I can only guess, because I don't have that particular wikipedia page, and I likewise don't have the references that page used, but apparently, the spectrophotometric method is using some sort of indicator, and is determining pK(solvent), to determine pH(solvent).
As I'd heard it, if you wanted a "correct"* value for pH, in a non aqueous solvent, say, 10% ethanol in water, what you'd do is fill the pH electrode with 10% ethanol (and the correct ions, usually conc. KCl saturated with AgCl, for a silver electrode), and standardize it with pH buffers, likewise 10% ethanol, with their concentrations calculated for the lower activity of the solvent. Then you can use it for an unknown sample.
*Disclaimer: this is of course not the definition of pH, the activity of hydrogen in pure water, but instead of hydrogen activity.
**Super Disclaimer: I have, often, measured and reported the pH of an HPLC mobile phase, that was not pure aqueous, with a pH meter meant only for aquesous solutions. I know the answer was technically wrong. But it was reproducible, and gave expected results, so sometimes people can just "go with the flow," depending on the rigor of their application.