SSI Thought Letter: Vol. 1 Issue 4
Expanding The Safety Ecosystem Model
It’s More Than A Notion
“Safety” as an ecosystem is a pervasive concept in the physical realm of human activity. In the literature, it is primarily applied to protecting human life, health, and property. The universe of technical discussions and components of safety at tactical levels reflects a wide array of midrange approaches to safety in various disciplines. In contrast, the sheer number and diversity of safety ecosystem depictions and models reveal a rich and robustly complex topic.
Yet, safety literature and discourse still have a ‘missing link’.
In recent decades, the term of art, “Psychological Safety,” emerged, then stuck the landing in the 1980s workplace when teams became the sweet spot domain. Small teams came into vogue as a preferred technology for business problem-solving, for reasons not well known, but timely. The entrenched art of Psychological Safety settled into a sense of comfort defined as “…being able to show and employ oneself without fear of negative consequences of self-image, status or career.” [end note 1] In psychologically safe teams, members are said to feel sufficiently accepted and respected such that they share the belief that the team is safe for interpersonal risk-taking. [end note 2]
The operative terms in those definitions are “feel” and “belief.” One of the most effective local legislators I’ve met in my career once said, “When anyone comes before this [legislative] body and starts their argument with, ‘I believe’ or ‘I feel,’ the presentation is over at that point. We need evidence to make decisions, not perceptions.” He had come to understand that in the end, feelings and beliefs exist largely in the mind (though the gut plays a crucial interoceptive role, too), and they can be and often are dead wrong.
Most can readily relate to the ease with which a visual perception of a dress or an auditory perception of a name can present as alternate realities for sane and rational human beings. Feelings and beliefs are subjective dependent variables whose value depends on that of another. Finding the other variable(s) in a seeming conflict is often crucial to arriving at reality, rather than relying on a winner-takes-all brawl between dueling perceptions.
In the case of the Laurel vs. Yanny debate, the moderator variable (or moderating variable) was revealed, and the discovery made it possible for listeners to hear both, reconciling perceptions with realities. That sweet spot represents a point of equilibrium, a coordinate that creates a neutral, impartial moment in aural space that compensates for the natural bias (valence) in our spectral hearing range. We can “hear” what is otherwise dichotomous.
That kind of moderated balancing relates to the Venn-like structure, as we recognize that the dynamic equilibrium insinuated by Lewin’s Life Space model insinuates a form of the law of conservation, a naturally closed system, but one that’s structurally unique to each individual. We operate within an “umwelt,” a natural range of abilities that allow us to sense our environment physically and perceive it cognitively. The examples of our limits are abundant, from our inability to hear a dog whistle to our natural inability to perceive a fourth dimension as it’s graphically represented in a tesseract. Therefore, we have tradeoffs, energy exchanges, and sacrifices of our mental, physical, and social capacities and faculties with each other in a zero-sum way. When we reach limits in one, we naturally attempt to compensate with others, both voluntarily and autonomically.
Thus, that realization in the 1980s spurred management theory around creating a perception of “psychological safety” in teams where work and personal stresses and strains posed a glaring conflict with the necessities of work-life performance for many. That dynamic generated research into the decision-making small team strata juxtaposed against the backdrop of a post-Civil Rights Era “diversity paradigm.” One challenge: varying cohorts couldn’t reconcile succumbing to a [false] sense of total safety in teams comprised of other organizational members affected by broader, lingering social realities of the day. It was, after all, in the 1980s when the American political strategist Lee Atwater crystallized a longstanding debate simply and succinctly, with the now axiomatic assertion that “perception is reality”.
A moderating organizational variable was missing. In keeping with that, one major theme of sociology over the past half-century has been helping people differentiate and reconcile their personal circumstances and experiences with realities consistent with those of their societal cohort at a macro level. American sociologist C. Wright Mills (1959) referred to the point of reckoning as “the intersection of biography and history,” where personal and private matters connect or correlate to broader social-historical conditions.
In short, it’s the scientifically neutral and objective assessment of the relationship between self and society. That same kind of assessment can be performed in mini or micro societies, like organizations, as well as at the macro societal level. It is the rough equivalent in nature to individuals’ sensitivities, their abilities to detect a specific range of frequencies of light or sound that reconcile the visual perception of dress colors and the auditory perceptions of names.
Edmonson (2010), Kahn (1990), Schein, Bennis (1965), Maslow (1950), and others embedded implications in their work regarding an individual’s feelings/beliefs congruent with safety (risk tolerance). What psychological safety meant for organizations from Maslow on, though, was motivating (i.e., convincing) employees to make a leap of faith off the socioemotional ledge of interaction risk. Anyone who has worked more than one eight-hour shift in a business organization knows they’re not entirely safe. The question central to psychological safety was how to manage risk tolerance/aversion to produce optimum team results.
Workers are routinely manipulated into believing that the workplace is safe enough to make the same kind of psycho-emotional leap. The Schein-Bennis work of the mid-60s preceded Kahn’s 1990 release. Even Maslow’s higher third or so of the pyramid was really about psychological safety. All of them, including Edmondson, did exactly what complexity theorist David Snowden warns against. They treated an ecosystem issue as if it were an engineering problem. Maslow came closest to defining a safety ecosystem, despite the later unfortunate connection to the hierarchical structure. However, he only obliquely implied the sociological “third leg” variable of the safety stool. At its least complex applications, the robust measure of safety encompasses six specific aspects, which we’ll share in upcoming installments.
Meanwhile, most would likely agree that employees should only believe/feel a psychological sense of safety when they are, in fact, safe in every respect to a reasonable degree. Psychologists have long recognized that there may be circumstances in which the sense of safety has no connection to reality. They’ve got terms to describe the incongruence between reality and belief, termed the “miscalibration of subjective probabilities.”
- Paranoia = Feeling unsafe when we’re probably safe.
- Overconfidence = Feeling safe when we’re not.
- Irrationality or Phobia = Feeling unsafe or safe, inconsistent with the evidence.
- Rationality = Assessment of safety that is consistent (accurately calibrated) with realities.
- Deception = Planted false or misleading evidence to give the appearance of safety despite the realities (facts). (Sometimes termed “impression management.”)
In a study of citizen confidence levels in police performance in encounters with police, the psychological safety levels were quite different when citizens’ race was taken into account. In the Pew Center Survey, “Only about a third of blacks but roughly three-quarters of whites say police in their communities do an excellent or good job in using the appropriate force on suspects, treating all racial and ethnic minorities equally, and holding officers accountable when misconduct occurs.” There is a clear and statistically significant race-based difference in feelings and beliefs that constitute a psychological sense of safety.
The sociological perspective on safety would have to be based on more than surveys of beliefs and feelings. As an example, an NAACP Criminal Justice Fact Sheet provides some insight into the difference between a psychological and sociological perspective, listing the following, among other things:
- A Black person in the USA is five times more likely to be stopped without cause than a white person.
- Police killings of unarmed Black Americans are responsible for more than 50 million additional days of poor mental health per year among Black Americans. This mental health burden is comparable to that associated with diabetes, a disease that strikes 1 in 5 Black Americans.
- 5% of illicit drug users in the USA are African American, yet African Americans represent 29% of those arrested and 33% of those incarcerated for drug offenses.
While individual persons who are members of a society may believe or feel a certain way about their relative safety, the sociological analysis of cohort circumstances can reveal the underlying (or over-arching) safety story more convincingly and more unambiguously than can the psycho-emotional story we may tell ourselves and others, pro or con.
The fundamental measurement can be the same whether an individual, a group, or a cohort. The conceptual meta-safety formula might start with a simple additive safety level resulting from the three domains. If all three are perceived as relatively safe (risk<0), the sum will be a positive number. When the perceived level is unsafe (risk>0), then to the extent that is true for one or both, the sum could be negative. That’s simplistic bottom-line math of a compound reality.
However, even with all its multifaceted possibilities, layered complexities, relative weighting, and the distinction between tangible and intangible aspects, as well as moderating influences between the variables of the three domains, a structured equation model, and more, it still yields the same fundamental result from the calculation. (Liminal spaces or overlap areas and nexus in these tripartite Venn models will be addressed in upcoming posts)
Any sense of perceived psychological safety, expressed as a feeling or belief, will depend on the measurable confidence levels in the tangible instrumentalities for creating both perceived and real safety in the other two domains. In short, evidence. The measurable performance of tangible, credible, physical, and social safety protections and precautions in place (environment, processes, procedures, infrastructures, etc.) would constitute the physiological and sociological safety variables. The net of the three domains will produce a safety measure (score) for each domain, and for the safety ecosystem as a whole. The objective would be a more robust analytical frame for understanding and developing ways to make safety ecosystem outcomes for each cohort a more neutral proposition, now including the discrete Sociological Safety domain objective of viable social existence, both individually and relationally.
Thus, with police performance (behaviors and outcomes), we can see the robust ecosystem of physiological, sociological, and psychological metrics differentiated by actual cohort (race, in this case) experience with citizen safety in policing. Sociologists would readily assert the correlation between perceived psychological safety and objective measures of sociological safety. (“Thank God, the police are here,” vs “Oh God, the police are here.”)
So, from that three-legged stool perspective, what can organizations use this for? It’s not hard to make the leap to applying sociological safety considerations in organizations, such as businesses, business sectors, governmental agencies, and even countries. We’ll shift back to business organizations first.
Work teams became the orbital center of the psychological safety universe for reasons discussed in another installment. However, teams aren’t the only sociological unit in organizations or societies by a long shot. Workgroups, departments/functions, cultures, affinity groups, and the enterprise itself all become sociological units internal to the organization. Sociologists tend to study similar themes about large group behaviors, as psychiatrists and psychologists do about individual and small group behaviors, with scale being one differentiator.
Invariably, discussions of establishing and sustaining psychological safety among team members, as well as the methodological approaches to doing so, rely on two risk calculations centered on safety.
a) Members are at low risk of physical harm from environmental hazards, and
b) Members are at low risk of real or perceived mistreatment and/or injury (physical or emotional) from other members of their sociological unit.
Larger groups are naturally subject to the same two risk calculations by their members, regardless of scale, from classrooms to entire countries.
The historical development of Labor Unions, for example, exemplifies the formation of internal sociological (transactional) subgroups in response to the safety needs of their members. More recently, the development of “Employee Resource Groups” (ERGs) as an interactive subgroup is getting a second look (i.e., the side-eye). Originally formed to foster a sense of organizational group safety via affinity, some are realizing ERGs may have fallen well short of expectations in their attempts to provide the kinds of psychological safety necessary for a sense of full inclusion, or a strong sense of belongingness in the organizations or the subgroups that held out that promise of safety/belongingness.
The risk calculation can be intuitive, performed without much conscious thought, and in response to real or perceived threats. It can employ the mind, as well as the heart and gut, as sensing instruments. There is a simple formula for it, which we’ll consider here.
Psychological safety metrics are descriptive, and the prescriptive measures are taken with the intent to alter member beliefs and feelings, their “sense” of safety.
The risk calculation can be illustrated with a longstanding but inexact judgment and decision-making heuristic from 1930s law. Liability was boiled down to a test called the “standard of care,” which is still used in liability cases. Again, it’s not math, and there’s no standard calculation. In law, it’s always an after-the-fact assessment by a judge or jury for an award. It’s a multifactorial conceptual formula that can generate huge swings in the value of court settlements.
For individuals and groups, however, it is a neat (i.e., useful, practical) way of conceptualizing a more holistic algorithm for assessment of both real and perceived safety, rather than a litmus test. It is the same calculated way in which psychologists attempt to “reverse engineer” psychological safety in teams, by assessing the component parts after the fact, then attempting to engineer more functional group norms for a specific purpose. It reflects the basic components that Occupational Safety professionals deal with in risk mitigation. In a very real way, it can be a handy, valuable tool for quick situation analysis in occupational safety, health, and even in personal prospective risk mitigation (protective measures and actions) in a wide range of social (group) settings.
Without a doubt, constructing a belief or feeling of psychological safety can have a positive short-term effect, boosting quality of decision-making, productivity, and relational stability for individuals, teams, or groups — even outgroups. That will only last as long and remain as strong as an organization can maintain employee confidence in the construct, and sustain the appearance of reliability and stability of the safety precautions and protections in place. As a result, the “reverse engineering” approach of psychological safety in teams requires some very special technologies and techniques, as the baseline is in constant modulation.
David Snowden drew an important analogy for governance and management. He pointed out that they are “much simpler when you understand the nature of the system, and you stop trying to treat an ecosystem as if it were an engineering problem when it's an ecological problem.”
Characterizing safety as an ecosystem, a complex network of relations, or an interconnected system of risk management is more than a notion. When taken as robust enterprise-level management of risk, mitigation, and contingencies in total over the three domains of physical, psychological, and sociological safety, the nexus represents what Snowden would call a complex adaptive system in which everything is connected to everything else, and in ways that we often cannot foresee or even see at all until it is too late. Unintended, unknown, and unknowable consequences can abound in ecosystems. Because of that, retrospect becomes safety management’s primary lens, the objective being, “Let’s work to prevent that from ever happening here — again.”
The extrusion of a “sociological safety” domain as a third leg of the safety stool makes it possible to treat safety as a robust ecosystem. It opens the domain of psychological safety to a more broadly generalizable calculation of “belief and feeling” about interaction with both physical and human domain systems. It elevates the psychology of safety well beyond the tight constraint of small team interaction or the psychology of incident management, encompassing the social psychology of even large sociological groups, or very small ones.
We know that members of outgroups can and do encounter highly dysfunctional (abnormal) psychologies in organizations and their sociological sub-groupings, too often by design and intent. The sociological safety domain accomplishes psychological diagnostics through groupings.
The recognition of a “sociological safety” domain both frees and crystallizes the role of psychological safety as the study of beliefs and feelings in connection with relational experiences at any level of interaction, including individual perceptions about self, others, and the environment. By formalizing C. Wright Mills’ incorporation of reconciliation of personal circumstances and experiences with realities consistent with those of their societal cohort at a macro level in a well-defined sociological domain, it then opens the way to a clarified intent of aligning (reconciling) psychological beliefs and feelings (perceptions) with organizational norms. (That’s not a typical way of applying mesosociological control concepts, but the implication isn’t subtle.)
Said another way, there’s no rational reason that small teams should be the exclusive level of human interaction that produces a sense of “psychological safety,” even at the micro/quantum level of strong individual performers.
The prevailing definition of “microsociology” (1939, Georges Gurvitch) has been “small-scale sociological analysis that studies the behavior of people in face-to-face social interactions and small groups to understand what they do, say, and think.” In tandem, sociological micro theory focuses on “individuals and small groups and the interactions between them.” Social action theory perspectives focus on smaller groups and the subjective states of individuals and their interactions within them. Yet, by the end of the 20th Century, the “psychological safety” brand conscripted the group subset of team interaction. The result is psychologists attempting to artificially manipulate small group perceptions to achieve micro-behavioral goals, but in a sociological vacuum, with little sense of how those behaviors and perceptions alter sociological structures and systems.
Optimism, Pessimism, and Realism are the three components of the faculty we use to calculate probabilities and assess risk. It has elements of intuition (the ability to understand something immediately, without the need for conscious reasoning), and involves our capacity for hindsight, insight, and foresight.
We may each tend toward a particular orientation. However, we utilize this faculty to make situational calculations at varying levels of analysis. We tend to categorize the components as positive, negative, and neutral (objective to the extent possible). For example, “Optimism often keeps us moving forward beyond the point of no return.” That can cause unfavorable outcomes. Conversely, “Pessimism often prevents us from doing what should not be done.” That would cause favorable outcomes. An inherent valence shouldn’t be applied to any of the three components except, in retrospect, knowing the situational outcome(s) in context. The combinatorial relative proportions of the three can be a matter of fuzzy math. But it is a calculation nonetheless, one we make naturally in the course of any given day or event, particularly regarding our safety.
IN THE END, the measured psychological “belief and feeling” of safety should objectively correlate with the level of confidence and reliability (performance) of tangible, material, corporeal, observable safety activities, behaviors, conduct, processes, procedures, systems, and infrastructure design for establishing and sustaining safety in the other two domains. There’s also no objective way, other than inferential, to measure or observe “psychology.” That’s not the case with physiological safety or sociological safety, as interactional functionality (which means that teams are included in the full spectrum of sociological groups).
In contrast, infrastructure, processes, systems, and procedures for improving physiological and sociological safety can be designed, built, and objectively observed and performance-tested using psychology’s inferences and correlations as the basis, in addition to analog feedback from non-theoretical “hard data” results, such as accident statistics or manifest dysfunctional behaviors.
This addresses the understanding of how the three domains relate methodologically. Psychology provides the inferential framework and correlational insights, but the “real world” validation of beliefs and feelings comes from observable performance, competence, and reliability in the physical and social domains. Although safety systems can and do emanate from structure, they are not limited to macro-level structures, but also encompass the observable interpersonal and group dynamics that can be directly measured through communication patterns, decision-making processes, conflict resolution effectiveness, and collective performance outcomes.
A state of Sociological Safety is the design, implementation, and performance optimization of observable social systems, structures, and interactional processes that demonstrably produce safe outcomes across all levels of group organization, throughout the spectrum of social entities, from pairwise dyadic interactions to teams, work units, institutions, and societal systems.
Sociological Safety operates through tangible, measurable infrastructure consisting of policy, protocols, communication, decision-making procedures, accountability mechanisms, coordination systems, organizational hierarchies, regulatory frameworks, and collective behavioral patterns. As social technologies, these can be directly observed, performance-tested, and objectively evaluated based on measurable outcomes such as incident rates, response effectiveness, coordination quality, and collective safety integration.
Sociological Safety® encompasses both macro-level structural arrangements (how organizations distribute authority, resources, and risk) and micro-level interactional functionality (how members communicate, coordinate, resolve conflicts, and make safety-critical decisions). It recognizes that safety emerges from the reliable performance of guiding social systems, not just individual psychological states or even physical protections.
The organizational meso-level serves as the connective tissue between macro-level structures (such as institutional policies or societal norms) and micro-level interactions (like team dynamics or individual behavior). Sociological Safety® fits here as the operational layer where social systems are actively designed, maintained, and tested for reliability. The meso-level is where social systems are enacted, and Sociological Safety provides the blueprint for making those systems perform reliably, inclusively, and adaptively.
The Sociological Safety objective is to structure and manage social environments in which safe behavior is structurally supported, unsafe behavior is systematically prevented or corrected, and collective safety performance can be continuously monitored and improved. This involves designing organizational cultures, team processes, communication systems, and institutional arrangements that function reliably under stress, adapt to changing conditions, and maintain safety performance regardless of individual psychological variations.
Sociological Safety utilizes psychological insights as design inputs but validates effectiveness through objective measures, including accident statistics, compliance rates, team performance metrics, system reliability data, and observable behavioral outcomes. The focus is on building social infrastructure that works, not just analyzing why existing systems fail.
The criteria for evaluating whether organizational frameworks support Sociological Safety pivot on whether they:
- Map observable social systems
- Identify measurable infrastructure components
- Show multi-level integration pathways
- Account for performance under stress
- Enable objective validation
As a side note, there is more than a century of sociological theory to draw on for this. Sociological Safety® isn’t reinventing the wheel. It’s cleaning the windshield. The concept of safety as a complex ecosystem of risk management and mitigation doesn’t require unnatural carve-outs (cherry-picking) or theoretical and metaphorical contortions to distinguish sociological safety theory from action.
Importantly, to gather data specific to this idea of a sociological safety domain, there’s nothing intrinsically wrong with advancing theory and conducting experiments in organizations of any kind and at any level. Experimentation has remained a natural and necessary part of organizational diagnosis and problem-solving, standard practice. The diagnosis can and should induce action when compelling and appropriate.
At the very least, theorists and experimental researchers should have a firewall or neutral third-party validation between them and their implemented recommendations. This separation isn’t just an ethical practice. It’s especially critical for sociological safety because the infrastructure must be objectively reliable, not just inferentially elegant. The safety system’s integrity depends on the manifest quality of implementation, not the researcher’s satisfaction.
However, when theorists and experimental researchers (consultants) become responsible for implementing and managing their own recommendations on safety, researcher bias and worse can creep into those implementations in organizational development, even constituting a conflict of interest. The potential for conflict of interest can become part of the risk equation. We’ve seen it often enough in consulting.
Enough has been said, for now, about a more robust Safety Ecosystem comprised of all three domains of Physiological, Psychological, and Sociological Safety®, as well as its alignment with organizational norms. We’ll take a deeper dive into the emergent Sociological Safety® domain in the upcoming installment, Volume 1, Issue 5, on the Theoretical Grounding of Sociological Safety®.
_______________
Notes:
1. Kahn, William A. (1990–12–01). “Psychological Conditions of Personal Engagement and Disengagement at Work”. Academy of Management Journal. 33 (4): 692–724. doi:10.2307/256287. ISSN 0001–4273. JSTOR 256287.
2. Edmondson, Amy (1 June 1999). “Psychological Safety and Learning Behavior in Work Teams” (PDF). Administrative Science Quarterly. 44 (2): 350–383. doi:10.2307/2666999. JSTOR 2666999.
