How deaf sleepers woke up before electricity
Not with silence, and not alone — with people. Before any electric device could shake a bed on schedule, waking up on time if you couldn't hear a bell mostly depended on who else was in the house, what light was doing outside your window, and, for a paying few, a stranger with a stick who couldn't actually solve your problem the way he solved everyone else's.
The knocker-upper, in more detail
Britain and Ireland's knocker-uppers are the best-documented piece of this history, and it's worth being specific about what the job actually was: a paid worker, often for decades, walking the same streets before dawn and tapping bedroom windows with a long pole, a rattle, or a pea-shooter until a client stirred and signalled they were up (JSTOR Daily). Named individuals survive in the record — Mary Smith worked 1930s London, Betty McEntee worked County Louth, Ireland in the late 1800s, and Charles Nelson had been doing it in East London for over 25 years by 1929 (JSTOR Daily). It was skilled, relied-upon work: one 1878 account has a knocker-upper in northern England, Mrs Waters, earning more in a week than many mill workers around her (The Vintage News).
Here's the gap worth naming plainly, because none of the sources we could find address it directly: the knocker-upper's whole method assumed a client who could hear a tap on glass. A pea-shooter rattling a window pane, or a pole rapping the frame, is a sound-based signal first and a visible one only if you happen to be looking at the right window at the right second. Surviving accounts of the trade don't discuss how, or whether, knocker-uppers adapted their method for deaf customers — which likely means either that they didn't serve many, or that whatever adaptation existed (a harder tap intended to be felt through the frame, perhaps, or simply being let in to shake someone directly) went undocumented rather than unpractised. We're not going to invent an answer the record doesn't give us.
Family, not hardware
For deaf people specifically, the more consistently documented solution wasn't a device at all — it was another person. A hearing family member, roommate, or household servant physically waking a deaf sleeper by touch was, by every account we could find, the default arrangement for as long as households were multi-person and mornings mattered. It's a solution with an obvious dependency: it requires someone else to already be awake, reliably, before you need to be — which in a working household usually meant whoever rose earliest for their own reasons (fire-lighting, milking, the first shift) took on the job for everyone else by default, deaf or hearing.
Light and orientation
The other consistently documented strategy didn't require another person at all: daylight itself. Historian Sasha Handley's research (summarized by Live Science) notes that early modern households in Britain sometimes oriented beds to catch the sunrise, partly for religious reasons and partly because morning light through a window is itself a strong, reliable cue for waking — one that works exactly as well for a deaf sleeper as a hearing one, since it asks nothing of the ears. It's not a precise alarm; sunrise moves by the season and tells you nothing about a specific hour. But paired with a consistent bedtime and the body's own circadian rhythm, it's a documented, low-tech method that predates any of the mechanical or electrical alternatives below by a wide margin, and it's still recommended today as one of the few wake-up strategies that doesn't depend on hearing at all.
1951: the first patent that mattered
The line from "another person" to "a machine" runs through a specific, verifiable document: U.S. Patent 2,580,598, "Alarm for Deaf Persons," filed by Pasquale Rody of Phoenix, Arizona on 23 January 1951 and granted 1 January 1952 (via Google Patents). It's worth reading the mechanism in full, because it's the direct mechanical ancestor of every bed shaker sold today. The patent describes a vertically mounted motor positioned beside a bed's side rail. Connected to the shaft is a horizontal bar with a resilient roller at each end. Wired to a standard alarm clock's time switch, the motor spins the bar at the set hour, and the rollers strike the bed's side rail — repeatedly, forcibly — carrying the resulting vibration through the frame and into the mattress. Rody's own filing frames it explicitly as an improvement on earlier electromagnetic designs, describing his roller-and-motor approach as more "efficient and positive in operation" and, notably, "commercially feasible" — language that suggests earlier attempts existed and hadn't caught on, though the patent record we could verify doesn't name what those were.