I’ll quote IIa IIae, 3, 8, 2, and Thomas’s answer to that objection:
Praeterea, visio est certior auditu. Sed fides est ex auditu, ut dicitur ad Rom. X, in intellectu autem et scientia et sapientia includitur quaedam intellectualis visio. Ergo certior est scientia vel intellectus quam fides.
Ad secundum dicendum quod, ceteris paribus, visio est certior auditu. Sed si ille a quo auditur multum excedit visum videntis, sic certior est auditus quam visus. Sicut aliquis parvae scientiae magis certificatur de eo quod audit ab aliquo scientissimo quam de eo quod sibi secundum suam rationem videtur. Et multo magis homo certior est de eo quod audit a Deo, qui falli non potest, quam de eo quod videt propria ratione, quae falli potest.
The objection is this: according to Romans 10, “faith is by hearing,” while “knowledge and wisdom” depend on “a certain intellectual vision.” “But vision is more certain than hearing,” so knowledge must be more certain than faith.
Thomas solves the problem by acknowledging that vision is the more certain, “other things being equal,” but that the relative certain in any intellectual act always depends on the provenance of the knowledge: since in faith we hear God, who cannot deceive, we can be more certain in it than we can in our own knowledge, whose vision is all too easily deceived.
But might it not be simpler to subsume all our knowledge of God under a single heading? Surely any knowledge we attain about God can only be the result of his prior speaking to us — even knowledge of God come from attending to the intricacy of creation is the result of speech, according to the Psalmist: “The heavens declare the glory of the Lord.” Why not rather say that God speaks to us faintly through his creation, and more clearly through his Word come to voice in Israel, and then in Christ?