Michael Devitt develops a hybrid causal theory of non-empty proper names and certain singular terms that semantically resemble them. He considers that the use of a name designates an object not by virtue of the different information we know about it, but by a causal network that starts from the first uses of the name to designate the object, through a “reference borrowing” from the previous uses. (Devitt 1981)
After the initial “naming ceremony”, to use this name, a speaker must already have the ability to use names (“a mental state which is brought about in a language user by perception of a naming ceremony (and in other ways to be described) and which is apt to produce (in part) certain sorts of utterances—utterances using the name in question”). The first uses are causally related to the object, and then transmitted through a causal chain called by Devitt the “d-chain”, d from “designation.” The ability allows the use of the name in an ordinary predication. But, unlike descriptive theories, it is not necessary for a person to have a substantial set of beliefs that involve the name.
The causal chains that are linked together form the causal network for a type of name. There can be multiple d-chains on which a person’s use of a name is based, and a causal network based on its symbol formed from the union of all these individual networks. Devitt introduces the term “name token” which designates the object on which the underlying d-chain it is founded.
Each time we hear a name being used, we must associate it with an ability.
The correct understanding of the use of an ambiguous name is based mainly on the (external) context for the clues, usually represented by a reliable guide in terms of the speaker. The designatum from the context depend a lot on what we already believe, especially about the speaker (what we think the speaker can designate by this name, and what we think he thinks we can designate through him and what he thinks we know about his designation abilities). Another important clue for interpreting a name token is the predicate used with it. In the case of an ambiguous context, we can ask the speaker to tell us his intentions.
The d-chain starts with the perception of the object. The clear cases are those of “face-to-face perception” of the object. The connection between a name token and its object can be mediated by a description. If this connection is to be a d-chain, it is descriptive.
According to Devitt, cross-reference is an example of how a token may depend for its reference on another. In his theory of reference borrowing, an analog type of dependency is the set of most single-term tokens. Also, if an object is selected by an attributive description to its name, then the resulting name will be assigned and the object is not involved in the causal network for this term at the beginning of the network but may be involved later. Otherwise the network becomes grounded in the object and an attributive name becomes designative. The grounding allows a causal theory to explain the change of reference and certain errors and misunderstandings.
A name can be indirectly grounded on its object, on certain types of representations of the object.
Devitt states that a name token designates an object if and only if “underlying the name is a d-chain grounded in the object”. The d-chains consist of three different types of links: “groundings which link the chain to an object, abilities to designate, and communication situations in which abilities are passed on or reinforced (reference borrowings)”. (Devitt 1981, 66) A person may lose the ability to designate an object when they cease to have thoughts that include tokens grounded in the object. An association between description and name consists in preserving the beliefs that the user would express using names and descriptions. But not all thoughts are beliefs.
A problem that may arise is when the causal chains for the term can be grounded in several objects or are not grounded in any object. Or, more abilities, and therefore more networks, can play a role in producing a designation term. Also, misunderstandings can lead to the involvement of several objects in a causal network.
Groundings plays an essential role in Devitt’s theory, representing the final link between all (designational) names and the world.
In the case of a designation change, there are two possibilities: the old name continues to coexist with the new one, or not. This possibility allows a plausible explanation for the gradual transition from one naming convention to another. For a designation change to take place, a network initially grounded in one object must become grounded in another. The object must be of the same more general category.
In the case of several networks, the similarity between them is a matter of degree, resulting in different notions of synonymy that can be defined in terms of similarity. Two networks become more similar if they are grounded in the same object.
Although most names are conceptual, some are attributive, respectively a network is grounded in an object through a description.
d-Chains are based on some notions that remain largely unexplained; the notion of perceiving an object is not sufficiently clear; as in the case of notions of thought, and of the notion of cause.
The basic idea of the Devitt’s causal theory of grounding is that the name is introduced into a formal or informal dubbing, in the presence of the object. (Devitt and Sterelny 1999) Perception must be causal. A witness to the dubbing will have the semantic ability to use the name to designate the object by virtue of the causal link; the perception of the object determined the thoughts that led to the use of the name.
The causal theory of the reference borrowing implies that people who are not at the dubbing acquire the semantic ability from those at the dubbing, which is also a perceptual causal process. The name is used in communication. Hearers can gain the ability to use the name to designate the object by virtue of causal chains that link the object, naming, and user through discussion.
A name has both reference and meaning. The meaning of a name is a particular property of the name, designating its bearer through a certain type of causal link between name and bearer. The aspects of reality that we must call to explain the reference are sufficient for meaning. The reference of a name is determined by the appropriate causal chains, and thus by its meaning.
The theory developed by Devitt shares with the descriptive theories the ability to account for certain special features of natural language: independent stimulus (the causal chain on which its use depends does not require the presence of the object); arbitrary and independent of the environment (any symbol from any environment can be placed in the appropriate causal relation with the object); and so it must be learned. Unlike descriptive theories, this theory can also explain the apparent abstractness of one’s names. Also, according to Devitt, it avoids the problems of descriptive theory, can solve the problem of identity statements, promises an explanation of the final links between language and the world, and the explanation regarding causality seems to be naturalistic.
Devitt concludes that causal theory about names cannot be a “purely causal” theory; it must be a “descriptive-causal” theory, so that a name is associated with a description in a grounding, thus appearing a descriptive element in the characterization of a d-chain. (Devitt and Sterelny 1999, 80)
In setting the reference, a name is ostensibly introduced at a binding event in the presence of an object that will henceforth become the referent.
The ability to refer is a mental state that contains a set of thoughts (beliefs, desires or hopes) that are grounded in the referent and associated with the name.
In the reference borrowing, those individuals present at the initial appointment spread to others the ability to refer through communication.
To refine the notion of fixing references, Devitt uses several groundings, basing a name on an object just like the initial naming.
Devitt’s theory successfully deals with the qua problem, stipulating that the one who grounds names must perceive both the object and the thought of the object under a general correct term. Some degree of matching is required for a name initiator to obtain a name in an object. (Yang 2017)
In the question of the incorrect causal source, Devitt’s theory also provides an intuitive answer.
Both Devitt and Evans believe that Kripke went too far in denying the descriptive elements. Thoughts associated with the name have a reference role. But I agree with Kripke that the name refers by virtue of a causal relationship.
But although Evans believes that the initial baptisms play no role in the reference, generating the qua problem and the incorrect causal source problem, Devitt accepted the idea that appointment is essential, but that there is nothing necessary about the initial appointment. Through multiple groundings, he avoids the problem of changing the reference due to the error, without having to eliminate fixing the reference as an essential component of the name theory. Also, Devitt manages to avoid the problem of the source of incorrect causation because the names can only refer if they are causally related to the object through direct perception, and the problem stipulating that those who ground must have the correct belief of the object under a generic term.
- Devitt, Michael. 1981. Designation. Columbia University Press.
- Devitt, Michael, and Kim Sterelny. 1999. Language and Reality: An Introduction to the Philosophy of Language. MIT Press.
- Yang, Wei Xiao. 2017. “On Causal Theories of Proper Names.” https://journals.lib.sfu.ca/index.php/sfu-ujp/article/view/127.
This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International. To view a copy of this license, visit http://creativecommons.org/licenses/by-nd/4.0/.
Sfetcu, Nicolae, “Causal Theory of Reference of Michael Devitt”, SetThings (12 septembrie 2019), URL = https://www.setthings.com/en/causal-theory-of-reference-of-michael-devitt/