I cried in front of her, which startled her. Then she said it again. That was the beginning. The years since have been a slow, uneven, miraculous expansion of her language, and this is what it actually looked like, not what the parenting books promised.
I'm writing this as a personal essay, not a how-to. The how-to articles are elsewhere on this site. This is just the story.
Before words existed in our house
For the first twenty-five months of her life, my daughter didn't have words. She had sounds, babbling, laughing. She had a particular way of grabbing my hand and pulling me across the room toward whatever she wanted. She had eye contact when she chose to give it. She had every kind of communication except the kind we count as "talking."
I noticed the gap around fifteen months. She'd had two words, "dada" and "ball," and then she didn't have those anymore. The words were retreating, not advancing. That regression was the thing that scared me.
I asked our pediatrician. She said wait. I waited. I asked again at the eighteen-month visit. She said wait more. I waited more. At twenty months, I called early intervention myself. The evaluator came to our house, watched my daughter play, and said, "She is autistic. We will get her speech therapy. We will get her OT. Tomorrow."
That was the beginning of the real work.
Six months on the floor
Early intervention came to our house twice a week. The speech-language pathologist, a woman named Courtney from a practice in suburban Philadelphia, sat on the floor with my daughter. She didn't do flashcards. She never said "say apple." She joined whatever my daughter was already doing and layered language onto it.
My daughter would be lining up cars, and Courtney would sit down and say, "Long line. Red car. Blue car." My daughter would be spinning a top: "Spin. Round and round. Spinning." My daughter would be holding a snack she couldn't open: "Open. Help. Open the snack." She never asked my daughter to say anything back. She just put words onto the moments my daughter was already living in, for six months straight.
I learned to do the same thing, slowly, badly. I narrated. I waited. I expanded the few sounds my daughter made. I bought books. I bought a sensory bin. I sat on the floor until my knees ached.
For the first six months, nothing visible changed. More sounds, maybe, more eye contact. Courtney said this was real progress and to trust it. I trusted it because I didn't have a better option.
I remember sitting in my car after one session, calling my wife, and saying, "I don't know if this is working." She said, "What's the alternative?" There wasn't one. So we kept going.
"More"
At twenty-six months, Courtney came over and we were sitting at the kitchen table. My daughter was eating goldfish. I wasn't paying attention. She held out her hand. I gave her another goldfish. She held out her hand again. I gave her another. She held out her hand a third time. I said, "Sorry, baby, that's enough goldfish."
She looked at me. She looked at the cup of goldfish I was holding. She said, "More."
I don't remember what happened next. I think I dropped the cup. I know I cried. My daughter looked at me with confusion, because I had never cried in front of her before. Courtney, sitting across the table, laughed and said, "Yes, more. More goldfish. Buddy, you said 'more.'" She gave my daughter more goldfish.
My daughter ate the goldfish. The word "more" became a word she used a hundred times a day for the next month, then seemingly a thousand times a day. It was the foundation for the next two years of language work. Every other word grew from that one root.
Words in clumps, then silence, then clumps again
Words didn't arrive in a steady drip. They came in bursts, sometimes a new word a week, sometimes a month of nothing at all.
Roughly in order, there were food words ("milk," "snack," "apple," and "more" extended into "more milk"), movement words ("up," "down," "go," "stop," "again"), people words ("dada," "mama," "grandma," "baby"), and emotion words ("happy," "sad," "tired," "ow"). Question words came much later: "where," then "what." "Why" didn't show up for two more years.
She also started using full sentences from Bluey before she could produce matching sentences of her own. She would say, "Looks like I'm gonna be the doctor today" with Bingo's exact intonation, when she wanted to play. That was gestalt language processing. Real communication. Not the typical developmental sequence, but absolutely functional.
We learned to validate the gestalts, to use them, to slowly help her break them into smaller, more flexible pieces. Marge Blanc's work on Natural Language Acquisition was the framework that finally made sense of what we were seeing.
Plateaus are not failure (but they feel like it)
I want to be honest: there were long plateaus. Months when nothing visible changed. Months when I was convinced we were doing something wrong.
Courtney would say, "Trust the process. The receptive language is growing. The expressive will follow." I trusted her because I had to. Most of the time she was right.
A few times, though, the plateau was a signal that something needed to change. The therapy approach needed to shift, or the sensory environment needed attention first. We added an OT. We reorganized play at home. Things moved again.
Here's the thing about plateaus: they're usually integration phases. The brain is consolidating, and new output follows. But knowing that intellectually and living through week six of silence are two very different experiences.
What actually worked
Looking back, the list is simpler than I expected. Starting early intervention at twenty months instead of waiting until three mattered most, since her brain was more receptive then, and that was the single biggest decision we made. Naturalistic, play-based therapy that followed her lead instead of drilling her made the difference day to day. Courtney taught me the techniques and I did them every day, because the twice-a-week sessions alone weren't enough; the daily reps at home were what compounded.
Reading the same books a thousand times turned her favorite stories into language scaffolding. She memorized them, and the memorized text became raw material for spontaneous speech. (If I never read "Brown Bear, Brown Bear" again, I'll be fine with that.) Once she was old enough, Bluey became a language source too. She pulled functional phrases straight from episodes. For kids who are gestalt language processors, shows with naturalistic dialogue can be genuinely useful, not just screen time.
Beyond all that: patience, not pushing for output and letting it come, and treating every form of her communication, pointing, gesture, sound, or word, as real. It was.
What I got wrong
I waited too long to push back on our pediatrician. The six months I spent trusting "wait and see" were six months without early intervention. If you're in that position, get the evaluation early. Don't wait.
I sometimes drilled when I should have followed. In the first month of home practice, I tried to get her to imitate words, and she shut down completely. Courtney coached me out of it: drill doesn't work for my kid, play does. The instinct to "teach harder" when progress stalls is strong and, in our case, exactly wrong.
I worried more than I needed to. The language came, not on the timeline I'd imagined, but on her timeline. The worry didn't speed anything up.
I sometimes compared her to other kids. That was corrosive for me and would have been harmful to her if I'd said it out loud. Her timeline is her own.
Where we are now
My daughter is six. She has full sentences, questions, jokes, opinions. She has favorite topics she will talk about for an hour if you let her (currently: octopuses and the show Numberblocks). She is autistic. She has language.
She still has things that are harder. Social communication takes more effort, and initiating with peers doesn't come naturally. Some sounds are still in development. The language is rich but uneven, growing wild in some beds and sparse in others. That's fine.
I don't know what her communication will look like at sixteen. I know she'll be okay. I know the work we did, and more importantly the work she did, got her here.
The tools around when my daughter was little just didn't fit her. Drill apps shut her down, flashcards led nowhere, and the clinical materials were built for neurotypical kids working on small articulation goals. What actually worked was what Courtney did on the floor: following her lead, narrating her play, pausing, expanding. There was no app for that, so I built one. Buddy is meant to sit between the human therapy sessions and parent floor time, the thing I wished I'd had when she was two. But the tool isn't the story. The story is her finding her words. The tool was just one piece of how she got there. If your child is going through this, the most important thing you can do is get into the early intervention system if you haven't already. The evaluation is free, the support is free, and earlier is always better. Don't wait for the pediatrician to bring it up. Call yourself. A few questions people ask me a lot: will a child eventually say their first word? Most do, especially with support, though some children don't develop verbal speech and use AAC (augmentative and alternative communication) instead. Both are real communication, and both are valid paths. Why did my daughter regress before she progressed? Regression is one recognized pattern in autistic language development, though the underlying reasons aren't fully understood. Many kids who regress do go on to make real progress with the right support. How long did it take her to go from first word to full sentences? About three years for her, though this varies enormously from kid to kid. Did Bluey actually help? For her, yes. The scripts gave her language she could absorb and reuse in context. This is gestalt language processing, a well-documented phenomenon (Marge Blanc's work is the key reference here). Not every child is a gestalt processor, but for the ones who are, shows with naturalistic family dialogue like Bluey can be a genuine resource. And if I had to name the single most important thing we did, it was starting early. Getting into early intervention at twenty months instead of waiting until thirty was the biggest decision we made, and it's the one I'd push every parent toward. If you want to keep reading, I've written more about this in the autism dad hub, the pillar guide on speech therapy at home for autistic kids, the longer story of building my daughter an AI speech friend, and what being an autism dad has looked like. For more practical guides, there's also advice on short speech practice sessions for toddlers who won't sit still, what ten minutes a day actually did for her speech, benchmarks like whether strangers should understand 75% of a 3-year-old's speech or whether it's normal if a 4-year-old is still hard to understand, and practical AAC questions including getting an AAC device through insurance, a plain-English AAC starter guide (also covered here), and when and how to start AAC with toddlers. There's a full library under parent guides if you want to go deeper.