An Assessment of Russell's Theory of Definite Descriptions

A definite description is a term that specifies a particular object or person. It refers to something specific, such as “The Prime Minister of Canada,” or “my best friend’s dog,” or even pointing to someone and saying, “the guy standing over there.” Definite description may create problems when assigning truth value to the sentence. For example, the following sentence is a paradox.

(1) The current King of France is bald.

In (1), ‘The current King of France’ is a definite description as it specifies the present King of France. However, France is a democratic nation with elected officials, not a monarchy, thus the current King of France does not exist . Assigning a truth value to (1), would lead to a belief about the status of the current king of France’s hair situation, not the present existence of the King of France.

Russell assesses this classical paradox. Russell’s theory states that the truth value of any sentence using definite descriptions could be properly calculated if it were broken down into its logical form. In this case, (1) is actually asserting three propositions simultaneously. The proper logical form of (1) is:
(1R) There is an x such that x is a current King of France and x is unique in that nothing else is the current King of France and x is bald.

Once broken into logical form in (1R), determining truth value becomes easier. In order for (1) to be true, all three assertions must be true. If one of the statements is false, then (1R) is false. In this case, there is not an x such that x is the current King of France, therefore (1) is false.

However, Russell’s theory of definite descriptions has been criticized. Donnellon refutes Russell’s theory, claiming that it allows no concession for referential use of definite descriptions. Russell’s theory accounts solely for the attributive usage of definite descriptions and therefore is not universal. For instance, consider the following hypothetical situation. Suppose there were a man named Doug who never completed grade 8 and therefore was unaware of the French of Revolution. Doug believes that the current political system of France is a monarchy, of which there is a king. If Doug were touring a castle in France, and there was a bald man dressed as a king, and upon viewing this man, Doug asserted, “The current King of France is bald!” Doug would be using the definite description referentially, not attributatively. The sentence is identical to (1), which we have assessed using Russell’s theory. In this case, Doug did not mean to say, (1R) “There is an x, such that x is the current King of France, and x is unique and x is bald.” What Doug meant is, “That kingly-dressed fellow over there is bald!” Donnellon shows that truth value can come apart between what the proposition of the sentence is communicating and what the speaker is communicating.

Kripke criticizes of Donnellon’s critique, observing that Donnellon made an excellent distinction between attributive use and referential use of definite descriptions. Donnellon failed to point out that Russell’s theory of referential use of definite descriptions is incorrect. According to Kripke, Donnellon provided a distinction between two concepts, not a counterexample to the thesis.

Kripke presents a thought experiment in which we imagine a language that is identical to English, except the sentences would have to be formed with Russell’s logical breakdown. In this thought experiment, Kripke stipulates the truth conditions of what the speaker wishes to communicate. If this Russellian language can also have a contrast between referential and attributive use, then Donnellon has not presented a valid counterexample to Russell’s theory. Let’s expand the earlier hypothetical situation involving Doug. If Doug were using Russellian language, he would say:

(2) There is an x such that x is the current King of France, and x is unique in that no y is x, and x is bald.

In this case, the Russellian language could still have the discrepancy between the attributive use of the conditions and the referential use. In this hypothetical case, Doug intends to assert that the fellow he sees dressed as a king is bald. By stipulating the truth conditions of the speaker’s assertion, Kripke finds the problem with referential use in this Russellian language exists. Therefore, Kripke notes that Donnellon makes an excellent point in distinguishing between referential and attributive uses of definite descriptions, but fails to show that Russell’s theory is wrong.

Kripke’s thought experiment is a strong argument only under the assumption that it is possible to stipulate the truth conditions in which speakers intend to use them. Donnellon might argue that this stipulation is impractical and impossible. The speaker’s intention belongs solely to the speaker, the intention of the speaker cannot be asserted into Russellian language because the only person privy to the intentions of the speaker is the speaker. Even if the speaker was questioned as to their intention, the speaker might have been lying. Donnellon could argue that Kripke cannot stipulate the truth conditions and Kripke’s thought experiment is not practical.