The Person Who Was Never Safe to Talk To: Finding That Safety With AI
Some things don't have clean names. There's a particular kind of damage that comes from growing up with — or spending years in a relationship with — someone to whom you could not speak freely. Not because they were overtly abusive, though sometimes they were. But because something about speaking your real thoughts or feelings around them always went wrong. It cost you something. The reprimand, the dismissal, the inexplicable coldness that followed, the way the conversation would somehow turn against you. You learned. Of course you learned. The brain is nothing if not efficient at learning what it costs you to be seen. So you stopped being seen. Or you learned to show a version of yourself carefully calibrated to be safe.
The Long Reach of Unsafe Relationships
The people who made speech unsafe for you often weren't strangers or acquaintances. They were parents. Partners. Friends you trusted. The intimacy of those relationships made the learning deeper, because it attached the fear to closeness itself. Not just: this specific person is dangerous. But: speaking from the self to another person who matters is dangerous. Research from the University of Minnesota's attachment lab found that early relational experiences characterized by emotional unpredictability or dismissiveness had measurable effects on communication patterns in adulthood — specifically on willingness to disclose, comfort with conflict, and the capacity to tolerate the vulnerability of being honest in close relationships. These weren't beliefs people could simply think their way out of. They were embedded in behavioral patterns laid down through thousands of early interactions.
What It's Like to Have This Pattern
People who carry this often don't identify it as a problem for years. They're articulate in professional contexts. They can speak in meetings and to strangers and in low-stakes social situations without difficulty. What they can't do — or can only do with enormous cost — is speak genuinely to people they're close to. The closer someone gets, the less themselves they can be. Intimacy activates the pattern. What comes out in close relationships is the performed self, the managed self, the one who was trained to stay safe. One sign of this is the experience of rehearsing conversations before having them — particularly conversations where you're going to say something vulnerable or honest. Another is the tendency to express real feelings only in writing, or after the other person has left the room. Another is the sense of relief you feel when a social engagement is canceled.
Finding Safety, Practice by Practice
Recovery from this pattern has to be relational — it can't happen in isolation, because the injury is to relational speech. But it also can't happen by jumping directly into the high-stakes intimacy that triggers the pattern most acutely. This is where structured, lower-stakes interaction environments can serve a genuine function. Some people find this in therapy, which provides the explicit security of a boundaried relationship with someone trained not to punish disclosure. Some find it in carefully chosen friendships developed slowly. Some have found genuine utility in AI companions — not as a substitute for human connection but as a space to rediscover that speaking from the self is possible.
The Tangent on Authenticity
There's an interesting philosophical question embedded here about what authenticity means when you've spent years curating a safe self. Who is the authentic self if not the one you've been performing? But most people who carry this pattern know the answer experientially — there's a difference between what they say and what they mean, and they feel the gap as a kind of persistent low-grade grief. The goal isn't to become a different person. It's to find the one who's been there all along, waiting for the conversation to become safe enough to join.
What Happens When Speech Becomes Safe
Research from Johns Hopkins found that people who described experiencing increased relational safety over time — through therapeutic relationships or consistently supportive personal relationships — showed measurable shifts in interpersonal communication patterns, moving toward more direct expression and less defensive management. This shift doesn't happen all at once. It happens the way trust is built: incrementally, through small experiences of disclosure followed by non-punishment. Over time the neural prediction shifts: saying the real thing doesn't always cost you. Eventually, it often doesn't. And eventually, the performance becomes less necessary, because the safety becomes more real. That safety is worth finding. And it starts wherever you can start.
Want to discuss this with Dr. Amara?
No signup needed · Start chatting instantly
Ask Dr. Amara About This →