“The Thought: A Logical Inquiry”. Gottlob Frege. Mind, New Series, Vol. 65, No. 259. (Jul., 1956), pp. 289-311 (‘Der Gedanke’ is the untranslated name of the same work)
Frege explores the cognitive phenomenon of taking something to be true. His central claim is that to take something as true is to enter into a relation with an abstract entity called a ‘Thought,’ which to Frege is a specific sort of meaning, expressible through sentences, which may be either true or false.* In the midst of formulating this answer, Frege clarifies what he means by ‘true’, explains what he thinks are some basic properties of Thoughts, and engages in a discussion of the precise ontological status of Thoughts.
Frege begins by clarifying what he means by ‘truth,’ since without a more clear notion of this term his thesis about what it is to take something as true does not have meaning. Frege eschews several commonsense uses of the word before stating that the sort of truth he wishes to discuss is that which is sought out by the sciences (326).
In an attempt to explicate the notion further, Frege gives a tentative catalog of things truth – which for the sake of discussion he assumes is a property – may be predicated of: pictures, ideas, sentences, and Thoughts. Frege rejects the claim that truth may genuinely be predicated of pictures and ideas, for he thinks such predication requires a correspondence theory of truth, or, a theory which states that truth consists in some correspondence between a picture or idea and item in the external world. Frege presents a convergent argument against all such theories (a cluster of independently reinforcing points): one, such theories go against the use of the word ‘true’ as they require a relation between two things, something the word typically does not assert; two, if two items corresponded perfectly, they would be identical, and this is not what a person predicating truth wishes to say; and, three, if one wishes to specify what sort of correspondence truth consists in, it may always be sensibly asked, “Is this definition true?” The coherence of the question for all such definitions indicates to Frege that none of them are capturing the essence of truth, because otherwise the question would be incoherent. These results lead Frege to conclude that truth is indefinable. (326-327)
Frege then asserts that, while we often speak of sentences as being true or false, what this talk actually consists in is ascribing truth or falsehood to the senses (contents) of such sentences. And the specific sorts of senses which may be sensibly ascribed truth or falsehood are, to Frege, the Thoughts. Frege thinks as well that the ‘is true’ predicate does not add any content to a sentence.
Frege then describes several of Thought’s basic properties in an attempt to reveal their nature more clearly. Thoughts are, to Frege, imperceptible: none of our senses ever interacts with a Thought. Frege uses the example of a specific sensed phenomenon: while the Sun rising may be sensed, that the Sun is rising is a Thought with a truth value and is never sensed, instead being grasped by some other means.
Frege argues that Thoughts may be expressed without thereby being asserted: Thoughts are expressed by propositional questions (which are not assertoric) as well as interrogative sentences (which are assertoric), indicating that the assertion of a Thought is a separate issue from its truth value (e.g., one can say something true and yet not assert it) – a difference Frege thinks can be explained by sentence-forms and the conventions surrounding their use. Thoughts and their associated truth values exist independently of use.
The final properties Frege discusses are the under- and over-determination of Thoughts by sentential content. Thoughts may both be expressed in sentences with more content than is needed to express the Thought, or not expressed at all due to a sentence lacking certain features. Regarding the former case, Frege cites expressive and poetic words as not assisting in the expression of Thoughts; logically, such words are extraneous, whatever their function in everyday language use. Frege uses the word ‘there’ to explain underdetermination: if a sentence uses the word ‘there’ along with an accompanying demonstration (e.g., a pointed finger), then we are not grasping a genuine Thought by looking at the sentence alone; certain extralinguistic facts must be known as well (in this example, we must know where the finger is pointing). (332)
Frege concludes his exploration with a discussion of the ontological status of Thoughts . As already mentioned, Frege does not think Thoughts are external, sensible objects: truth attaches to Thoughts, but not to sensed objects, and so they cannot be the same. Frege next considers the claim that Thoughts are ‘Ideas’, a term he uses to refer to the internal items of a person’s mental life, viz., sensations, desires, intentions, and so forth. Frege thinks Thoughts cannot be Ideas, for Ideas have specific properties that Thoughts do not: they are possessed by persons, and they are a constituent of a person’s consciousness. If Thoughts were mere items of a given person’s consciousness, they and their truth would be relativized to that person, for it is impossible to share bits of one’s consciousness with another. Yet Frege takes it as obvious that Thoughts – he here uses the example of the Pythagorean Theorem – are mutually grasped entities the truth of which has nothing to do with any given person’s consciousness. (336) Since identical things will have all the same properties, and Frege has just found that any given Thought and Idea will have divergent properties, it follows that Thoughts are not Ideas.
Frege disarms the skeptical claim that, for all we know, Ideas are all that exist. Frege levels two arguments against the claim: one, at least one independent object is needed to best explain our experiences. To Frege, selves – conscious entities that possess Ideas – cannot be explained in terms of Ideas. He finds it is far more reasonable to posit the self as an independent object than it is to attempt an explanation of selves as specific portions of conscious content. The second argument is pragmatic: to accept the claim that only Ideas exist is to give up on all substantive inquiry, all of which assumes the existence of external objects in order to be meaningful. This is an unacceptable outcome and thus skepticism is to be rejected.
Thus, Thoughts are the bearers of truth, but do not exist as external objects nor as Ideas. Given that there can be independent objects – a claim which follows from Frege’s refutation of skepticism – the only option left is to posit a “third realm” where Thoughts exist. (337) This realm is outside of time and space, although its constituents are ‘graspable’. This grasping in turn leads us to action; Thoughts as such have an indirect causal impact on the world. Through explaining what it is to treat something as true, Frege has discovered what he takes to be the nature of thinking more generally.
*Frege’s ‘Thoughts’ are what we’d now call propositions, though their exact nature — including whether they exist or not — is still a matter of debate.