The article examines self-censorship as a lived professional mechanism rather than as a purely structural indicator. The central problem of the study is the tension between the formally guaranteed freedom of expression in a democratic system and the recurring, everyday practices through which journalists narrow down, soften, postpone, or abandon public-interest information – often without an explicit prohibition, yet with very real anticipated costs.
In this article, self-censorship is defined as the conscious and voluntary withholding of factually reliable information that is considered important, in cases where no formal legal or institutional obstacle prevents disclosure. This distinction is crucial: ethical editorial decisions (e.g., source protection, harm prevention, professional standards) follow a logic of responsibility, whereas self-censorship begins at the threshold where information is deemed publishable in principle, but its projected price becomes ‘too high’. The paper also uses the term ‘(self-)censorship’ to mark a continuum between direct censorship and internalized restraint, while the term ‘self-censorship’ is reserved for the operationally delimited phenomenon described above.
The aim of the research is to reconstruct how Lithuanian journalists experience, reflect upon, and rationalize self-censorship situations, and how these experiences reshape professional identity-including the relationship to journalism’s normative democratic mission (the ‘watchdog’ role). Empirically, the study is grounded in a qualitative design: 15 semi-structured, in-depth interviews (13 journalists and 2 newsroom leaders), conducted and analysed in 2025–2026 by using thematic coding and episode-based (micro-event) analysis, and informed by a phenomenological orientation to experience, temporality, and meaning.
The theoretical framework is deliberately multi-level. At the micro level, Bar-Tal’s conceptualization provides criteria for distinguishing self-censorship from routine selection. At the macro level, Bourdieu’s field theory and Herman and Chomsky’s propaganda model illuminate how ownership, advertising dependence, ratings logic, source regimes, and organized backlash (‘flak’) shape what becomes sayable long before a single sentence has been written. Whereas, at the meso level, gatekeeping theory locates where restraint materializes in the newsroom chain topic pitching, editing, publication decisions, authorship regimes, and ‘unwritten rules’. The study additionally draws on Noelle-Neumann’s spiral of silence to conceptualize audience pressure and reputational sanctioning as environmental forces that narrow the corridor of ‘safe speech’. Finally, the phenomenological–affective line (Merleau-Ponty) and the theory of emotional labour (Hochschild) help show how structural pressure enters the body: as micro-pauses, tonal discipline, anticipatory avoidance, and a sedimented reflex of ‘better not’.
The analysis indicates that self-censorship most commonly operates not as a dramatic act of silencing but as anticipatory self-editing – as an internalized gate which activates before a statement exists. It emerges through overlapping mechanisms: structural filters (ownership, advertising, ratings), editorial gatekeeping (expectations and routinized ‘maps’ of what will pass), and audience sanctions (reputational risk, online hostility, labeling). Crucially, restraint is frequently normalized through rationalization: decisions are re-described as ‘professionalism’, ‘neutrality’, ‘responsibility’, or ‘common sense’, making the mechanism both socially acceptable and less visible to the actor herself.
In terms of professional identity, the thesis shows a recurring shift from the journalist as a normative watchdog toward the journalist as a risk manager – someone who continually calculates reputational, organizational, procedural, and emotional costs. In this transformation, the boundary between democratic mission and self-protection becomes a daily negotiation, whereas the ‘electric fence’ of discourse increasingly functions as an internal reflex: not merely shocking a thought, but teaching it to reroute before it has become a sentence.

This work is licensed under a Creative Commons Attribution 4.0 International License.