous than as enduring: labels persist while meanings shift | moderate | conceptual, with the H-P Way and literature on individual identity |
| C2 | Discrepancy between identity and construed external image triggers change in identity or in image | moderate | builds on Dutton and Dukerich (1991) and Dutton et al. (1994); illustrated by Shell |
| C3 | Identity instability is adaptive for the organization | weak | argued; illustrated by selected cases of successful change, no counter-cases |
| C4 | Image does not wholly replace identity | moderate | three limiting arguments, one study (Elsbach and Kramer 1996) |
Concepts
- Organizational identity: members’ collective understanding of central, distinctive features presumed relatively permanent.
- Continuity versus endurance: continuity keeps labels while meanings shift; endurance would keep both.
- Construed external image: members’ beliefs about how outsiders see the organization (Dutton et al. 1994’s relabelling).
- Reputation: outsiders’ “relatively stable, long-term, collective judgments”.
- Adaptive instability: identity’s instability, through its interplay with image, as an aid to change.
Connections
It builds on Albert and Whetten’s definition, Dutton and Dukerich’s Port Authority study and Gagliardi’s “change to remain what they have always been”. Cloutier and Ravasi’s identity trajectories (LIT-1141, not reached) is a later longitudinal study of the continuity the paper asks researchers to study. Feldman and Pentland (LIT-1124, NOTE-1386) make a parallel move for routines: the ostensive aspect persists as a gloss while performances change what it means, as identity labels do here.
Bearing on the record
- CASE-103. The case’s split of a fanbase’s reputation into members’ behaviour (P1), others’ belief (P2) and the belief’s effect on the fanbase (P3) is close to the paper’s split between identity, reputation and construed external image, and the paper’s process model is A5’s insider-identification path: members compare who they think they are with who they think outsiders think they are, and change one or the other. The paper also makes reputation the slow variable (“relatively stable, long-term”), against an identity that can shift quickly, which fits the case’s reputation that outlasts the behaviour that earned it. It does not study a case where the reputation is wrong, and it does not consider the case’s selection path, a reputation that attracts or repels members.
- CLAIM-329. The paper’s continuity of labels with changing meanings is a challenge to any persistence criterion that requires identity content to stay fixed, as A6 expected. It does not supply the claim’s causal-transmission criterion: its continuity is members’ interpretive work and outsiders’ demand for recognizable partners, and its examples (the Marlins’ gutted roster among them) are posed as puzzles, not answered.
- No instruction for machine-learning practice; nothing for the anthology.
Limitations
- No data. The illustrations are chosen successes of deliberate change (IBM, Oticon, British Airways, an unnamed Pittsburgh firm told in “a personal conversation”); the paper does not show that instability is adaptive on average or say when it is not, beyond warning of “the unintended substitution of faddish image for key values”.
- “Adaptive” is not defined against a measure of fitness or survival.
- Slips. Bourdieu is “Bourdieau” in the text and the reference list; “Wnen” for “When” on p. 68; Shell’s “ca series” on p. 67. The Brent Spar account says Greenpeace’s misinformation about the platform’s contents was “eventually refuted”; the paper gives no source for it.
Open questions
- When does a change of meaning under the same labels stop being the same identity continuing? The paper’s criterion is members’ felt continuity, which it also says revisionist history manufactures.
- Is adaptive instability more common than maladaptive instability, and how would one count?