Published by Unseen Progress, an independent publisher of caregiver research. Last reviewed 2026-05-10. Part of the selective mutism research overview.
Short answer. Sliding-in (also called stimulus fading) is the core behavioural mechanism of evidence-based selective mutism treatment. A new person is introduced into a setting where the child is already speaking to a trusted adult — first at the threshold, then a few feet away, then in the room, then in the conversation — until the child is speaking to the new person. The mechanism is well-validated (Bergman et al., 2013; Kurtz, 2020; McHolm, Cunningham & Vanier, 2005) and parents can run a simplified version at home with realistic expectations and the right structure.
Selective mutism is an anxiety-driven freeze response that fires when a new (or insufficiently safe-feeling) person enters a setting in which speech is expected. The freeze response collapses the child's ability to speak in that moment, even when they want to. Sliding-in works by not triggering the freeze. The trusted adult and the child are already in a comfortable, speaking conversation. The new person is introduced so gradually that the child's nervous system never crosses the threshold from speaking to frozen. The new person ends up inside the speaking circle without the freeze ever being triggered.
McHolm, Cunningham & Vanier (2005) describe this as a transfer of stimulus control: the speech behaviour, originally bound to the trusted adult, transfers to a new adult by way of overlapping presence rather than direct demand.
The Bergman/Kurtz protocols and the McHolm family treatment manual describe broadly similar steps. A simplified parent-friendly version:
Pick an activity the child already speaks during with the trusted adult — a board game, a counting game, reading aloud, naming animals on flashcards. The activity should naturally pull short verbal responses ("blue", "six", "lion") rather than free conversation. Do this in a quiet, private room.
The new person stands outside the open door, not visible to the child but audible. The child continues the speaking activity with the trusted adult. The new person says nothing.
The new person enters the room and sits as far from the activity as the room allows, ideally turned slightly away. They do not speak to the child. The trusted adult continues the activity. The child's job is to keep speaking to the trusted adult — not to the new person.
Over several minutes — or several sessions, depending on the child — the new person moves progressively closer. They remain a quiet observer. The child is still answering the trusted adult.
The new person sits at the table. They participate in the game (move a piece, turn a card) without speaking. The trusted adult continues to be the one asking the child questions.
The new person asks a question that has a single-word answer and that the trusted adult could answer for the child if needed ("What colour is that?"). The trusted adult waits. If the child answers, the new person responds warmly but briefly and continues the activity. If the child doesn't answer, the trusted adult asks the same question and the child answers them — and the activity continues without comment.
Repeat the protocol with the new person across different rooms, then in different settings, then with different new people. Each round generalises faster than the last as the child accumulates evidence that new-person exposures are survivable.
Do: Stay calm, neutral, and unsurprised by every brave attempt. Praise quietly and specifically afterwards ("I liked how you answered her, that was brave"). Plan the session in advance — who, where, what activity, what question. Stop while the child is still in the speaking zone, not after a freeze.
Do not: Announce that the new person is coming as if it is a test. Coach the child to speak. Apologise to the new person on the child's behalf. Repeat the question. Praise loudly, especially in front of the new person. Push for a second session immediately after a hard one.
The Kurtz PCIT-SM protocol explicitly trains parents in these "do not" behaviours because they are the ones that, despite good intent, reinstate the demand frame and reactivate the freeze.
In structured therapy with a trained provider (Bergman et al., 2013; Kurtz, 2020), measurable shifts in expected-speaker settings appear within 3–6 months and full functional communication often takes 1–3 years. Parent-led sliding-in at home, used as a complement to structured therapy, generally produces faster gains within the family's own social network — extended family, neighbours, family friends — than in school settings, because the parent has more direct control over the variables.
1. Skipping steps because the child seemed fine. A child who tolerated the new person at the threshold has not yet spoken to them; the next session needs to start one step back, not three steps forward. 2. The new person breaking protocol. A grandparent who promises not to speak and then says "hi sweetheart, did you have a good day?" the moment they see the child resets the freeze. Brief the new person privately and explicitly. Show them the protocol on paper. 3. Doing too much in one session. A 20-minute session that ends in the speaking zone is more useful than a 60-minute session that ends in shutdown. 4. Treating sliding-in as a one-off event. Each new person needs their own sliding-in. Generalisation accelerates with practice but is not automatic. 5. Pairing sliding-in with a verbal demand from a parent. "See, I told you you could do it" said immediately after a brave answer turns the next session into a performance test.
Sliding-in is the engine of behavioural treatment for selective mutism, but it is not the whole treatment. It is paired with:
The Selective Mutism Association (selectivemutism.org) and the SMart Center (selectivemutismcenter.org) both publish parent-facing protocols that align with this structure.
---
Unseen Progress publishes long-form caregiver research and builds research-backed daily trackers for the families covered. See the full selective mutism research overview for the complete framework.