When words split on the road: Recognition costs of across-line destination-name order in pavement markings

Destination names are often painted directly onto road surfaces. When lane width is limited, these names are typically split across two lines, creating a design choice regarding their vertical order. While current models of reading propose that preserving canonical order (Top-to-Bottom) facilitates word recognition, real-world implementations often violate this principle. In two go/no-go lexical decision experiments, we examined whether vertical ordering of split destination names influences recognition efficiency across two display settings. Spanish city names were presented as split markings in either canonical Top-to-Bottom order (e.g., BARCE above LONA) or inverted Bottom-to-Top order (e.g., LONA above BARCE); in the inverted arrangement, the first fragment appeared on the lower (nearer) line. Experiment 1 used a neutral asphalt texture, whereas Experiment 2 embedded the markings in a realistic road scene. Bayesian mixed-effects analyses showed clear evidence for an advantage of canonical ordering, with response times substantially faster for Top-to-Bottom than for inverted layouts (164 ms in Experiment 1; 211 ms in Experiment 2). From a traffic and road-design perspective, preserving canonical reading order when splitting destination names appears to be a simple, low-cost strategy to reduce processing demands and increase effective preview time for lane-choice decisions.