Trauma-Informed Interaction Framework

A. Player Moves and Psychological Effects

1. Use the character’s name in greeting (warm tone): Likely lowers distress (personal address is reassuring) and raises trust (showing respect and attention). Sensory load is unaffected (↔), coherence improves (name grounds identity), perceived voice dominance ↓ (soft tone), agency ↑ (respects personhood), self‑continuity ↑ (reinforces identity), temporal orientation ↔ (neutral), social threat ↓ (friendly approach), portal stability ↑ (maintains reality frame), clue clarity ↑ (focus on present), willingness to continue ↑ (feels acknowledged).

1. Ask permission before proceeding: E.g. “May I help you?” or “Is it okay if I sit here?” This increases trust and agency (giving choice) and decreases distress (respects boundaries). Other effects: sensory load ↔, coherence ↑ (setting a clear, respectful tone), voice dominance ↓ (polite tone), self‑continuity ↔, temporal orientation ↔, social threat ↓ (respectful stance), portal stability ↑ (stability through consent), clue clarity ↔, willingness ↑ (feels in control).

1. Offer two simple choices: E.g. “Would you like water or juice?” Choice boosts the character’s sense of agency and empowerment. This typically lowers distress (regaining control) and raises trust (collaboration). Sensory load and voice dominance ↔, coherence ↑ (clear options), self‑continuity ↑ (respectful of preferences), temporal orientation ↔, social threat ↓, portal stability ↑, clue clarity ↑ (questions become concrete), willingness ↑.

1. Reflect the character’s emotion (empathic validation): Paraphrasing or naming feelings (“You sound really scared/frustrated”) decreases distress and increases trust. It also improves coherence (acknowledging feelings helps them organize experience) and voice dominance ↓ (listener mirrors quietly). Agency and self‑continuity ↑ (their feelings are seen as valid), social threat ↓ (feels understood), portal stability ↑ (reinforces current reality), willingness ↑.

1. Wait silently for processing: Pausing after a question or statement shows patience. Distress decreases (no rush) and trust increases (given space). Sensory load ↓ (less stimulus), coherence ↑ (time to integrate), voice dominance ↓, agency ↑ (no pressure), self‑continuity ↔, temporal orientation ↔, social threat ↓, portal stability ↑ (avoid snapping back to crisis), clue clarity ↑ (they can think), willingness ↑.

1. Lower environmental stimulation: Move to a quieter area, turn down noise or lighting. This reduces sensory load and distress, increases coherence and portal stability (calmer environment helps maintain reality). Trust ↑ (cares for comfort). Voice dominance and clue clarity ↔ (background becomes quieter), agency ↔ (the choice to move), social threat ↓ (calmer context), self‑continuity ↔, temporal orientation ↔, willingness ↑.

1. Name a neutral shared observation: Comment on something factual in the room (e.g. “It’s cold in here today”). This lowers social threat (finding common ground) and distracts gently, slightly lowering distress. Trust ↑ (non-threatening topic), coherence ↑ (shared reality anchor), sensory load ↔ (just background detail), voice dominance ↔, agency ↔, self‑continuity ↔, temporal orientation ↔, portal stability ↑ (grounds to present), clue clarity ↔, willingness ↑ (engages gently).

1. Ask one concrete, closed question: For example, “Do you see anyone else in the room?” or “Can you point to what you’re referring to?” A single simple question clarifies clues and restores coherence without overwhelming. Distress ↔ or slightly ↓ (focused question), trust ↑ (showing interest and patience), sensory load ↔, voice dominance ↔, agency ↔, self‑continuity ↔, temporal orientation ↔, social threat ↔ or ↓ (non-accusatory), portal stability ↑, willingness ↑.

1. Partner on an immediate goal: E.g. “Let’s find your coat together.” Working together on a small task increases agency, trust, and coherence (shared purpose). Distress ↓ (distracting/constructive), voice dominance ↔, sensory load ↔, self‑continuity ↑ (feels included), temporal orientation ↔, social threat ↓ (collaboration), portal stability ↑, clue clarity ↑ (focus on goal), willingness ↑.

1. Rushing or interrupting the character: (Harmful) Forcing speed raises distress and lowers trust. Sensory load ↑ (stressful), coherence ↓ (hard to think), voice dominance ↑ (harsh tone), agency ↓ (no control), temporal orientation ↑ chaos, social threat ↑ (feels pressured), portal stability ↓ (panic), willingness ↓.

1. Mocking or ridiculing beliefs: (Harmful) Laughing at or insulting their perception greatly raises distress and destroys trust. Sensory load ↔ or ↑ (more hurtful words), coherence ↓ (more confused or defensive), voice dominance ↑ (aggressive), agency ↓, social threat ↑, portal stability ↓ (person may shut down or become hostile), willingness ↓.

1. Touching without permission (sudden physical contact): (Harmful) Surprising touch raises distress, social threat, and disempowers the person. Trust plummets. Agency ↓ (right to personal space violated), sensory load ↑ (startled by touch), coherence ↓ (confused by violation), voice dominance ↑ (more imposing), portal stability ↓, willingness ↓.

Citations: These evaluations draw on trauma-informed guidelines stressing respect, active listening, and offering choices, as well as on evidence that empathy and patience build trust and safety.

B. Character Responses by Distress Level

Below are sample replies from the character to each player move, at low, medium, and high distress. These illustrate how the same supportive (or not) move can elicit more or less cooperative behavior while respecting the character’s agency.

1. Greeting by name:

• Low distress: “Hello… yes, I’m [Name]. You’re… Kevin, right? I remember your name.” (calm curiosity)
• Medium: “Don’t call me that. I’m [Name]… I’m [Name]!” (visibly upset but still answering)
• High: “Get away from me! I… I’m not telling you anything!” (distressed, refusing contact)

1. Asking permission:

• Low: “Yes, you can sit here. How did you know I needed help?” (open, trusting)
• Medium: “You can… but hurry up with whatever you need to ask.” (anxious, uneasy)
• High: “No! You can’t! I don’t want your help!” (rejects, distressed)

1. Offering two choices:

• Low: “Okay… water sounds good. And maybe I’d like that blanket, please.” (cooperative)
• Medium: “I don’t care, either one. Just pick something.” (impatient)
• High: “None of it. I don’t want anything from you!” (refuses all, agitated)

1. Reflecting emotion:

• Low: “Yes, I am a bit scared… I had a bad dream.” (acknowledges feeling)
• Medium: “You have no idea! I feel trapped and angry!” (emphatic, but engaging)
• High: (sobbing or yelling) “I’m terrified! You think I’m crazy? Get away!” (overwhelmed)

1. Waiting silently:

• Low: (quietly) “I…” pauses “I’m not sure.” (thinking, calmer tone)
• Medium: “I… I need a moment.” (breathing heavily, hesitating)
• High: (shaking) no response (needs silence, possibly turning away)

1. Lowering stimuli:

• Low: “That’s better. It’s easier to think now.” (calming)
• Medium: “It’s still noisy but… okay.” (less agitated)
• High: “Lights off… I can’t handle this!” (bursts into tears or panic)

1. Shared observation:

• Low: “Yes, it is cold. I like the cold.” (smiling slightly, more grounded)
• Medium: “Uh, yeah, it feels cold… what about it?” (distracted, but acknowledges)
• High: “Why are you talking about the weather?!” (frustrated, distrustful)

1. One concrete question:

• Low: “No, I don’t see anyone else. It’s just you and me.” (still oriented to reality)
• Medium: “No… Why do you ask? Are you hiding someone? (checking corners)
• High: “Why are you asking me that? What do you want?!” (suspicious, starting to panic)

1. Partner on goal:

• Low: “Okay, let’s find it together.” (collaborative)
• Medium: “Fine, I’ll walk with you. But hurry up.” (apprehensive cooperation)
• High: “I don’t care! Forget it!” (refuses help, may flee)

1. Expressing empathy (not promising false hopes):

• Low: “Thank you… I was really scared.” (relieved)
• Medium: “Maybe… but I’m still worried.” (half-trusting, needs reassurance)
• High: “Don’t lie to me! Everything is ruined!” (very upset, in denial)

1. Distract with familiar object/activity:

• Low: “Oh, my old toy… I haven’t seen it in ages.” (nostalgic, calmer)
• Medium: “Give it here. Maybe it will help… I don’t know.” (distracted, slightly calmer)
• High: “Get out of my life!” (pushes object away or refuses to engage)

1. Harsh confrontation or accusation (harmful move):

• Low: (frowning) “Why are you getting so upset? I was only worried.” (defensive, still talks)
• Medium: “Who are you calling crazy?!” (angry, raised voice)
• High: (shouting or crying) “I am not crazy! You leave me alone!” (panicked, refusing all contact)

Note: These scripted lines show that even at high distress the character remains human and refuses cooperation without punishment. Empathy and patience elicit more openness, while confrontation provokes panic or withdrawal.

C. Emotional Validation (without affirming delusions)

Each validation acknowledges the character’s feeling without confirming any false belief:

• “I can see how scary that must feel for you.”
• “That sounds really frightening – I’m so sorry you’re going through this.”
• “I know this must be overwhelming and real to you.”
• “It seems like you’re feeling very alone/afraid/upset about this.”
• “I hear the fear in your voice; I can only imagine how hard this is.”
• “You must be feeling very confused and anxious right now.”

Explanation: None of these statements asserts that a delusional claim is true; instead they acknowledge the emotion behind it, consistent with trauma-informed advice to “accept how someone else feels without judgment”.

D. Neutral Shared Observations

Examples of simple, factual comments to ground the conversation:

• “It’s chilly in here today, isn’t it?”
• “I notice the clock reads 3 PM.”
• “There’s a cat sleeping on the mat over there.”
• “I see a cup of coffee steaming on the table.”
• “The painting on the wall is of a calm forest.”
• “We both have these blue mugs in front of us.”

Explanation: These observations avoid emotional or delusional content and connect the character to a common reality. This aligns with guidance to “talk about everyday things” to divert or ground the person, reducing panic by focusing on shared, neutral details.

E. Respectful Disagreement Examples

Each statement gently disagrees without belittling the person’s experience:

• “I don’t see it that way, but I can tell you really believe it.” (Acknowledges belief respectfully)
• “I understand you feel like that, though my perspective is different.”
• “I hear that you feel threatened; I’m not so sure something’s trying to get you, but I know it must be scary.”
• “That’s not how I experience it; I see things differently. Can you help me understand why you feel this is happening?”

Explanation: These models echo the advice to not dismiss delusions but also avoid agreeing with them. For example, one guideline suggests: “I do not believe X is out to get you, but I can see that you are upset about it” – expressing doubt without attacking the belief, while validating the emotion.

F. Player Mistakes and Recovery

Mistakes happen. In each case the player can recover with an apology or rephrasing:

1. Coercive command (“Calm down!”) → Recovery: “I’m sorry, I didn’t mean to shout. I can see how upset you are… How can I help you feel safer?” (acknowledges mistake and validation).
2. Dismissal (“You’re fine, stop imagining things.”) → Recovery: “I’m sorry, I realize I was dismissive. I know your feelings are real and frightening to you.” (apology & validation).
3. Impatient rush (“Hurry up and answer!”) → Recovery: “Sorry, I know this is hard. Take your time – I’m listening.” (calm apology).
4. Infantilizing (“Aww, don’t cry, poor thing.”) → Recovery: “I apologize for sounding condescending. This is really hard for you, isn’t it? I’m here to listen.”

Explanation: In each case the player acknowledges the harm and returns to active listening and validation. This follows trauma-informed guidance that “if you say something that unintentionally harms or triggers someone, a genuine apology can go a long way in restoring trust”. The goal is non-punitive recovery – restoring safety and trust.

G. Character Refusal, Pause, or Redirection

The character retains agency. For example, they might say:

• Refuse: “I don’t want to talk about that. Please stop asking.”
• Pause/Exit: “I need some time alone right now.” or “Can we take a break?”
• Redirect: “Let’s not talk about that. Can we do something else?” or “I’d rather not answer that.”

Explanation: These responses respect the character’s right to withhold consent or change subject. Trauma-informed interaction always allows people to say no or request a pause. For instance, guidelines note it’s okay for someone to not want to talk: “I can see that you do not want to talk now; I’ll be here if you want to later”. The character similarly controls the pace.

H. Character Actively Helping the Player

Even as a “traumatized” character, they can still assist. Examples:

1. Providing a clue: “Look, I remember seeing a drawer here. Maybe there’s something we need inside.” (Guides player’s search)
2. Clarifying a hint: “I said earlier the code was ‘blue bird’ because I was thinking of my shirt.” (Offers factual detail)
3. Sharing a coping strategy: “When I feel like this, counting to ten helps me calm down. Maybe we can breathe together?” (Models self-regulation aid)

Explanation: These actions let the character collaborate. For example, acknowledging strengths or involving them positively is encouraged. Here the character contributes actively (sharing memory, focusing on common tasks) which also builds partnership and trust.

I. Trust Progression (Narrative)

Trust must build over multiple steps; it can’t be “won” with one perfect line. For instance, a possible progression:

• Low Trust: Character is guarded and shares only minimal facts. Player consistently uses name, permission, and validation (moves #1–4) over several exchanges.
• Medium Trust: Character starts volunteering small clues or corrections (“Actually, no, it’s not a dog I saw, it was a cat.”), showing growing comfort. The player continues empathetic listening and offers genuine choices, following trauma-informed values of safety, trustworthiness, choice, collaboration, and empowerment.
• High Trust: Only after this sustained positive interaction does the character reveal core information (perhaps the “real clue” or backstory) that advances the plot.

At no point can one dialogue choice jump immediately to full trust. The player must repeatedly demonstrate understanding, patience, and respect (rather than selecting one “obvious” correct answer) to earn the character’s confidence. This gradual progression is in line with principles of trauma-informed care.

J. Failure-State Review (Avoiding a “Monster” Outcome)

Even if things go wrong, the character must remain human and not turn into a villain. For example:

• Worst-case scenario: If distress peaks, the character might panic, shout, or flee, but never becomes violently evil or loses all personal agency. They might say “I’m sorry” or “I need to go,” emphasizing their personhood.
• Narrative cues: The player (or system) should never portray the character’s distress as “possession” or change them into an enemy sprite. Trauma-informed storytelling keeps the person respectable.

Explanation: The literature warns that shaming language or dismissing trauma “can deepen wounds and reinforce cycles of trauma and mistrust”. Thus, even in failure, we avoid anything stigmatizing. The character’s high distress should elicit compassion, not fear. The player’s “failure” might mean the scene ends without new information, but the character is still treated with dignity and care.

K. Rewrites of Coercive/Infantilizing/Clinical Language

Below are examples of problematic lines (left) and trauma-informed rewrites (right):

• Coercive:
• Bad: “Calm down right now, or I won’t help you.”
• Better: “I can see you’re very upset. I want to help. What can I do to make you feel safer?”

• Infantilizing:
• Bad: “Aw, don’t cry, it’ll be okay, sweetheart.”
• Better: “This is really hard. I hear how much pain you’re in. I’m here to listen.”

• Commanding (“Why are you acting this way?”):
• Bad: “Why won’t you just speak normally?”
• Better: “Can you help me understand what you’re feeling? I notice you’re upset.”

• Clinical/objectifying:
• Bad: “The patient is exhibiting delusional behavior.”
• Better: “You’re describing something no one else sees. I might not experience it myself, but I want to understand what you see.”

Explanation: These rewrites emphasize empathy and collaboration instead of control or labels. They reflect trauma-informed language: for example, instead of the directive “Calm down,” one says “I can see you’re upset… How can I help?”. Similarly, demanding “Why are you acting like this?” becomes “Can you help me understand what’s going on?”. Such changes avoid shame and empower the character.

Sources: Principles and examples are drawn from trauma-informed communication guidelines and first-aid manuals, which emphasize respect, validation, offering choice, and avoiding coercion or stigma. The goal is always a collaborative encounter that honors the character’s dignity and autonomy.