Sorry for the stupidity, but what is wrong with or missing from the usual Tarskian semantics for formulas with free variables? Arnon Avron