Foundationalism and empirical reason: On the rational significance of observation