Text-to-911 is live in the majority of PSAPs nationwide. The infrastructure is deployed. The capability exists. What doesn't exist, in most centers, is any structured training for how to dispatch from a text conversation instead of a voice call.
That gap matters more than most people think, because text and voice are not interchangeable communication channels. They transmit fundamentally different information, and a dispatcher trained on voice calls is losing critical data when they process a text-to-911 session without understanding what's missing.
What text strips away
On a voice 911 call, the dispatcher is receiving at least three layers of information simultaneously. The words - what the caller is actually saying. The paralinguistic cues - how they're saying it, including tone, pitch, pace, volume, breathing patterns. And the background audio - what's happening in the environment behind the caller's voice.
These three layers combine to produce the dispatcher's situational assessment. "Someone's hurt" spoken calmly is a different call than "someone's hurt" screamed at full volume with crashing sounds in the background. The words are identical. The dispatch is different. The experienced dispatcher processes all three layers automatically and produces a resource decision that reflects the totality of what they heard.
Text gives you one layer. The words. No tone. No pace. No breathing. No background. "Someone is hurt" arrives as characters on a screen with no indication of whether the sender is calm, panicked, injured themselves, or even still present at the scene. The dispatcher has to make the same resource decision with two-thirds of the information removed.
The speed problem
A voice call processes in real time. The dispatcher asks a question, the caller answers, the information arrives in seconds. A text conversation has latency built into every exchange. The caller has to read the question, type a response, send it, and wait for the next question. Each round trip might take 30-60 seconds compared to 3-5 seconds on a voice call.
For a dispatcher used to voice call pacing, this latency is excruciating. The instinct is to send multiple questions at once to speed things up, which overwhelms the caller and produces garbled responses. Or to fill the silence by dispatching based on incomplete information rather than waiting for the next text, which is the text equivalent of dispatching too fast on a voice call.
Neither instinct is wrong in a voice environment. Both produce problems in a text environment. And nobody trained the dispatcher on the difference.
Training for the medium, not just the technology
Most text-to-911 training covers the technology: how to receive a text, how to respond, how to transfer, what the interface looks like. That's necessary. It's also insufficient.
What dispatchers need is training on the communication dynamics of the medium. How to compensate for the absence of tone. How to manage the latency without either over-querying or under-querying. How to recognize urgency when the only signal is word choice and response speed. How to structure questions for text that work differently than the scripted questions designed for voice.
The caller who stops responding to texts might be in danger, might have lost service, might be hiding and unable to type, or might have gotten distracted. On a voice call, the open line gives you background audio to differentiate. On text, you get silence that means everything and nothing simultaneously.
The technology is deployed. The training for the human on the other end of it hasn't caught up.