Paulding Light

Local memory offers a cast of candidates. The most popular holds that the light is the lantern of a railroad brakeman killed trying to stop a train from striking cars stalled on the valley's tracks — an explanation that survives the awkward detail that no railroad is recorded in that valley. Other versions nominate a slain mail courier, a Native American dancing on the power lines, or — per Detroit Free Press columnist John Carlisle — a grandparent hunting a lost grandchild with a lantern that needs constant relighting, which would account for the flicker.

The light also answers to the Dog Meadow Light, after the pond on the valley floor, and to the Watersmeet Light, after the nearest town with a gas station. Every variant agrees on the essential point: someone died out there, and the light is the unfinished business. The commerce is more certain than the casualty — the viewing area keeps a steady season, and the nearby outfitters sell their share of sweatshirts to people who drove six hours to look at the dark.

Investigators have been driving to Paulding since the sightings began. An October 1990 effort using telescopic and spectroscopic analysis concluded the lights were automobile headlamps and tail lamps. In 2010, Michigan Technological University students — organized by doctoral candidate Jeremy Bos under the university's SPIE optics chapter — took a telescope to the overlook and resolved the glow into headlights on cars traveling US-45, roughly five miles north.

They ran the correlation test: team member Bill Norkus stationed himself at the highway and logged every passing vehicle while the others logged light appearances, and the match held. Then they manufactured the phenomenon outright, driving the stretch with hazard lights on and watching a flashing yellow ghost appear at the overlook. Faculty advisor Michael Roggemann later described the mechanism — evening temperature inversions over the valley refract the distant lights into a hovering mirage.

None of this has closed the file. Visitors continue to report the light flaring red, splitting, or moving in ways they insist headlights do not; the SyFy series Fact or Faked: Paranormal Files devoted an episode to it, and the argument between skeptics and believers is described locally as a kind of municipal sport. When NPR asked about the debunking in 2019, residents mostly wished the researchers would stop spoiling it.

Bos, now a Michigan Tech professor, attributes the persistence to confirmation bias — people see what they drove out to see. Perhaps. On a clear night the cars still line up at the dead end, the light still rises over the valley, and the brakeman — who never existed, by the record — still gets the credit.