The Truth of Performative Utterances

Foreword

This is my second, and arguably best, paper from Philosophy 516: Language and Meaning. It deals with J. L. Austin and his "performative utterances". Austin argues that these utterances have no truth value, however I provide arguments on why this shouldn't be the case, and provide an alternative view based on fidelity conditions that seems quite good.

Body

In “How to Do Things”, J. L. Austin promotes a distinction between sentences that simply “describe a state of affairs” and those that do not. For example, “All men are mortal.” makes a descriptive claim about the way the world is, regardless of whether it can be verified or not. On the other hand, there are some utterances that Austin says cannot be true or false. Of those that cannot, Austin defines a ‘performative sentence.’ For example:

1. “I do”
2. “I name this ship the Queen Elizabeth”
3. “I give and bequeath my watch to my brother”

What is noteworthy about these types of sentences is that performing the utterance is the same as, or is part of, doing the action. Another way to think of this is by considering a regular declarative sentence: “It is 35 degrees in Madison right now.” This statement is either true or false, regardless of whether or not someone uttered those words. But, a statement like “I name…” seems to not be able to be evaluated unless those words are being spoken. This is what a performative sentence is; uttering is doing.

Austin claims that these sentences do not hold truth or false values on intuitive grounds. An initial objection would be that this reduces complicated concepts, such as marriage or betting, to mere words. But, Austin acknowledges that it is rarely the case that the utterance is sufficient for the action, or even necessary. Rather, along with other conditions, an utterance is a leading incident of some act.

Next, consider a statement such as “I promise…” Such an utterance seems to require a speaker to be ‘genuine.’ Someone who makes a promise in a sarcastic tone does not seem to be making a genuine promise. This leads one to wonder if the utterance is simply an outward description of the inward performance. Austin is not a fan of this viewpoint and argues that such a promise wasn’t false, but rather void or given in bad faith. This is an important distinction. Imagine I make some utterance “I promise you that I will drive you home.” while secretly believing that I will not drive you home. Austin argues that this provides a loophole for the promiser. It is not the case that this person did not promise, they did promise, rather it was a promise given in bad faith. Austin further argues his point by considering the American legal system. He points out that performative utterances are taken in as evidence as the action they perform.

Austin provides a set of conditions for performative utterances to work well:

1. there must exist a conventional procedure having a certain effect, including the uttering of certain words by certain persons, and the particular persons and circumstances in a given case must be appropriate according to this procedure;
2. The procedure must be executed by the participants correctly;
3. Where the procedure is intended for persons with certain thoughts and feelings, the persons involved should have these thoughts or feelings.

The third condition is not necessary for an utterance to have occurred, rather it is better understood as a maxim of conversation; things are easier if (3) is true. The other two are necessary for an utterance to be performative. It is clear that any random utterance of “I do” does not make someone married, just as saying “I bet” after a horserace outcome is determined does not mean you placed a bet.

It seems like (3) does not need to be a separate condition. If such internal feelings are necessary for the conventional procedure, then they are simply absorbed by (1). If, on the other hand, they are not required for a performative utterance, Austin is just giving us a maxim that makes ordinary language flow better. Though he acknowledges this, there is no reason to pick that maxim over the plentiful list of others.

I propose that for a particular utterance, we can write specific fidelity conditions that are more detailed than the overarching definitions Austin gives. For a marriage, we may require that an officiant be there to witness the declaration of intent, and two individuals to say it to each other. Ignoring the variety of cultural practices between marriages, it seems possible to write some list of things that should be true for an utterance such as “I do” to be valid, past the conditions Austin gives. Some utterances may require no conditions: such as something like “I say…”

Some issues with Austin’s theory present themselves when considering the negation of performative utterances. Consider “I do.” and “I do not.” both concerning marriage. The first, as Austin understands it, is performing some action. Consider, though, the second. What action could the second utterance possibly be? Not getting married? Consider any other case in which we use “I do.” “Do you want ice cream?” “I do.” “Do you want to go biking?” “I do.” These examples are not of you acting, but rather affirming the question; describing a way the world is (viz. The opinions you hold). This is seemingly what happens in a wedding. Whether it is explicitly stated or not, “I do” is a response to “Do you take this person to be your lawfully wedded partner?” In this sense, saying “I do” is simply an affirmation of the question. What then, could the difference be between “I do” in marriage and in commonplace questions? A clear difference is that saying “I do” changes the state of the speaker (from unmarried to married). The issue here is that the statement “I do not” does not change the state of the speaker: they remain unmarried. Rather they are just reporting an internal state in response to the question. In this case, it seems that there is no meaningful difference between “I do not” (in response to an officiant), and “I do” (in response to common questions). They are rather declarative responses.

The issue that this presents is that now the negation of “I do” (in response to an officiant) becomes a completely different type of thing. Having “I do not” be evaluable to a truth value, and Not(“I do not”) be completely undefined should be a violation of some logical law, and a conclusion that we ought to avoid.

Despite Austin’s distaste for evaluating such statements as truth or false values, it warrants investigation. I propose a theory in which we define the truth of a performative statement as how it follows a particular utterance's fidelity conditions. Let us define a performative utterance as true if the speaker follows all of the fidelity conditions, and false if they violate any.

Let us again investigate why Austin doesn’t like the viewpoint that performative utterances can still be evaluated as true or false, because of the issue with promising. Making a bad-faith promise will depend on your internal beliefs. I intend to resolve this issue with the idea of fidelity conditions. Under our theory, “I promise..” has some list of fidelity conditions that, so long as they are followed, means that the person who uttered those words performed that action Assume initially, that there is nothing more to making a promise than uttering “I promise…”; there are no fidelity conditions to satisfy. Thus, regardless of your intentions to keep the promise, your utterance makes you hold a promise. This is exactly the conclusion Austin wishes us to reach, now any person who does not follow through on their promise did indeed make such a promise, and any intention otherwise makes their promise void. Now, every utterance of “I promise…” is true, as the utterance makes it the case that they promise. This seems to be the type of promise that Austin is interested in, the only condition he is worried with is performing the utterance. Sure, other conditions are useful for ordinary language, but not necessary for the utterance to be an action. Thus, we have reached a theory that aligns with Austin’s previous discussion, follows some set of fidelity conditions, and can be evaluated as true or false. Someone who prefers more conditions than just an utterance with a promise, such as sincerity, would just add them as more fidelity conditions. Thus, any utterance of “I promise…” under such a theory is true if it is sincere, and false if it is not.

If promising no longer provides an issue to this theory, we have run out of counterexamples. We have defined a system that allows us to have performative utterances be evaluable, without any clear issues. This is preferable to a system that does not do so, as the evaluation of sentences is a useful characteristic. Argument’s against this on simply intuitive grounds are not strong enough.