The Pause Before the Answer: Psychological Safety as the Missing Condition Between Visionary Leadership and Genuine Teacher Voice

A Note to the Reader

This paper is about a moment that most teachers will have experienced. A question is asked and the answer that follows is not quite the one that was there. It is careful, appropriate, and stays on the surface. The fuller thought does not make it out. That moment is small and passes quickly. This paper argues that what it contains matters enormously.

1. Introduction

During the data collection for an earlier study on visionary leadership and team performance in educational institutions (Chin, 2024), one moment stayed with the researcher long after the interview had ended. A teacher paused before answering a question about leadership and shared vision. The pause was not long. It was enough. What followed was careful: the words measured, the response general, the answer shaped more by what was appropriate to say than by what the teacher might have genuinely thought. It was not what she said but what she did not say that lingered. It did not feel as though she had nothing to offer. It felt more like she did not feel safe to say it.

The 2024 study had been designed to examine how visionary leadership fosters shared vision and teamwork, and how these contribute to team performance in educational settings. Its findings were clear on what leaders did and what teachers experienced. Participation structures were present. The invitation to contribute was genuine. Leadership aimed, sincerely, for inclusivity and collaboration. What the study could not explain was why some teachers held back their views even when the opportunity to share them had been formally given. It could describe the space. It could not explain what determined whether a teacher felt able to inhabit it. This was not merely a limitation of one study. It pointed towards a question that the visionary leadership literature, for all its insight into what leaders do, had not yet asked with sufficient precision.

This was not a question of leadership. The leader had created the conditions. It was not a question of structure. The mechanisms for participation were in place. It was not, primarily, a question of teamwork. Colleagues were present and the culture was described as collaborative. The unresolved question sat elsewhere, in the distance between the opportunity that had been offered and the voice that had not fully arrived. What people say is only part of the data. What they hold back is also part of it. The 2024 study had no means of capturing the second part.

This paper argues that what was missing from that study, and from the visionary leadership literature more broadly, was not a structure, a strategy, or a leadership behaviour. It was a condition: the felt sense of safety that determines whether participation becomes a genuine voice, or remains, as this paper's title suggests, present but withheld. That condition is psychological safety (Edmondson, 1999), and its absence from the visionary leadership literature is the gap this paper seeks to address. Specifically, this paper proposes that psychological safety is the missing condition between the participation structures that visionary leaders create and the genuine teacher voice that those structures are designed to release.

The argument proceeds in six sections. Section 2 examines what the visionary and transformational leadership literature has assumed about the relationship between participation and voice, and where that assumption fails. Section 3 introduces psychological safety as the construct that was present, though unnamed, in the 2024 findings. Section 4 draws on the literature on employee voice and silence to argue that withholding is not absence but choice, made in the presence of perceived risk. Section 5 proposes the conceptual model that places psychological safety between shared vision and genuine teacher voice. Section 6 extends the argument to the conditions of the research encounter itself, where the same human assessment applies.  The paper concludes by setting out the implications for school leadership practice and the empirical questions that the conceptual model generates for future research.

2. What the Literature Assumed

Transformational and visionary leadership research has, for several decades, offered a compelling and internally coherent account of how leaders produce genuine engagement among their teams. Bass (1985) established the foundational premise that transformational leaders inspire followers, communicate a compelling vision, and lift individuals beyond their immediate self-interest toward collective goals. The process is motivational at its core. When a leader articulates a direction clearly and sincerely, the assumption is that followers will orient themselves towards it. Sashkin and Sashkin (2003), extending this framework, argued that visionary leaders do not merely state a vision. They model it, embody it, and foster the conditions under which others commit to it. Taylor, Cornelius and Colvin (2014) reinforced the same logic: that effective leaders define purpose, align members, and unify effort toward shared goals. The language of alignment, commitment, and unity runs through this literature like a current. Strong leadership, clear vision, genuine inclusion: these three elements, in combination, were understood to produce engagement.

It is a persuasive account. It is also, in important ways, incomplete.

The incompleteness is not in what these scholars argued. Each made a defensible and well-evidenced claim within the terms of their own frameworks. The incompleteness lies in what the framework did not ask. Transformational and visionary leadership research focused its attention on what leaders do. It examined how vision is communicated, how inclusion is structured, and how participation is invited. What it did not examine, at least not systematically, was what happens inside the person who receives that invitation. Specifically, it did not ask whether the teacher sitting across from a leader who has shared a vision sincerely and with genuine intent feels safe enough to respond to that vision fully and without reservation.

This limitation is one of application rather than original intent. Bass (1985), Sashkin and Sashkin (2003), and Taylor, Cornelius and Colvin (2014) were not arguing that participation alone guarantees voice. The field, in applying their frameworks, quietly assumed it did. This gap has not gone entirely unnoticed within the transformational leadership tradition itself. Scholars working on teacher emotions and the internal conditions of followers have pointed towards this dimension as an important but underexplored aspect of leadership effectiveness, though it has yet to be connected systematically to the question of psychological safety.

The assumption embedded in the literature was this: if you invite people to speak, they will speak. The invitation was treated as sufficient. The willingness to respond was treated as a natural consequence of the quality of the leadership behaviour that preceded it. This study's 2024 findings did not contradict that assumption directly. The leaders in that study were inclusive. The structures for participation were present. The invitation to contribute was genuine. And yet something in the data resisted the clean sequence the literature had predicted. Some teachers contributed fully. Others offered answers that stayed on the surface. A few reduced their participation over time, not because the invitation was withdrawn, but for reasons the study's design could not reach.

The missing question did not become visible during the study. It became visible afterward, in the quiet work of sitting with what the data could not explain. Leadership research had asked one question: Does the leader share the vision? It had not asked the question that sits just beneath it: does the teacher feel safe enough to respond to that vision fully and without reservation? Bass's (1985) own framework included individualised consideration, which attends to followers' individual needs and development. Yet individualised consideration addresses growth and support rather than the specific interpersonal safety assessment that precedes voice, and it does not resolve the question of whether a teacher feels safe enough to speak freely in a given moment. The question of whether the leader shares the vision and the question of whether the teacher feels safe enough to respond are not the same question. That conflation is where the framework assumes more than the evidence supports.

Two recent empirical studies encountered the same gap from different directions. Awodiji and Oluwalola (2025), examining distributed leadership and school goal achievement in Nigeria, found that inspiring a shared vision did not significantly predict outcomes, a result that sits uncomfortably alongside the confidence of the transformational leadership tradition. The authors did not dismiss the finding. They absorbed it and noted that whilst distributed leadership relies heavily on vision-sharing, the relationship between vision and achievement is not always direct. Di-at and Abaya (2026), studying school leadership practices and teacher organisational commitment in the Philippines, reached a related conclusion by a different route. Leadership practices across six domains, including vision, mission and goals, rated very highly among teachers, yet the strongest effects were on affective and normative commitment rather than on active voice or substantive engagement. Teachers felt attached. They did not necessarily speak freely. Di-at and Abaya (2026) did not examine voice directly. Their study was designed to measure commitment, not participation. But the pattern in their data points towards exactly the gap this paper seeks to explain: that structural attachment and genuine voice are not the same condition, and that leadership practices, however strong, do not automatically produce one from the other.

What neither study could explain was why. The present paper proposes that the explanation lies not in the quality of the vision, nor in the sincerity of its communication, but in the presence or absence of the condition that determines whether teachers engage with a shared vision genuinely or perform engagement carefully. That condition is examined in the section that follows.

3. Psychological Safety: Recognition, Not Discovery

When the researcher returned to the literature after collecting the 2024 data, the encounter with Edmondson's work did not feel like discovering something new. It felt like recognition. The moment with the teacher who had paused, who had given a careful and general answer where something fuller had seemed possible, had remained unexplained. Not wrong. Not dishonest. Simply held back. Reading Edmondson's account of how people assess interpersonal risk before deciding whether to speak, the response was less one of surprise and more one of quiet arrival. This, it turned out, was what that had been.

Edmondson (1999) defined psychological safety as a shared belief held by members of a team that the team is safe for interpersonal risk-taking. The definition is precise in what it includes and, equally, in what it does not. Psychological safety is not confidence. It is not personality. It is not the absence of disagreement or the presence of warmth. It is a shared belief, held collectively, about the consequences of speaking. When that belief is present, people do not need to weigh whether their words will be taken well before offering them. When it is absent, that weighing begins. And once begun, it tends towards caution. Psychological safety is not the same as trust (Edmondson and Lei, 2014). Trust is an individual's assessment of another person's intentions, whilst psychological safety is a shared group-level belief about the consequences of speaking (Edmondson and Lei, 2014). The two are related but not identical, and conflating them risks obscuring what makes psychological safety distinctive as a condition of voice.

In plain terms, psychological safety is not about being allowed to speak. It is about not needing to hold yourself back when you do. That distinction matters because the leadership literature examined in Section 2 had attended carefully to the first condition and overlooked the second entirely. Participation structures create permission. They do not create the felt sense of safety that permission requires in order to become genuine voice. A teacher can be formally invited to contribute to a shared vision, can sit where the conversation is happening, can even add her name to the list of those who attended, and still hold back the thought that matters most. Confidentiality was assured. The pause remained.

Edmondson and Lei (2014) traced this construct across five decades of organisational research, showing that psychological safety operates at the individual, group, and organisational levels simultaneously. At the group level, which is most directly relevant to the present argument, psychological safety describes the climate within which team members decide, moment by moment, whether speaking is worth the risk. Frazier, Fainshmidt, Klinger, Pezeshkan and Vracheva (2017), drawing on 136 independent samples representing over 22,000 individuals and nearly 5,000 groups, confirmed that psychological safety is among the most robust predictors of voice behaviour, learning, and performance in team settings, and that its relationship to performance operates primarily through voice rather than as a direct effect. The empirical weight behind the construct is, by now, considerable. A recent evidence review confirmed that psychological safety remains one of the most actively researched constructs in organisational behaviour, with implications extending across leadership, voice, learning, and performance (Capezio, Barends, Rousseau and Wietrak, 2023).

What is striking, however, is that this weight has not been brought to bear on the visionary leadership literature in any sustained way. The two bodies of scholarship have developed largely in parallel. Leadership research has examined what leaders do. Psychological safety research has examined the conditions under which people respond. The connection between them, the question of whether the conditions a leader creates actually produce the felt safety that genuine voice requires, has remained largely unasked. This paper asks it.

The reason it matters for the 2024 findings is this. The themes that emerged from that study, belonging, commitment, open communication, inclusivity, and empowerment, are not causes of team performance. They are descriptions of what a psychologically safe environment produces. The 2024 paper described the outcomes of psychological safety without naming the condition beneath them. Edmondson named the condition. The present paper connects the two. What the 2024 data observed was the presence of psychological safety in some teacher relationships and its absence, or its fragility, in others. What the study could not do was explain why some teachers felt it and others did not, or why the same leadership context produced both responses. That explanation requires a closer examination of what makes people choose silence over voice, and that is the subject of the section that follows.

4. What the Pause Contained

The teacher who paused before answering was not pausing because she had nothing to say. The pause was not confusion and it was not emptiness. What looked, on the surface, like a brief hesitation was, in that moment, a moment of weighing: a rapid and quiet assessment of whether what she actually thought was safe to say. The safer answer that followed was shaped by that assessment. The silence was not empty. It was chosen.

This distinction matters enormously for how qualitative data from leadership studies should be read. If silence is treated as absence, as the neutral space between one answer and the next, then what a participant does not say simply does not register. It falls outside the data. But if silence is understood as a choice, shaped by the participant's assessment of risk, then what is not said becomes as meaningful as what is. The data collected in the 2024 study included what teachers said. It did not, and perhaps could not, include what they had decided not to say. That is the gap this section addresses.

Morrison (2014), reviewing two decades of research on employee voice and silence, established that silence in organisational settings is rarely passive. When employees withhold information, concerns, or genuine views from those in positions of authority, they are not simply failing to contribute. They are making a choice, shaped by their assessment of what speaking will cost them. Morrison identified the conditions that make silence more likely: perceived risk of negative evaluation, uncertainty about how responses will be received, and the absence of a felt sense that speaking will make any difference. These conditions do not require a hostile environment to operate. They function just as effectively in settings that are formally inclusive, formally collaborative, and formally safe. The form does not dissolve the fear.

Detert and Edmondson (2011) went further, identifying what they called implicit voice theories: taken-for-granted beliefs about when and why speaking up at work is risky or inappropriate. These beliefs are not always conscious. They do not always present themselves as fear. They manifest more quietly, as an instinct to wait, to observe how others respond before deciding one's own position, to hold back the clear thought that is already fully formed. Detert and Edmondson (2011) found that in formal settings, these implicit theories become stronger, more automatic, and more difficult to override. There is a shared understanding, felt rather than stated, that one assesses the situation before speaking. The distance between what one thinks and what one says is measured carefully, and adjusted accordingly.

This is not, it should be noted, irrational behaviour. It is learned behaviour (Detert and Edmondson, 2011; Milliken, Morrison and Hewlin, 2003). The assessment of interpersonal risk is shaped by experience: by previous moments when speaking freely produced consequences, by observations of what happened to others who did not hold back (Milliken, Morrison and Hewlin, 2003), by the accumulated evidence of how a particular environment responds to speaking freely. Over time, that assessment becomes a habit. One filters more. One speaks less.  Not because there is less to say, but because the habit of protection has become more automatic than the impulse to contribute. Detert and Edmondson (2011) found these implicit voice theories to be widely held and significantly predictive of workplace silence, even after accounting for individual personality and organisational context.

What this means for the 2024 findings is this. The teachers who gave careful, general answers were not disengaged. They were not indifferent to the shared vision their leaders had articulated. They were, in all probability, engaged enough to have a view, and uncertain enough about the consequences of expressing it to keep that view contained. The invitation to speak had been extended. The safety to accept that invitation fully had not been established. And without that safety, the invitation produced performed participation rather than genuine voice. The structures were present. The conditions were not.

Silence, then, is not the absence of something to say. It is the presence of a decision not to say it. And that decision, quiet and often invisible to the researcher collecting the data, is where the gap between participation and genuine voice actually lives.

5. The Missing Stage

Existing models of visionary and transformational leadership represent the relationship between leadership and team performance as a direct sequence. The leader communicates a vision, structures participation, and invites contribution. The team engages. Performance follows. Provided the leadership is clear and inclusive, the steps between invitation and outcome are treated as largely unproblematic.

This paper proposes that one stage is missing. Between the invitation and the response, a person decides. That decision determines whether the invitation produces genuine voice or performed participation. The model proposed here places that moment of decision at the centre, and identifies psychological safety as the condition that shapes which way it goes.

The model operates in four stages.

The second is the invitation to contribute. Teachers are asked to share views, contribute ideas, and engage with the vision that has been offered. This is where most models locate the mechanism of engagement. This is also where most models stop asking questions.

The third is the moment of decision. Before responding, a person makes an assessment, rapid, largely automatic, and shaped by accumulated experience, of whether speaking is safe. Detert and Edmondson (2011) identified this as the operation of implicit voice theories: taken-for-granted beliefs about when and why speaking up carries risk. Morrison (2014) established that this assessment operates in formally inclusive settings as readily as in hostile ones. The quality of the invitation does not dissolve it. What determines its outcome is the felt condition that surrounds the invitation, not the invitation itself. Morrison (2014) further noted that leaders are often entirely unaware of this tendency, assuming that employees feel free to speak and that the absence of dissent signals agreement. The silence is invisible to the person in authority precisely because it is performing the appearance of participation. Motivation describes how much a person wants to contribute (Bass, 1985). The moment of decision is different. It is not about wanting. It is about whether, in this encounter, at this moment, speaking feels safe.

That felt condition is the fourth stage: psychological safety. Edmondson (1999) defined it as a shared belief that the team is safe for interpersonal risk-taking. Frazier et al. (2017), drawing on 136 independent samples representing over 22,000 individuals, confirmed its robust relationship with voice behaviour, learning, and performance. In the model proposed here, psychological safety does not sit alongside the decision point. It shapes it. When it is present, the decision tends towards voice. When it is absent, the decision tends towards self-safety, the quiet preference for giving the answer that risks the least. A response was given. The contribution was recorded. Whether it reflected what the teacher genuinely thought is what the model cannot confirm, and what the interview alone could not reveal.

The distinction this model draws is between genuine voice and performed participation. Genuine voice is substantive, honest, and generative. It contributes to team performance not by adding numbers to a participation count but by bringing real thinking into a shared process. Performed participation occupies the same structural space without producing the same effect. It satisfies the form of inclusion whilst leaving its substance unrealised. The two are invisible to any measure that counts only whether people spoke. They become visible only when one asks what determined whether people spoke freely. Most models of visionary leadership stop at leadership and engagement. This model says: leadership invites voice, but psychological safety determines whether voice actually happens. That addition, one stage between the invitation and the outcome, is what the 2024 data pointed towards without being able to name.

Psychological safety does not arise from nothing. Research has identified leadership behaviour as its primary antecedent, with leader inclusiveness, modelling of fallibility, and responsiveness to voice all contributing to whether safety develops within a team (Edmondson and Lei, 2014; Frazier et al., 2017). The antecedents of psychological safety are well documented elsewhere. The contribution of this paper is to show where safety sits in the leadership-to-voice sequence. The moment of decision as proposed here draws on the antecedent literature on voice behaviour, especially Detert and Edmondson (2011) and Morrison (2014), and applies that body of work to the visionary leadership sequence where it has not previously been located. The empirical question of what produces or inhibits psychological safety in specific school contexts remains an important direction for future research.

It is what the pause contained.

6. The Same Condition in the Research Encounter

The argument this paper has made about teacher voice in schools applies with equal force to the conditions under which teachers speak to researchers. It is the same condition appearing in a different encounter.

When the researcher sat across from the teacher who paused, every procedural requirement had been met. Ethical clearance obtained. Consent given. Confidentiality assured. The process was proper. What the process could not do was change the felt reality of the encounter: that one person was asking and another was responding, and that the document on the table could not guarantee how safe the teacher felt in that moment.

Cohen, Manion and Morrison (2000) described the fundamental ethical tension in social research as a costs and benefits ratio. The participant makes the same assessment from the other side of the table. The teacher who gave a careful answer was not being uncooperative. She was being cautious, as any person might be when asked to speak in an unfamiliar encounter.

Milliken, Morrison and Hewlin (2003), in their study of employee silence, found that 85 per cent of respondents had been in situations where they felt unable to raise an issue even when they considered it important. The most frequently cited reasons were not indifference. They were the fear of being labelled negatively and the fear of damaging valued relationships. That fear does not disappear when the conversation moves from a workplace to a research interview. It follows the person into the encounter.

Formal procedures address this asymmetry structurally. They do not dissolve it experientially. Strip away every procedural assurance and ask what remains that no document can guarantee. What remains is the participant's felt sense of safety. That feeling cannot be administered. It can only be earned through the quality of the human relationship the researcher builds before, during, and after the interview.

In both cases, the same mechanism governs the response:  a quieter voice makes an assessment about whether speaking freely is safe. Kawabata and Gastaldo (2015) argued precisely this point in the context of qualitative research: that silence is not an absence of communication but a communication strategy, and that researchers who treat hesitation and withholding as empty space in the data miss what those moments actually contain. Participants' meaning-making, they noted, is often embedded in what is left unsaid. Their argument was developed in a culturally specific setting, drawing on interviews with marginalised workers in Japan. The principle, however, extends beyond that context to any research encounter in which the felt sense of safety shapes what a participant chooses to say and what they choose to withhold.

The British Educational Research Association (BERA) inquiry into research and teacher education is instructive here. Beauchamp, Clarke, Hulme and Murray (2013) argued that teachers are intellectual professionals whose knowledge and judgement carry weight. When a teacher chooses not to speak freely, she is not failing the research. She is exercising professional judgement. The researcher's task is not to overcome that judgement but to create conditions safe enough that she might choose differently.

This is, in the end, the same task that visionary leaders face in their schools. Whether a teacher is speaking to a principal about a shared vision, or to a researcher about her experience of leadership, psychological safety, the felt sense that speaking is safe, shapes what is said and what is left unsaid (Edmondson, 1999). Leadership cannot create that felt sense by designing structures alone. Research cannot create it by completing forms alone. Both require something no document can verify: the felt sense that this situation, this person, this moment, is safe enough to speak in. This paper does not attempt a full treatment of research ethics. It notes one specific implication of the psychological safety argument: that qualitative data on voice and participation may be shaped by the very condition the research is designed to examine.

7. Conclusion: What the Pause Contained

This paper began in a room, with a teacher who paused.

The pause was brief. What followed was careful: an answer that stayed on the surface, that offered what was appropriate rather than what was real. The researcher moved on, as the study required. But the moment did not move on. It stayed, unresolved, at the edge of the data, pointing towards something the 2024 study could describe but not explain.

What it pointed towards was this. A question the study had not been designed to ask. Why did the teacher hold back when the invitation to speak had been genuinely given?

That observation, held for long enough, became the argument of this paper. Visionary and transformational leadership research has attended carefully to what leaders do: how they communicate a vision, how they structure participation, how they invite contribution. What it has not attended to, with the same care, is the condition that determines whether the invitation is accepted genuinely or performed carefully. That condition is psychological safety. Its absence is not visible in the structures of participation. It is visible only in the pause before the answer, in the careful phrasing that follows, in the gradual withdrawal of a voice that was never fully given.

The model proposed here places psychological safety between the invitation and the response. It does not replace the existing literature on visionary leadership. It extends it, by naming the stage that existing models skipped over: the moment of decision, shaped by the felt sense of safety, that determines whether participation becomes genuine voice or remains present but withheld.

That model now requires empirical validation. The questions it generates are not abstract. They are the questions this paper could not ask in 2024, because the study that prompted it was not designed for them. What does it feel like to be invited to contribute to a shared vision? What determines whether a teacher speaks freely or holds back? What would make a teacher feel safe enough to say what is genuinely meant?

These are the questions that the present paper cannot answer, and that future empirical work in this area will need to pursue.

Psychological safety is not a fixed attribute of a leadership context. It is a relational and dynamic condition, one that can vary across individual teacher-leader relationships even within the same school and under the same leadership.

What this paper can offer, in the meantime, is a reframing. Creating conditions for genuine contribution is not the same as guaranteeing it. Procedural conditions can be designed and documented. The felt sense of safety that genuine voice requires cannot be designed or documented. It can only be earned through the quality of the human relationship that surrounds the conversation, whether that conversation is an interview or a group session. The condition is the same in each case. What changes is only who is in the conversation and what each person brings to it.

No procedure can guarantee that feeling. No assurance, however sincerely made, can substitute for it. Safety is not something that can be declared. It is something that has to be felt.

The condition this paper has sought to name is not new. It has been present in every room where one person asked and another decided, in the space of a pause, whether to answer honestly. Safety, when it is genuine, does not depend on being needed. That is precisely what makes it rare. What is new, perhaps, is the willingness to ask what the pause contains, and to take the answer seriously. The pause has always been there. The question is whether it has been heard.

Acknowledgement: This paper began over coffee, when a friend asked a simple question: what is the point of asking, if people do not want to speak up? The question landed because the author had seen something similar in the data from an earlier study on visionary leadership and team performance (Chin, 2024), where a teacher had paused before answering in a way the study could observe but not explain. The question asked over coffee and the pause observed in the data were, it turned out, two versions of the same silence, and this paper is the attempt to name it. The author is grateful to the teachers whose participation in the 2024 study made this paper possible. Their voices, spoken and withheld, are what it is about.

Conflict of Interest: The author declares that the study was conducted without any existing commercial or financial associations that could be interpreted as a potential conflict of interest.

Funding: This research received no external funding.

Declaration of Generative AI and AI-assisted Technologies: This study has not used any generative AI tools or technologies in the preparation of this manuscript.