Reference and Definite Descriptions by K. Donnellan

How do we communicate through the use of definite descriptions? Are there flaws in Russell’s and Strawson’s Theories of Meaning? Here’s what Donnellan thinks.

Introduction

About Donnellan

Keith Donnellan (born in 1931) is an American philosopher who specializes in philosophy of language. His paper ‘Reference and Definite Descriptions’, published in 1966, is considered to be his most important contribution to the field. He is believed to have been influenced by his teacher, Donald Davidson, and by his fellow graduate student, Saul Kripke, as well as by thinkers like W.V. Quine, Ludwig Wittgenstein and J.L. Austin.

Donnellan’s philosophy is also influenced by the broader intellectual context of his time — characterized by the rise of analytic philosophy and the development of formal logic for the purpose of philosophical analysis.

Overview of the text

Donnellan’s theory of meaning is commonly known as the referential/attributive distinction. In his paper Reference and Definite Descriptions, he attempted to challenge the prevailing traditional view of the function of definite descriptions (such as ‘the President of Italy’) as singular terms referring to a particular individual. Instead, he highlights the twofold use of definite descriptions — referentially and attributively.

He emphasized how the meaning and function of a definite description depends on the context in which it is used. He attempted to make clear the distinction between the attributive and referential uses of definite descriptions by pointing out how they have different implications when it comes to determining the truth and falsity of the sentences they are a part of.

Read Bertrand Russell — On Denoting and On Referring by P.F. Strawson to better understand this blog

Attributive and Referential Uses

Donnellan points out that both Russell and Strawson have ignored the distinction between the two uses of definite descriptions. He then goes on to make this distinction clear.

Distinction between Denoting/Attributing and Referring

When a definite description is used attributively, it denotes the object that uniquely fits the description. Something about whoever or whatever is being asserted. If there is no object that fits the description, then no object is denoted by it. For Russell, this was the sole purpose served by definite descriptions.

Donnellan points out that the referential use of descriptions is different from its denotative view. When it is used referentially, the definite description allows the listener to single out the person or object being talked about, and then state something about that particular person or object. Russell had failed to recognize this referential use of descriptions. But Strawson did not. He recognized and accepted the referential use of definite descriptions, but failed to notice how it could be used non-referentially as well (in the case of the same sentence serving both purposes, too).

This can be explained with the following example: “The feminist poet from Amherst might have been an agnostic.” In this sentence, the definite description (the italicized part) is not only fulfilling its referential role (object referred to here is Emily Dickinson) but also serves an attributive purpose (as it is attributing the qualities of being a feminist and a poet from a certain place, to the object).

Note that when a description is being used attributively, the attribute of being so-and-so is of importance, but this is not the case when it is being used referentially.

Assumptions made by Russell and Strawson

Donnellan highlights two ‘common’ assumptions about definite description that were made by Russell and Strawson.

1. Definite descriptions function in an objective manner — they are independent of any particular context.
2. Presupposition or implication that something fits the description — and if something does not, then its truth value is affected. (For Russell — it becomes false, for Strawson — there would be a truth value gap)

Examples given by Donnellan

Consider the following example: “Smith’s murderer is insane.” Suppose Smith was considered to be an extremely likeable person, which is why the person who killed him is thought of as being insane. This is the attributive use of the description. In another situation, consider that Jones has been charged with Smith’s murder. And he happens to exhibit peculiar behavior in the courtroom. Here, if one states that “Smith’s murderer is insane”, they are making a referential use of the definite description (Smith’s murderer) to point/refer to Jones.

There is one common assumption in both of these uses — that Smith had been murdered. But if it turns out that he had committed suicide, instead, the above mentioned uses would come into question. To resolve issues like this, Donnellan points out how uses other than assertions can be facilitated through language.

He clarifies this with the help of another example. Suppose these is a person in a party who is holding a martini glass. If a person asks — “Who is this person drinking a martini?”, if we are recognizing the attributive use of the description, we would relate ‘the man’ with ‘drinking a martini’. But if we are only concerned with the referential use of the sentence, we would only focus on ‘the man’ drinking a martini.

Here, Donnellan says — what if there is water in the glass, and not martini? The question about the identity of the man would seem to become invalid. Because if no one is drinking a martini, there would be no direct answer to the question — nothing would satisfy the description for the attributive use and nobody could be referred to for referential use. Donnellan emphasizes how the presupposition that both Russell and Strawson made — that someone or something fits the description — is a mistake.

Criticisms Against Russell’s and Strawson’s Theories

Donnellan clarifies his proposed distinction between definite descriptions’ attributive and referential uses by explaining how they apply on the examples given by him.

Russell’s theory, Donnellan states, applies only to the attributive use of descriptions. He completely overlooked the referential use.

In case of the Smith’s Murderer example, Donnellan points out that the definite description can be used attributively even when the speaker has a particular individual (Jones) in mind, who fits the description. Similarly, the description can be used referentially even when the speaker does not have a particular individual in mind. This is allowed by Donnellan’s theory, but not by Russell’s or Strawson’s.

In the case that nothing fits a description, Russell and Strawson contend that the truth value of the sentence is affected. For Russell, such a sentence (take the Martini Man example) would become false. For Strawson, it would be neither true nor false. But Donnellan believes for it to have the possibility of being true. He says that the speaker may successfully refer to something through a definite description even if it does not correctly describe the person/object it is referring to. For example, in the case of the man holding a martini glass (which actually contains water), the purpose of referring is solved, even though the description is incorrect.

Additionally, he points to the difference between Russell’s usage of the term ‘denoting’ and ‘referring’. A definite description can be said to denote an entity if the said entity satisfies the description uniquely. One can use such an expression without being aware if there is indeed something that is being denoted. But one cannot refer to something without being aware of the existence of the object/person in question. For Russell, denoting is seen as a purely semantic or referential relation between words and things, but Donnellan emphasizes upon his distinction to highlight the importance of the speaker’s intentions for determining the meaning and reference of definite descriptions.

Criticism of Donnellan’s Distinction

While Donnellan’s distinction has been influential, it comes with its own challenges and drawbacks. Some of them can be explained as below:

1. Ambiguity: At times, it may be unclear if the definite description is being used attributively and referentially. In such cases, it becomes difficult to determine which interpretation is intended/correct.
2. Internationalism: The heavy dependence of Donnellan’s theory on speaker’s intention has been called problematic on two grounds — first is the difficulty in determining the speaker’s intentions, second is the circularity in determining meaning that arises from such a focus (because speaker’s intentions depend on the meanings of the words used).
3. Exclusivity: The distinction might be seen as being exclusive as it allows for only two possible uses of definite descriptions (referentially and attributively).
4. Lack of Empirical Support: Because of its seemingly scarce empirical support, some have argued that Donnellan’s distinction is more of a theoretical construct than it is a reflection of actual language use.

Takeaway and Conclusion

Donnellan believes that both Russell and Strawson fail to present a correct theory of use of definite descriptions. Russell ignores the referential use, and replaces ‘referring’ with ‘denoting’ (which is a mistake, as explained by Donnellan). Strawson, on the other hand, mixed up the two (attributive and referential) uses of definite descriptions as he overlooked the difference between them.

Donnellan stresses upon the importance of the context, or the speaker’s intention in the chosen use of definite description (and not just the object that may or may not fit the description), while understanding the function of language and of definite descriptions.

Donnellan’s distinction is a significant development in the philosophy of language as it challenges traditional views and points to the importance of context while grasping meaning. It has implications for debates regarding reference, nature of meaning and identity. Even though it faces challenges and criticisms, it remains an important contribution to the field.