Philosophy Things

De Dicto, De Re and Frege's Account of Intentionality

Frege’s account of intentionality is an example of purely conceptual, de dicto belief. Contrary to a de dicto belief is a de re belief, one that is not purely conceptual, but rather, part conceptual and part contextual. This notion of de re belief is one held by Keith Donnelan, Tyler Burge and John McDowell, whom have each presented their own accounts of how intentionality. In the end, McDowell presents a farfetched idea of a de re sense while claiming to stick to Frege’s de dicto account of intentionality.

Frege’s account of intentionality illustrates how it is that a thought or sentence can come to be about something in the world. Frege argues that linguistic items express senses, and sentences express thoughts. Both senses and thoughts are abstract entities that connect or mediate one’s language to the world. Each sense is comprised of a set of conditions that uniquely characterizes and points out that referent. Additionally, senses on their own are not complete and therefore cannot be assessed as either being true or false. Thoughts, on the other hand, are complete and do have truth-value.

In his essay, “Proper Names and Identifying Descriptions,” Keith Donnellan presents an attack on how a proper name refers. He calls this method the “principle of identifying descriptions” (POID) and claims that it is a two part thesis: 1) that the user of a proper name must be able to supply a set of definite descriptions that uniquely point out the referent, and 2) that the referent of a proper name (if there is one) is that object which uniquely fits a “sufficient” number of definite descriptions in the identifying description (definite descriptions are what articulate identifying descriptions). This principle is similar to Frege’s account of intentionality in that it is grounded in mediated reference.

Donnellan rejects this principle and provides four examples, two in particular that he feels justify that rejection. In the first example, Donnellan presents a plausible case in which reference occurs while the speaker is unable to provide any identifying descriptions: A young child is sleeping at a home while his parents are throwing a party. The child wakes up and meets a strange man, who tells him, “You have a nice set of parents” and then walks away. It is not until a couple weeks later that the child tells his parents, “Tom is a nice man.” But because the parents throw many parties and know many men named “Tom,” they are unable to distinguish whom the child is referring to, yet still it appears that the child is actually referring. Donnellan claims that this is a violation of the first part of the POID.

In a second example, Donnellan presents a plausible case in which the definite descriptions don’t point out a unique referent. For example, a student attends a party and believes he has met the famous philosopher, J.L Aston-Martin, who actually is just a look-alike. The student talks with the man for a while, and then as he is leaving, J.L. Aston-Martin (the look-alike) trips over his foot. The next day, the student tells his friends how he met J.L. Aston-Martin (referring to the famous philosopher), and then tells them how he tripped over his foot (referring to the look-alike at the party). The problem with this counter-example is that it violates the second part of the POID. Since the student’s definite descriptions do not accurately characterize the proper name, it seems that the descriptions are not properly referring.

Both of Donnellan’s counter-examples to the POID cast doubt on Frege’s account of intentionality because on Frege’s view, due to the reliance on mediated reference, one is committed to the POID. If Donnellan’s criticisms are accurate, then the child at the party case shows that definite descriptions are not needed to point something out uniquely, and the J.L. Aston-Martin case shows that definite descriptions don’t always point to something particular.

Two general methods of belief are “belief de dicto” and “belief de re.” Belief de dicto is a form of indirect reference that is entirely conceptual. It relates a believer to a complete thought: a complete thought in the sense that the representation is either entirely true or entirely false. Belief de re, on the other hand, is a form of direct reference that is partly conceptual and partly contextual. It relates a believer to a general, partial characterization and to a thing. For example, “Orcutt believes of someone in particular that he is a spy” represents a de re belief. The “of someone in particular” provides the contextual part, and the “a spy” provides the conceptual part.

In his essay, “Belief De Re,” Tyler Burge argues that belief de re is the more fundamental method than belief de dicto. He begins his argument by showing how indexicals are the linguistic phenomena of a sentence. For example, “This is a marker” is a de re belief, because the word “this” is an indexical that’s used to supply the contextual part, and the words “is a marker” are used to supply the conceptual part. This account is completely inconsistent with the traditional understanding of Frege’s view because, under Frege’s view, reference is made by purely conceptual, de dicto method. Because de dicto beliefs don’t reduce to de re beliefs, it is not possible to hold both Burge’s and Frege’s views while maintaining compatibility.

John McDowell, in his essay “De Re Senses,” proposed the idea of an essentially de re thought. This idea rests on the following two tenants: 1) that it is about something in particular, but not uniquely characterizing, and 2) that the sense does not exist if what it is about fails to exist. The child at the party and the J.L Aston-Martin cases are examples that employ such de re senses. Many find the existence of these entities that McDowell is proposing to be farfetched, mostly because of basic Fregean thought that in order to think about something in the world, one must be able to think about properties that point out that thing. But Burge criticizes Frege for this method and uses Donnellan’s cases to show that Frege’s view needs refinement if it is to account for such scenarios as the child at the party and J.L. Aston-Martin.

If there were such a thing as de re senses, they would serve to strengthen Frege’s traditional view of intentionality. Since Donnellan has pointed out that reference does not always work by specification, de re senses would help Frege include Donnellan’s scenarios under his theory and ultimately defend it against future criticisms.