Mission impossible? The paradoxes of stretch goal setting

Stretch goal setting is a process involving multiple and nested paradoxes. The paradoxical side of stretch is attractive because it holds great promise yet dangerous because it triggers processes that are hard to control. Paradoxes are not readily managed by assuming a linear relation between the here and now and the intended future perfect. Before adopting stretch goal setting, managers should thus be prepared for the tensions and contradictions created by nested or interwoven paradoxes. Achieving stretch goals can be as difficult for the managers seeking to direct the process as for designated delegates. While the increasing popularity of stretch goal setting is understandable, its unexpected consequences must be taken into account. The inadequate use of stretch goals can jeopardize the social sustainability of organizations as well as their societal support systems.


Introduction
Management by objectives is a core device that managers are obliged to use in a significant number of organizations (Drucker, 1954;Greenwood, 1981). Conventional wisdom for good management practice positions managing by goal setting (i.e. target setting, performance monitoring, and incentives/people management) as a central practice (Bloom et al., 2012). Despite criticism (Joullié, 2016;Ordóñez et al., 2009), the achievement of specific, difficult goals as a source of performance improvement is generally accepted as a key discipline of management practice. Goal setting forms a core element of organizational behavior's positivist focus (Miner, 1984), which we do not contribute to here: instead, we are more heedful of goal stretch as a mode of governmentality (Dean, 2010). Goal setting imposes focus, constrains behavioral strategies, canalizes feedback transmission, and valorizes goal attainment as a psychologically gratifying process (e.g. Locke et al., 1981;Latham, 1984, 1990). Merely achieving goals is now passé, however. Outstanding managers are expected to exceed in performance and thus achieve "stretch goals" "that are considered virtually unattainable" (Thompson et al., 1997: 48). The person who is constitutively minded to meet stretched goals has a quite specific subjectivity-not just performing in role but seeking to performatively extend their role achievement as a matter of course. It is a subjectivity that is disposed to be constantly striving to meet that which is stretched.
An element of neurosis is likely where stretch goals constitute a "radical" managerial "tool" (Sitkin et al., 2011: 556), promoted as the key that can unlock potential, increase performance, and achieve extraordinary results if only because today's extraordinary becomes tomorrow's normal and a further extraordinary must be achieved. To be constantly shifting the frontier of normalcy is hard enough for dedicated professional athletes; for ordinary managers for whom the extraordinary becomes a normal expectation, the regimen for achieving it constitutes a constant existential uncertainty as one lives in expectation of not being "good" enough to meet the escalating expectations. Stretched performance targets can be so ambitious that achieving them might qualify as "mission impossible" (Thompson et al., 1997). Finkelstein (2016) crystallizes a dominant managerialist approach to achieving these goals as an issue of leadership: some bosses become "super" because they make "exceptional people do the impossible" (p. 106). Impossibility, similar to excellence, is a moving target. Today's impossibility becomes tomorrow's norm prefiguring an endless horizon of potential innovation. Motivating people to achieve "virtually impossible" goals can be a recipe for failure that creates excess pressure for innovation. Processes perceived as a cause of extraordinary performance are likely, when widely adopted, to fail, and their failure becomes a further incentive to managerial demands to do better in the future. Much as Dylan's (1965) muse, we have to realize, analytically, that there is no success like failure, and failure is no success at all because, organizationally, it invites even greater stretching. Sitkin et al. (2011) argue that stretch goals lead to positive outcomes only in those rare cases where organizations combine sufficient slack resources with strong recent performance (that builds up self-confidence, or self-efficacy). Otherwise, expecting ordinary people to achieve the extraordinary can be a managerial weapon of mass destruction: both management and workers can be deeply stigmatized by the failure that follows with escalating effects on stakeholders urging them to do better in the future. Moreover, leaders who trigger exceptional performance on the part of a few employees can diminish the motivation of the rest of the organization.
We contribute to the emerging meta-theory of organizational paradox (Briscoe, 2016;Eisenhardt, 2000;Putnam et al., 2016;Schad et al., 2016;Smith and Lewis, 2011) and problematize the difficulties of using paradoxes that cannot be strictly managed or controlled. Paradox refers to persistent contradictions that require simultaneous address. Paradoxes are potentially nested (i.e. one level informs other levels) and interwoven (one tension informs other tensions); they are embedded in other paradoxes and display multiple interactions . Existing theorization of stretch is extended conceptually here by seeing stretch goals as involving a combination of paradoxical processes, nested and interwoven (e.g. Fairhurst et al., 2016). The tackling of one paradox may thus activate another previously dormant paradox. By activating stretch, managers may end up opening an organizational Pandora's box with unpredictable, but predictably extreme, consequences, which failure to achieve targets can only compound. In this way, we contribute to two main theory streams: we explore the paradoxical dimension of stretch which remains under-theorized and we use this discussion as a stepping stone to discuss the important topic of the nesting of organizational paradoxes.
We begin by defining stretch goals before articulating the relationship between stretch and paradox as well as the simultaneous triggers of paradox and contradiction of stretch goals. Next, we conceptualize the process tensions activated by stretching goals. Finally, we discuss paradoxical outcomes of stretch, such as success becoming a source of future failure, turning the positive into the negative (Miller, 1993). The study, overall, complements Sitkin et al. (2011) by discussing the process paradoxes that need to be tackled while adopting the powerful but demanding managerial technology of stretch.

What are stretch goals?
The notion of "stretch goals" refers to "goals that are extremely difficult and extremely novel" (Zhang and Jia, 2013: 994). They can be vertical, that is, aligned with current activities, or horizontal, that is, orthogonal to existing activities (Kerr and Landauer, 2004). Stretch goals are conceptually interesting because, according to standard goal theory, setting extremely difficult goals should diminish, not augment, their motivational appeal. In practice, as research suggests, stretch goals will be valid only under rare circumstances (Sitkin et al., 2011). In empirical contexts, setting such goals is a practice supported by managerialist value systems.

Stretch goals in practice
Stretch goals are seductive frames for management practice. Collins and Porras (1994) made the notion popular in the C-suite, by defining them as "hairy goals." Using managerial controls to push people beyond comfort zones in performance is a practice increasingly expected of managers in hypercompetitive contexts (D'Aveni, 1995). The setting and acceptance of goals otherwise perceived as unreasonable can be justified in familiar terms of "leadership." The popularity of success cases and the infrequency of reports of failed cases privilege attention to positive deviance that key figures, such as Jack Welch, served to mythologize (Hughes, 2001;Sitkin et al., 2011) as did key business cases: for instance, Southwest Airlines adopted stretch goals when top management dictated that planes be turned around at the gate in only 10 minutes (Chatterjee, 2005).
In practice, the stretch goal is set by managerial fiat and then employees are held responsible for working out how to attain the metrics. Setting stretch goals is a form of governmental device by which employees become governed by their own freedoms (Dean, 2010). Stressing the positives, there are several often-cited cases: Medtronic was "transformed" in the 1990s through setting stretch goals and then figuring out how to attain them (George, 2006). Steve Jobs became famous for stretching people to achieve the unachievable at Apple as well as at Pixar through his now famous "reality distortion field" (Catmull, 2014;Isaacson, 2011). Elon Musk, the CEO of Tesla Motors and SpaceX, is known for his "overly optimistic deadlines" (Vance, 2015: 68). Rocky Flats, the "most dangerous location in North America because of the quantity of radioactive materials, explosive devices, and volatile chemicals stored on site and the levels of contamination in the facility and surrounding land and groundwater" (Lavine and Cameron, 2012: 135), was cleaned in a record time as a stretched goal that management thought impossible to achieve (Cameron and Lavine, 2006;Lavine and Cameron, 2012). 1 While good news travels best in management circles, there is also a dark side to stretch goals as well. Practice is replete with examples-consider Nokia: Because there was always some lunatic who promised they'd do all these ten fine and wonderful things within the timeframe given by TMs [Top Managers] even though it wasn't true at all [that it would be possible] … TMs trusted these people when they said it's going to work out. (Vuori and Huy, 2015: 23) Second, there is the now infamous case of Volkswagen, where in a hostile and fear-ridden environment (Edmondson and Lei, 2014) employees cheated when pressured to reach an "impossible" goal (Ewing, 2015: B3).

Goal difficulty
If achieving a goal is perceived as "mission impossible" (Thompson et al., 1997: 49), it will predictably lead people to desist Latham, 1984, 1990) according to goal-setting theory. However, goal-stretching leaders confront managers and employees with impossible goals.
Perceptions and assumptions of possibility and impossibility are critical in defining and operationalizing stretch goals. In terms of the mentalities "governing the subjects" seeking to achieve these goals, the stretched goal, although its achievement is a managerial imposition, ideally, from a managerial perspective, becomes incorporated and embrained in employee volitions as a "technology of the self" teaching responsibilization and normalization (Foucault, 1991). Therefore, there is no objective measure of difficulty. Subjectivity matters and what is difficult for one person may be impossible to another (Vroom, 1964). That organizations are sometimes able to make the impossible possible problematizes the notion of impossibility as a "normal" expectation (Lavine and Cameron, 2012).

Extreme ambition and novelty
Stretch goals involve extreme ambition and novelty. Stretch goals are ambitious by definition: they aim to push the organization beyond its current limits by means of developing novel solutions. Increasing effort should not be qualified as stretching but as a high-intensity goal (see Table 1). While confronting high-intensity goals, people have to work harder but not necessarily smarter. Stretching goals involves difficulties and requires some element of novelty in approaching a problem. As Steve Kerr pointed out (in Sherman, 1995), if people already know how to reach a goal, it is not a stretch goal. A stretch goal is defined as a target that, beyond being characterized by an extreme level of difficulty, requires local innovation(s). It is because intensification of effort, per se, will not do that stretch goals require new strategies to approach the target. Stretch goals thus prompt people to work not only harder but also smarter. The problem is that stretch is a complex process, vulnerable to tension and contradiction, as will be explained next. Paradoxes permeate the whole process.

How stretch unleashes paradox
Stretch goals entail a new approach to an existing problem or challenge. According to the literature, three reasons explain why stretch goals are considered managerially viable solutions. First, when responding to jolts, stretch goals serve to trigger the organization's reactive responses. In this case, the literature predicts that stretch goals are triggered by fear. The fear of failure or of important external menaces poses existential organizational threats. Under these conditions, stretch goals function because of negative pressure and are activated by a perceived sense of urgency (Kotter, 1996). Second, when utilized as mechanisms against stagnation, they are supported by benchmarks. The isomorphic explanation is of the type "If our best competitors can do it, we have to be able to do it too." Selecting the right benchmarks can help support a stretch goal strategy. Finally, stretch goals are normally associated with a positive organizational identity, stressing what is core and enduring in an organization (Gioia et al., 2000) and its culture as Takeuchi et al. (2008: 100) argue, referring to Toyota.
Despite the emphasis on positives, organizations that might benefit from the adoption of stretch goals have to consider carefully their paradoxical component and the risks associated with stretch. Paradox refers to oppositions of elements that are valid per se but that persistently contradict each other when their articulation is demanded. Paradox articulates contradiction, interdependence, and persistence (Schad et al., 2016), characteristics that stretch goals richly exemplify. Stretching goals involves a contradiction between realism and imagination because "impossible" goals are imagined to be possible as a future perfect achievement (Pitsis et al., 2003). Implicitly, "possible" goals, by being less challenging and/or less meaningful, are constituted as less imaginatively exciting and their achievement insignificant. Interdependence occurs when a sense of organizational crisis is socially constructed and transmitted for which, if the crisis is to be met, the antidote is defined as collective achievement of stretch goals. Inherent contradiction and interdependence in goal achievement may not be resolvable, even where stretch goals are promoted as being achievable by persistent effort by those aiming to reach them.
Stretch goals are rich in paradoxical tension. They may become identity markers when used successfully. Success is only ever temporally specific. Today's success becomes tomorrow's Psychological safety is a fragile psychological process. Leader boldness in adopting stretch may itself threaten safety. Therefore, stretching perturbs the organizational functioning. 7 Feedback vs feedforward Stretch involves the use of past knowledge to leverage future knowledge creation. It involves feedback and feedforward.
Stretch goals, being seemingly impossible, are necessarily a leap into the unknown. In this sense, feedback and feedforward are unreliable navigation tools to create the unknown. 8 Learning vs forgetting Stretch involves forgetting as a precondition for learning.
The two processes of learning and forgetting are difficult to master. Combining them can be even more complicated. 9 Pride vs humility Stretch involves cultivating "humbition" as the synthesis of boldness and groundedness.
Leaders who stretch are often more ambitious than grounded.
mundane achievement setting the scene for further effort. Their positive use may be a target of criticism as depicted in Table 2. In this case, an organization institutionalizes extreme goals as a dimension of identity, as symbols of a culture of ambition, creativity, and innovation, serving to differentiate the organization as a culture of innovation (e.g. Apple; Isaacson, 2011). The internal choice dimension becomes reinforced, imbuing it with extra legitimacy as the organization embraces a stretch goal not because it is forced to do so but because it decides to do so to affirm its prized identity. Stretch goals, it is proposed, will be more effective when supported by a sense of internally generated crisis justified by external circumstances that legitimate rhetorics of innovation.

Paradoxical triggers
Organizations may feel tempted to articulate a stretch goal when confronted with a significant environmental jolt or when the previous routines are no longer adequate or possible (Meyer, 1982). Southwest Airlines' stretch goal of 10-minute turnarounds was adopted because of a reduction in aircraft in an already small fleet. To survive, a whole new approach to operational efficiency was necessary, stressing in accord with the literature that how leaders trigger stretch goals as well as how they use them is essential to the sensemaking and sensetaking elements involved in stretching (Gioia and Chittipeddi, 1991).
There are different process routes to stretching. When responding to jolts or external threats, leaders need to legitimate these as credible. In Bruch and Ghoshal's (2003) terms, they need to identify the "dragon" to be defeated in order to legitimate the rescue mission of saving a "princess." When stretch goals are not a response to an immediate threat but are used against stagnation, the pursuit has to be legitimated as a positive organizational purpose. In this case, the organizational response to challenge must be a self-activated process. Such activation can be facilitated by positioning challenges as system threats. When such a definition of the situation is accepted, it is real in its consequences, with the threat legitimating extreme action that enables stretch goals to become adopted. Extreme pressure may be perceived as not only normal but also as a desirable condition, a state of organizational excitement (Hewlett and Luce, 2006) that makes command leadership legitimate in the face of a crisis that has become defined as a system threat (Grint, 2005). Positive, volitional representations of stretch, however, can collide with the reality of stretch goal setting, for they elicit the paradoxical challenges that we explore next.

Paradox 1: being unreasonable and realistic
Why do people accept unrealistic goals? A first possibility, the negative case, is that in tough times, people accept extraordinary goals because difficult times require radical measures. In the second case, the positive approach, people may accept the call of a powerful and seductive vision. Building a piece of Olympic infrastructure to make a memorable event occur was a vital element in inducing future perfect strategies in the case that Pitsis et al. (2003) researched, for instance. Organizations convince people to accept an unusual amount of challenge and risk where some larger purpose than immediate interests is in view.
Stretch goals, paradoxically, are more effective when they combine realism, by building on existing elements (knowledge, experience, technology), and unrealism, where they involve a daring departure from existing elements (routines, mindsets, business models). In the case of culturally sustained stretch goals, a previous history of stretch and achievement can mark the organization as one in which those who dare to risk achieve and are celebrated. These cultures encourage strong member identification with charismatic leadership (Van Knippenberg and Sitkin, 2013). When such identification is strong, individuals accept the prevalence of collective over individual interest. As pointed out by Spector (2015: 8), "the presence of a strong, attractive individual leader exacerbated the tendency to submerge the individual into the group." However, there is a paradox: as the same author adds, revisiting Freud, unconditional admiration of a leader renders followers vulnerable to all the excesses that unbridled egoism can unleash. Subordinates of charismatic leaders sometimes accept being dominated and challenged beyond what seems reasonable because of the egotistically grand visions their leaders espouse, as biographical studies indicate (Stone, 2013;Vance, 2015; see also Pfeffer, 2015) or because they are coerced rather than inspired. Followers accept such ill treatment, rationalizing their "subjugation" (Pfeffer, 2015), including acceptance of impossible goals in order to express their dedication, despite being imposed on by the leader's high expectations. Such leaders manage to synthesize in their followers a mixture of admiration and fear. Ordinarily, this type of behavior is called bullying, but some organizations institutionalize abuse as a normal mode of being: bullying toward a vision. Members seek to match the leader's best expectations, fearing that resistance is not viable (Courpasson et al., 2012). The interaction between processes (e.g. top-down and bottom-up) helps to explain the unpredictability of the outcomes of stretch due to the interaction between nested/interwoven paradoxes. For instance, subordinates motivated by fear may feel psychologically unsafe (see paradox 6). This in turn can trigger a dynamic that cannot be anticipated by stretch goals setters and that results from cross-level interaction rather from linear dynamics, as predicted by paradox theory .
The very adoption of stretch goal setting, in summary, implies the articulation of two paradoxical tensions. To create a sense of urgency (Kotter, 1996), a combination of high internal control and strongly encapsulated external justification is required, suggesting, simultaneously, management being both in control and not in control. The paradox is that stretching is represented as being both simultaneously realistic and aspirational. Realism comes from past experiences of success, a sense of confidence in the organization's capabilities, and the fact that new goals are projected from previous successes. Extreme aspirational goals need to be phenomenologically experienced as both potentially individually and collectively achievable, as well as extraordinarily demanding. The 10-minute turnaround at Southwest offers an illustration. People fighting for their jobs (Freiberg and Freiberg, 1996: 34) were seeking to control their destiny in a situation spiraling out of control (Tichy and Sherman, 1993).

Paradox 2: crisis-endogenous versus exogenous
Stretch goals can be used endogenously in an autogenic crisis, one that is intentionally, internally generated (Barnett and Pratt, 2000), where they are used to disturb situations in which organizations, insisting on improving routines incrementally, become trapped in increasingly exploitative practices preventing them from renewal (Miller, 1993;Piao and Zajac, 2016). The lack of external justification will potentially reduce motivational power because of the absence of a credible legitimation. An autogenic stretch goal indicates that the organization is in control and that senior executives are triggering action rather than reacting to crisis. Communicated convincingly, especially when accompanied by some sense of urgency, it will induce people to accept the challenge. However, the sense of crisis needs to be represented not as a result of caprice but as necessary for present or future reasons. Paradoxically, to tighten control through stretch, the management has to first represent itself as losing control.

Tensional process
To sustain positive stretch goals, organizations should anticipate paradox by managing four potential process contradictions: conserving assumptions to revise assumptions, combining small wins and small losses, combining rich feedback and immunity to feedback, and synthesizing psychological safety and propensity to risk.

Paradox 3: conserving and revising assumptions
Existing assumptions are the first obstacle to remove when introducing a stretch goal. Given the element of novelty that stretch goals contain, organizations need to be open to questioning and revising prior assumptions. As Rousseau (1997) pointed out, stretch goals "motivate high performance by mandating creativity and assumption-breaking thinking" (p. 528). Organizational assumptions taking the form of dominant logics are powerful and resistant because they are not conscious but taken-for-granted and thus not subject to discussion and revision (Bettis and Prahalad, 1995). Stretch goals, because of their element of novelty, involve questioning existing logics. Only if the existing logic is made apparent and substituted by new assumptions can the target become possible. To facilitate these processes, organizations often translate ideas and benchmarks from adjacent fields. Southwest's use of the racing pit is a case in point (Vezmar, 1996). The company was able to translate knowledge performed with extreme precision and expertise in the pits in motor racing into routines for handling the refueling, cleaning, and provisioning of commercial airplanes.
Stretch goals involve using the past to build the future. Questioning of assumptions is critical for stretch but needs to be combined not only with the perception that new organizational states are possible but also with the conviction that positive path dependencies are conserved (Sitkin et al., 2011). Positive psychological capital on the members' part (Youssef and Luthans, 2007), conceived as synergistic combinations of hope, efficacy, resilience, and optimism, seems crucial to the operation of stretch goals. Hope, fostered by stretch goals, offers a positive sense of possibility and nourishes perseverance. Efficacy or self-confidence is also critical: low self-efficacy leads to the rejection of stretched goals, while high self-efficacy leads to accepting excessively challenging goals. As the goal-setting literature suggests, if goals are to motivate, they have to be seen as acceptable (Thompson et al., 1997). Resilience, the ability to adapt to adversity and flourish in face of challenging demands (Kossek and Perrigino, 2016), is important because, given a high level of difficulty, setbacks are likely to happen. When setbacks happen, the capacity to recover is critical. Finally, optimism is fundamental to preserve a positive orientation toward goals.
High levels of autonomy and empowerment are also necessary. To engage with stretch targets, individuals or teams have to perceive that they have the resources to redesign existing operations and the capabilities to use them by grasping the assumptions upon which they rest (Kahn, 1990). Thompson et al. (1997) explained that a high degree of control over resources and the capacity to introduce change and structural accommodation (i.e. the capacity to modify structures, policies, and practices), as well as protection from bureaucratic inertia, all play a role in the process. In the absence of these conditions, accepting stretch goals will be riskier. To the extent that they facilitate exploratory learning, the above elements are important (March, 1991;Sitkin et al., 2011). Stretch goals that are imposed by management will be best achieved by an organization that practices participation rather than management by fiat. Paradoxically, prevailing assumptions of participation and deliberation must coexist with the shared meta-assumption that these assumptions are dynamic and can be revised authoritatively, without consent.

Paradox 4: small wins and small losses
It is conceptually predictable that low expectations, if not matched by a clear sense of progress, may end up confirming initial expectations that a stretch goal is indeed impossible and lead individuals to desist. To counter the negative effect of initially low expectations, small wins can sustain a sense of progress toward a goal (Kotter, 1996;Weick, 1984). Small wins support momentum and can be used to build self-efficacy, which in turn can serve to sustain virtuous circles and facilitate a positive efficacy-performance spiral (Lindsley et al., 1995). Paradoxically, however, they can support the status quo, thus reducing the willingness to respond to stretch.
Small losses resulting from failed experiments fuel the progress of exploration (e.g. Birkinshaw and Haas, 2016) when they are not perceived as threatening but experiential-but catch 22 is that stretch goals are potentially perceived as threatening rather than experiential. The consideration of losses, their causes, and their implications may provide organizations with "humbling evidence", neutralizing the pitfalls of overconfidence, hubris, and conceit (Van Zant and Moore, 2013). Owens and Hekman (2016) suggest that collective humility fosters team and organizational performance. Protecting the organization from overconfidence is especially important because unreasonable confidence symbolically projected by the adoption of stretch goals (Bjaalid et al., 2015) is not only nurtured by overconfident, unrealistic leaders but can also be amplified by them as they become entrapped in the vanity and pathology of their projects (Carr, 2015;Stone, 2013;Vance, 2015).
In adopting stretch goals, organizations with cultures that offer employees rich opportunities to celebrate wins and to learn from losses will do better. But the effects of wins and losses contradict each other. Organizations need to avoid being caught in both overoptimistic interpretations of victories and in over-pessimistic interpretations of failures. The difficulty rests in the fact that stretching leaders are not necessarily prepared to receive bad news. Consequently, the necessary degree of psychological safety and voice may progressively lead to a narrative of grand illusions with limited adherence to reality. Hence, we propose that stretch goals will be more effectively adopted when an organization combines small wins and small losses.

Paradox 5: rich feedback and immunity to feedback
Organizations often get trapped in routines that impede learning. Argyris (1993) dubbed this mode of thinking, "Model I." Model I organizations repeat their existing processes with variations that do not favor exploration, exposing themselves to the risks of progressive, increasingly exploitative change. They keep on doing what they do even as these repertoires of action cease to be appropriate scripts in a changing world. Proficiency is no guarantee. When change becomes evident, the same organizations that stuck to their scripts and failed to learn will often adopt a pattern of radical change initiatives, increasing their exposure to risk. Consequential learning demands both stability and change (Farjoun, 2010).
The role of feedback in goal pursuit is well established, given the importance of experimenting, getting feedback, and experimenting again. The challenge is in filtering out "noise." Receptivity to feedback can lead to drift and reactive inconsistency, whereas impermeability to feedback can lead to disconnect with the environment. Accomplishing stretch goals requires the judgment both to filter and dismiss feedback. Managing stretch goals requires heightened sensitivity to rich feedback. Clarity of purpose is required to facilitate decision-making and to separate feedback from noise. Excess of openness to feedback may be confusing rather than illuminating. Hence, stretch goals will be more effective when an organization is able to distinguish the feedback that should be considered and the feedback that should be ignored, via rich conversations that reduce ambiguity (Daft and Weick, 1984;Dutton, 2003), supported by psychological safety, as discussed next.

Paradox 6: a propensity to risk in psychologically safe environments
The literature on stretch goals argues for the necessity of "safefailing" (Thompson et al., 1997). If failure is punished, people will only accept goals as legitimate that they know are possible to achieve or they will cheat because they fear acknowledging that they are unable to reach the goals, as in the case of Volkswagen (Ewing, 2015). Stretching goals involve pursuing risky procedures and outcomes, redesigning organizational operations, acquiring new knowledge. Psychological safety can be defined as the "perceptions of the consequences of taking interpersonal risks in a particular context such as a workplace" (Edmondson and Lei, 2014: 24). A measure of psychological safety, the employee's "sense of being able to show and employ one's self without fear of negative consequences to self-image, status, or career" (Kahn, 1990: 708), is favorable for learning processes. Paradoxically, without psychological safety, it is unlikely that individuals and collectives will pursue "unsafe" processes and outcomes. In contrast, a context characterized by strong psychological safety makes individuals more willing to adopt risky behaviors (e.g. searching and giving advice, assuming ignorance and asking for advice, assuming and sharing mistakes, and thus helping others to avoid the same mistakes and failures) that are indicative of more effective pathways toward stretch goals. Hence, we propose that stretch goals will be more effective when they combine a propensity to risk (as a facilitator of exploratory learning) with a sense of psychological safety (which renders failures and honest errors acceptable). In summary, stretch goals are difficult to manage results from the paradoxical challenges raised as they unfold. Maintaining a team or organization focused on stretch goals is an exercise in the handling of contradictions, which makes the process of stretching managerially difficult to handle.

Paradoxical outcomes
Organizations adopting stretch goals need to avoid the dangers of over-optimism resulting from success and of decreased collective confidence resulting from failure (e.g. Heynoski and Quinn, 2012). Stretch goal pursuit requires feedback and feedforward, learning and forgetting, as well as celebrating and cultivating pride and humility. Interactions between combinatory demands render the outcomes of paradoxical processes highly unpredictable.

Paradox 7: feedback and feedforward
Once a deadline for a stretch goal is met, organizations that use the opportunity to extract feedback from the experience in order to "raise questions that should be asked on the next project" (Catmull, 2014: 217) will continue stretching. By looking at the completed process as a source of feedforward furthering future process improvement, organizations judge practices that should be retained as well as those that should be discarded or changed. Feedback can thus be complemented or preceded by feedforward (Kluger and Nir, 2010). Anticipatory effects derived from completed processes can be retained to inform and enrich new projects. The use of feedforward deflates the importance of the past and helps to balance a focus on the past with consideration of future possibilities (Kluger and Nir, 2010).
The incorporation of feedforward in organizational processes has been defended (Kluger and Nir, 2010), but its organizational relevance is still marginal when compared to feedback. Feedforward is the use of information regarding future positive possibility, in contrast to what has been concretized in the past. Feedforward helps articulate the past and the future, to gain endurance and resilience. Instead of dealing with failure as individuals, the combination of feedforward and feedback involves collective sensemaking, a perception of social support that is crucial for the cultivation of resilience. The combination of feedback and feedforward may endow organizations with the sense of temporality that is crucial in facilitating a process understanding of events (Deroy and Clegg, 2015). Combined use of feedback and feedforward can sensitize organizational members to the fact that successes and failures are merely episodic manifestations of ongoing processes. Introduction of feedforward controls (Fowler, 1999;Koontz and Bradspies, 1972;Kreitner, 1982) can constitute an important but neglected dimension for managing stretch goals. Hence, we propose that stretch goals will be more effective when the organization combines feedback and feedforward. But this guarantees nothing: managers can use feedback selectively to materialize their intentions.

Paradox 8: learning and forgetting
One of the functions of the combined use of feedback and feedforward is to articulate processes of organizational learning and forgetting. Forgetting (both one's and others' mistakes) is important to learn, and learning is important to forgetting at both the individual and collective levels (Abramson and Senyshyn, 2010). The bias in favor of learning is evident in the literature and the importance of learning is not contested here, but in order to learn, organizations have to manage forgetting, namely, forgetting old habits (routines) that are no longer positive or productive as well as individual and collective mistakes and failures made honestly. Limited attention has been paid to the process of organizational forgetting (for an exception, see De Holan and Phillips, 2004).
The use of stretch goal setting weaves learning and forgetting together. It is as important to learn new practices under revised assumptions so as to forget old practices articulated under previously developed assumptions (see paradox 3). The paradoxical need to combine learning and forgetting indicates the necessity of dedicating explicit, formal attention to the processes of forgetting rather than leaving them to chance. Discarding old knowledge may be as tough as acquiring new knowledge. Interestingly, it seems easier to assume that forgetting happens in an effortless, natural way. It does not: old habits die hard and invalid routines, labels, and interpretations may prove extremely robust and difficult to counter (Cunha and Tsoukas, 2015;Masuch, 1985;Rosenhan, 1973). Organizations can manage forgetting by uncovering and exposing assumptions, bringing outsiders in, using devil's advocacy, learning from newcomers or from other non-traditional tactics (Sutton, 2002). Hence, the pursuit of stretch goals will engage the organization systematically in the managing of learning and forgetting.

Paradox 9: celebrating pride and humility
When organizations stabilize an interpretive pattern among their members that expresses mastery in a given activity, they can become victims of overconfidence (Lindsley et al., 1995). Overconfidence can be problematic in the process of managing stretch for a number of reasons. First, it can create the illusion that an organization is so competent that reaching the impossible is normal. Hubristic expressions of overconfidence are not uncommon (Hayward et al., 2006). Second, success can lead to simplification (Miller, 1993), and the more success is celebrated, normalized, and internalized, the more the organization will render itself vulnerable to the Icarus paradox (Miller, 1990), that is, to the assumption that there are no limits given the organization's proven competence. As Andrew S. Grove, cofounder of Intel Corporation, suggests, "Success breeds complacency. Complacency breeds failure. Only the paranoid survive" (in Baumol et al., 2007: 228).
Organizations should not only celebrate their achievements-including their small wins, as explained above-but they should also simultaneously display humility (Owens et al., 2012;Vera and Rodriguez-Lopez, 2004) to temper celebration. In other words, they can cultivate "humbition" (see Taylor, 2011). In this vein, an attitude of wisdom means that an organization treats success as carefully as it treats failure: both constitute occasions to learn (Cannon and Edmondson, 2004;Gino and Pisano, 2011).
Accepting that success and failure coexist, that both constitute paths for learning and that they should be approached with emotional equanimity, can aid organizations in being better prepared to deal with the challenges of stretching goals. Organizations seeking to embrace a "mission impossible" need to develop an emotional capability that balances an emotional response to demanding targets characterized by a creed of no euphoria in victory-no defeatism in failure. Readiness to adopt stretch goals is thus a matter not only of ambition and functional capability but also of members' emotional readiness and maturity (Huy, 1999). Existing research indicates that change processes involve significant emotional work and that those involving extremely demanding goals may be proportionately more demanding in terms of emotional capability (Vuori and Huy, 2015). Hence, stretch goals will be used more effectively when the organization celebrates success but simultaneously cultivates an attitude of doubt with regard to success.
In summary, it is critical to assume that the achievement of a stretch goal is not only the end of a process but also the continuity of organizational processes. Learning about success and failure and sustaining momentum are critical to inform the future. Finished projects constitute opportunities for learning from the past as well for projecting the future. The articulation of the connections occurring across the described paradoxes is depicted in Figure 1. Each paradox finds its foundation in a tension (represented by the dotted line) and is logically connected to other elements related to stretch goals (continuous lines). As described so far, stretch goals generate paradoxes that appear in different phases (triggers, process, outcomes) and articulate at different organizational levels (individuals, teams, organizations). Some of the interweaving of mutual connections is depicted in Figure 1.
Implications. The attractiveness of stretch goals to leaders and managers anxious for fortune and fame is obvious. In terms of ethics, stretching can be an excuse for the overambitious to justify their claims. The implications are consequential, especially if stretch is perceived as associated with the idea of winning at all cost. In this case, stretch goals can constitute a direct path to unethical achievement. It is important to explore the implications of stretch objectives over the triggering of organizational chains of reaction. As Sitkin et al. (2011) pointed out, stretch goal setting may be adequate only to a very limited set of organizations, being more appealing to those organizations that unfortunately are not well prepared in terms of slack and recent performance. Stretch triggered by despair or an unbalanced ambition (i.e. not by "humbitious ambition") can be destructive (Ordóñez et al., 2009).
We continued the discussion initiated by Sitkin et al. (2011) by showing that paradoxes are not only present in stretch but that they are also potentially nested. The recourse to stretch can open possibilities not envisioned by the organization or its managers as it is riddled with paradoxes that will potentially interact with other paradoxes impossible to pre-specify. We thus opened a conceptual window on cross-paradox articulation, a topic that is still under-explored, which may help  explain why organizational processes sometimes interact at levels of complexity that escape understanding and why managerial actions sometimes transform positive intentions into vicious circles difficult to make sense with as well as to break (Masuch, 1985). As Tsoukas and Cunha (in press) pointed out, To keep an organization in a state of equilibrium managers have to sustain negative feedback loops. The problem is that because referents are arbitrary a vicious circle can, up to a point, be perceived as virtuous such as when an organization is growing at very high rates. In this sense one circle may be perceived, paradoxically, as either virtuous or vicious depending on the selected referents, until the system breaks down.
Stretch can be an extreme example: positive or negative, depending on the eye of the beholder.

Conclusion
The recent surge of interest in stretch goals may have led some managers and researchers implicitly to assume the superiority of these goals to traditional approaches. Previous research by Sitkin et al. (2011) clearly specified that stretch goals are productive only within a very narrow combination of organizational circumstances. We have added detailed consideration of paradox to the discussion of stretch goals: because stretch goals trigger paradoxical requirements, they are hard to manage. The fact that extreme goal adoption triggers nested paradoxes, that is, paradoxes that trigger other paradoxes previously dormant at different levels, renders the process especially demanding from a managerial viewpoint. The article thus contributes to the literature by uncovering the paradoxes that precede what Sitkin et al. (2011) labeled as the paradox of stretch goals and by paving the way for the empirical analysis of nested paradoxes in stretch goal setting. If traditional approaches entail stating targets and rewarding their achievement in a form of governmentality that asks no more of the subject than to perform at a constant level, then asking the subject to achieve a constantly escalating level of performance in excess of the goals set breeds a certain paradoxical subjectivity as we have discussed. As a form of governmentality producing eager neurotics-eager to achieve but constitutively anxious about doing so-stretching goals may achieve remarkable results in the short terms but in the longer terms lead to nested paradoxes as we have outlined. Our work suggests, in conclusion, that organizations need to represent stretch goals in the here and now not only as a radical tool and a managerial temptation but also as opening a radical and potentially dangerous Pandora's box of paradox for the future.