Beardsley on literature, fiction, and nonfiction

Beardsley on literature, fiction, and nonfiction

Szu-Yen Lin*

Philosophy, School of Humanities, University of Auckland, Auckland, New Zealand


This paper attempts to revive interest in the speech act theory of literature by looking into Monroe C. Beardsley’s account in particular. Beardsley’s view in this respect has received, surprisingly, less attention than deserved. I first offer a reconstruction of Beardsley’s account and then use it to correct some notable misconceptions. Next, I show that the reformulation reveals a hitherto unnoticed discrepancy in Beardsley’s position and that this can be explained away by a weak version of intentionalism that Beardsley himself actually tolerates. Finally, I assess the real difficulty of Beardsley’s theory and its relevance today.


Szu-Yen Lin is a doctoral candidate in philosophy at the University of Auckland. He received his MA in philosophy from National Chung Cheng University in Taiwan. His research areas include aesthetics and philosophy of the arts, in particular the philosophy of literature. He is currently working on the problem of literary interpretation. Apart from being a philosophy researcher, Szu-Yen is also a mystery writer, having published 8 novels and 28 short stories in Chinese, one of which also appeared in Ellery Queen’s Mystery Magazine.

Keywords: illocutionary act; representation; intentional fallacy; intention; interpretation

Citation: Journal of Aesthetics & Culture, Vol. 8, 2016

Copyright: ©2016 S.-Y. Lin. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License, allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.

Published: 18 March 2016

*Correspondence to: Szu-Yen Lin, Philosophy, School of Humanities, University of Auckland, Arts 2 Building, 18 Symonds Street, Auckland 1010, New Zealand, Email:


Speech act theory was originally proposed and developed by J. L. Austin to explain deeds performed by utterances in particular contexts.1 His focus is mainly on illocutionary acts, that is, speech acts performed in using language. The operational schema can be described at the most general level: an illocutionary act is generated by intentional text production under appropriate conditions, according to certain language conventions.

The influence of Austin’s work has gone beyond the philosophy of language. Many literary theorists and philosophers have applied this machinery to literary studies. One notable attempt is the effort to define literature based on Austin’s remark that an illocution loses its force in artistic contexts. Among others, Richard Ohmann distinguishes literary texts from non-literary ones by defining the former as mimetic illocutions.2 This model is adapted by Charles Altieri to discuss poetry.3 Along similar lines, John Searle argues that fictional utterances, in contrast to ordinary ones, are pretended speech acts.4 An opposing view is given by Mary Louise Pratt, who tries to blur that distinction, showing that literary and non-literary discourses can be analyzed on the same model.5 The said distinction is again challenged and criticized by Stanley Fish.6

The application of speech act theory to literature7 has faced many challenges, both from philosophers and literary critics. Joseph Margolis rejects speech act theory itself and its literary applications, claiming the whole project is doomed at the very start8; Roger D. Sell criticizes the theory’s leading to depersonalized and decontextualized readings of literary works, advocating a historical literary pragmatics that treats writing as an act of communication between real authors and real readers.9 Though the heyday of the literary deployment of speech act theory has gone, there are occasional efforts to revive its interest. A notable example is by J. Hillis Miller, who attempts a speech act reading of literary works by Proust and Henry James.10

Beardsley proposed his speech act theory of literature when such deployment became a trend. Similar to Ohman and Searle, he espouses a formalist and anti-pragmatist position on reading literature, holding that fictional works are illocutionary-act representations. This construal of literature is in one sense dated, as opposed to the more recent trend of literary pragmatics, and this may be part of the reason why Beardsley’s work in this respect has received scarce attention. But what is often neglected is that Beardsley’s speech act theory of literature is strongly linked to his early and more influential proposal: the intentional fallacy. His literary applications of speech act theory are deeply connected to the intentional fallacy and should be understood in that context. If we think that the intentional fallacy is still relevant today, that the problem of intention and interpretation is still a thought-provoking problem to be resolved, then reconsidering the Beardsleyan speech act theory may prove helpful and enlightening for both the philosophy of literature and literary studies.

In what follows I aim (1) to offer a reconstruction of Beardsley’s account—even such a reconstruction is hard to find in the literature; (2) to use this reconstruction to correct some misconceptions; (3) to show that the reconstruction reveals a hitherto unnoticed discrepancy in Beardsley’s view, and that this can be explained away by a weak version of intentionalism that Beardsley himself actually tolerates; (4) to assess the real difficulty of Beardsley’s theory and to indicate the possible ways out.


The rule-governed nature of illocutionary acts is crucial in that Beardsley notably denies the relevance of speaker’s intention in determining the correct illocutionary act being performed. An illocutionary act can be performed unintentionally in the sense that it is not necessary for the speaker to have the corresponding intention to perform that illocutionary act.11 Notice that this intention is not the default intention in “intentional text production.”

To demonstrate the point, suppose the following conditions hold: (1) I know that you are looking for an Agatha Christie to read. (2) I believe that you have read At Bertram’s Hotel. (3) I do not know that actually you have not read it. (4) I think that it is not worth reading. (5) I gather that you plan to read At Bertram’s Hotel. Under these conditions, my utterance of the sentence to you “A Murder Is Announced is ten times better than At Bertram’s Hotel” constitutes the illocutionary act of advising you to seek A Murder Is Announced to read. This act is performed intentionally, that is, with the intention to give you advice. But at the same time, there is another illocutionary act being performed unintentionally: that of discouraging you from reading At Bertram’s Hotel.

It follows that performing an illocutionary act has nothing to do with having as its condition an intention to perform that act. We do not need to know the speaker’s intention to determine which illocutionary act is being performed, which is solely a function of relevant conditions and conventions.

Beardsley claims that the nature of fiction is representation. In its broad sense, a representation is a symbolic vehicle standing for something else. By this definition, words, texts, paintings, or sculptures are all representations. Plainly, Beardsley’s use of the term is narrower.

Representation involves what Beardsley calls selective similarity.12 For X to be able to represent Y, X must bear some crucial aspects that can be perceived to be a characteristic of Y but falls short of being Y. In the pictorial context, we know that a painting is a depiction of, say, a smiling face because its content exhibits selective respects distinctive of a real smiling face. The painting is hence a representation of a smiling face. In the case of dramatic representation, we know that a mime is representing the act of, say, climbing a ladder because her body language shows crucial aspects that are characteristics of ladder climbing. Her act is one of imitation.

Representation, according to Beardsley, is best understood as some kind of reference: X refers to Y if X represents Y. Such reference may be intentional at the very beginning of the practice of representation, in the sense that what is referred to is determined by the referrer’s intention. However, once the conventions of a particular kind of representation are established, unintentional reference becomes possible. This is called the detachment of reference.13

Such being the case, we can have different conventions for representing things. Needless to say, the content of representation in most cases corresponds to what one intends to represent, but what one does represent in the end depends on the relevant conventions. It might also happen that one intends to perform some action but ends up representing it instead, if the action is incomplete. This is to say that representation can be unintentional in the sense that the intention to represent is not necessary for producing a representation. Of course, when the performance fails one might end up representing nothing at all, if the similarity produced is insufficient.14

So far we can see that a representation is in place when an attempted action is incompletely performed but is still discernible in what has been done. Then, what is a representation of an illocutionary act per se? Verbal representation works in the same way as visual or dramatic representation: an illocutionary-act representation obtains when the conditions present are not sufficient for the act in question actually to occur. I will define an obtaining condition of an illocutionary-act representation as an infelicity condition of an illocutionary-act performance. And this will presuppose that the attempted action is always of a recognizable kind as in the current context talking about unintelligible actions is out of place. An example of the obtaining condition for an illocutionary-act representation will be the absence of uptake. That is, the utterance is not heard and understood by the audience. The illocution hence loses its force and becomes a representation.

The case of intentional verbal representation is a special case of refraining from illocutionary commitment in order to produce a fiction.15 For instance, when you catch someone stealing your wallet and say “You stole my wallet!”, you are performing an illocutionary act of accusation. But, when an actress during a performance says “You stole my wallet!”, she is not performing an illocutionary act of accusation but only representing one; she is refraining from an illocutionary commitment in order to produce a fiction or imitation.

This is not to exclude the case in which the actress intends to represent some other illocutionary act but ends up representing the one of accusation due to a slip of tongue. The point is that, as representation is a matter of convention, what is represented is solely determined by convention. Furthermore, just as a failure to perform an act may end up with a representation, so a failure to perform an illocutionary act might end up with an illocutionary-act representation because of the absence of uptake from the audience, for example.

It is now clear that a representation or imitation is essentially a fiction. The drawing of a smiling face is a fictional smiling face; the represented action of climbing a ladder by a mime is a fictional action of ladder climbing; the illocution-act representation of accusing someone is a fictional illocutionary act of accusation, and so on.


Before we see how the said notions find their places in literature, some important terms should be specified. First, Beardsley uses the word author to refer to anyone who intentionally produces a text. Thus a parrot or a speaking dreamer is not an author. To mean something, someone intentionally produces a text and in doing so intends to say something. The word “say” covers two different kinds of speech act: illocutionary acts and their representations.16

Second, a compound illocutionary act is defined as a set of illocutionary acts performed in one single text. A literary text, either fictional or nonfictional, can be seen as reflecting a compound illocutionary act.17

Third, in every literary work we can envisage an implicit speaker18 who performs the compound illocutionary act, and “whose words the work purports to be.”19

Now, let us consider the case of literary fiction. Typically, these are fictional narratives such as novels, short stories, and plays. The crucial question is: why does Beardsley think of them as representations or imitations of illocutionary acts? This is because we can always discern in them the infelicity conditions of an illocutionary-act performance, that is, the obtaining conditions of an illocutionary-act representation. These conditions can be seen as the marks of fictionality, which make fictional the illocutionary acts performed in the work.

A first point to consider is whether we should identify the implicit speaker with the author. If the answer is no, the author must be representing the illocutionary act performed by the implicit speaker. This speaker typically appears in two ways: (1) she tells the story behind the scenes and never reveals herself; (2) she narrates from the first person perspective as one of the characters in the story world.20

The identification is certainly wrong in the second case: no one would identify Dr. Watson with Conan Doyle, who in most Holmes’ adventures is only imitating the compound illocutionary act performed by a fictional speaker.

It is perhaps less clear why such identification in the first case is also illegitimate. A quick example may help. Consider the following passage from Ngaio Marsh’s mystery novel Artists in Crime, in which an artist’s model is murdered:

She examined the body. She states that the eyelids fluttered and the limbs jerked slightly. Miss Bostock attempted to raise Gluck. She placed her hand behind the shoulders and pulled. There was a certain amount of resistance, but after a few seconds the body came up suddenly. Miss Seacliff cried out loudly that there was blood on the blue silk drape.21

If we take this passage to be an illocutionary act performed by the implicit speaker, it would be one of describing a crime scene. However, given that the event is not real (no model called Gluck dies in the way described by Marsh), and that the names mentioned here do not refer to actual people (there might be someone called Gluck, Miss Bostock, or Miss Seacliff around us, but we cannot find a perfect match in the real world), the illocutionary act of describing this particular event is not likely to occur. The act is thus pretend. This explains why a fictional narrative is a representation of a compound22 illocutionary act in nature; there are always nonreferring names and descriptions in such a work that make fictional the narrator and the illocutionary acts performed by her.23

Beardsley is unusual in holding that, apart from narrative fiction, there is a second literary category that should also be taken as fiction, and thus as a form of representation: lyric poetry.24 Sometimes the infelicity conditions in such poems are easy to discern. Beardsley’s example is the sonnet by John Keats that begins with the line “Bright Star! Would I were steadfast as thou art—.” The speaker here is addressing the star but apparently uptake from the addressee is impossible. The purported illocutionary act is not likely to happen, and Keats is hence only imitating the act of praising performed by the implicit speaker.

One compelling reason for treating lyrics as illocutionary acts is that they are typically spoken from the first person point of view to express personal feelings and sentiments. It seems natural and right to identify the speaker in the poem with the author, and thus to attribute those feelings and sentiments to the author, when no obvious infelicity conditions are present. Take for instance the poem “Thoughts on a Silent Night” by the ancient Chinese poet Li Bai:

Before my bed a pool of light—
O can it be frost on the ground?
Eyes raised, I see the moon so bright;
Head bent; in homesickness I’m drowned.25

Though researchers do not seem to have any evidence that Li Bai based this poem on true experience, it is very intuitive and tempting for us to assume he did, and hence to identify the speaker in the poem with him. But one would find this kind of identification problematic if one came to know who wrote the following poem only after reading it:

With Northern grass like green silk thread,
Western mulberries bend their head.
When you think of your home on your part,
Already broken is my heart.26

This lyric poem, titled “A Faithful Wife Longing for Her Husband in Spring,” is a poem about a woman grieving over her husband’s absence due to military service. Simply put, the illocutionary act performed by the lonely wife seems to be one of complaining about the absence of her beloved. At first glance, one might suspect that it was written by a female poet. But surprisingly, this poem is also Li Bai’s work. It will be absolutely wrong to identify the speaker in the poem with the author because Li Bai is a male and does not have any husband. It is impossible that he ever had the emotional experience described in the poem. Li Bai in this case is representing the illocutionary act performed by a lonely wife. One would be yet more surprised to find that Li Bai actually wrote more poems from a female perspective, imitating various types of illocutionary acts that express women’s emotions and feelings.

So far it is clear why Beardsley rejects the identification of the speaker in a lyric with its author. If this identification is universal, what is true of the speaker must be true of the author, and vice versa. But the truth is that there is no necessary link between what is expressed in a lyric and its author. A lyric is hence best treated as an illocutionary-act representation, rather than as an illocutionary act.

Another reason for rejecting lyrics as genuine illocutionary acts is the so-called address without access, which constitutes the primary mark of fictionality.27 A published poem is not addressed to a reader the way a love letter is to him or her. The latter is done in a pragmatic context, which the former lacks.28 The absence of a pragmatic context gives us no reason to identify the implicit speaker with the author, who could just be imitating someone else’s speech acts. This is because the act of publishing a poem gives it an impersonal public character. This second locutionary act, to be distinguished from the first of producing the text, detaches the poem from the occasion of utterance and gives us no guarantee that the author is performing an illocutionary act. The poem is hence “freed of illocutionary dependence on its occasion.”29

It can be seen that the above reason also applies to narrative fiction. I see no problem in thinking of it also as the reason for treating literary fiction as representational in general.

What about nonfictional works? They tend to have strong realistic elements and lack marks of fictionality, or infelicity conditions.30 As Beardsley indicates, many pieces of discursive prose are duly classified as nonfiction, including history, philosophy, religious meditations, and personal essays.31 And it seems wrong to say that the illocutionary acts made in these works are pretend. For example, if a poet publishes an antiwar poem during wartime, it seems inappropriate to see the utterance as a mere representation devoid of the illocutionary force of opposing war.32

One might object that here Beardsley is self-defeating, given his view that “address without access” is a mark of fictionality. Beardsley has a response. When considering the antiwar poem it is hard to deny that publishing such a poem in the context of its production is very close to using or presenting that text, given that there are no other marks of fictionality present. It is better to treat the context in question as a pragmatic one in which it is legitimate to identify the author with the implicit speaker.

There are indeed reasons for making this move. As Beardsley points out, nonfictional works tend to be published in a more formal style and targeted at a particular audience.33 For example, many of Confucius’s sayings are addressed to political rulers; a large part of Kierkegaard’s literary writings is addressed to the Church. There is a sense of intimacy with the reader in this kind of work, in which the author is keen to secure proper uptake from the intended audience; that is, she is not addressing without access. But this strong sense of communication is absent from fictional works.

The conclusion we can draw at this stage is that, by rejecting the said mark of fictionality, published nonfictional works are justified as genuine illocutionary acts performed by their authors.

Beardsley draws on what has been argued so far to shed new light on his anti-intentionalism on literary interpretation. As I see it, his complete argument based on speech act theory comes in two parts, in which he tries to show that intention is twice removed from an illocutionary-act representation.

First, the author’s intention is not constitutive of illocutionary meaning (of which illocutionary act is performed) because performing an illocutionary act does not require the author’s intention. As nonfictional works are genuine illocutions, it follows that the author’s intention is neither necessary nor sufficient in determining their meanings.

But even if authorial intent did play a role in fixing the meaning of an illocution, the same would not hold with its representation. These are cases in which authors appear to withhold illocutionary force in order to produce representations. As the content of a representation is fully determined by convention, authorial intent is logically independent of meaning-shaping. And since fictional works are illocutionary-act representations by their very nature, the author’s intention is then irrelevant to what such works mean.34

Beardsley concludes that the intentional fallacy comes from the following confusions: (1) of illocutionary-act performance with its representation; (2) of authorial meaning with textual meaning (what was intended with what was actually said); (3) of the real author with the implicit speaker; (4) of the subject of the work with the occasion of its utterance. These confusions are implicated in the intentionalist argument that consists of two sub-arguments: from biographical data35 to probable intentions, and from probable intentions to work meaning. The first argument is legitimate, while the second is invalid and is a fallacy. It is this latter type of reasoning that Beardsley is strongly opposed to.36


We can clarify Beardsley’s view further by answering some representative objections before we continue with the issue of interpretation.

George Dickie, though sharing Beardsley’s anti-intentionalism, is not satisfied with the speech act argument and points out two problems.37 The minor problem is that the argument applies only to fictional works. However, Beardsley’s complete argument works with all literary works with respect to the autonomy of textual meaning, as I just showed.

Dickie’s major criticism is that Beardsley is not addressing the real debate, which is on locutionary meaning rather than illocutionary meaning. This criticism is doubtful for the following reasons. First, it seems that illocutionary meaning does figure in the real debate. For example, Beardsley’s major rival E. D. Hirsch implicitly raises the question about illocutionary meaning when asking whether the statement “Pass the salt” is an order, a command, or an entreaty.38 Moreover, if we accept that irony, allusion and metaphor take place at the illocutionary level, a matter that participants in the debate argue over, they constitute strong counterexamples to anti-intentionalism.39

Second, what kind of “meaning” is the real debate about? Christopher New distinguishes four levels of meaning figuring in the debate: (1) the vocabulary level (the correct words used); (2) the locutionary level; (3) the illocutionary level; (4) the suprasentential level (themes and theses). Dickie’s favorite example of misspeaking is at the vocabulary level40; Sirridge engages anti-intentionalism at the same level. Noël Carroll defines the meaning of a work as its “themes and theses,” which correspond to the suprasentential level.41 Peter Lamarque also thinks that the debate should be engaged at this level.42 All this shows how divided the opinions are. The most we can say is that locutionary meaning is undoubtedly part of the real debate, but it is not the debate.

Third, an important point: Beardsley does not abandon the discussion of verbal meaning but addresses it indirectly. As the meaning of a word or phrase will affect which illocutionary act the sentence in which it occurs reflects, such determination of meaning is done by considering which choice of meaning contributes most coherently to the illocutionary act reflected in the text.43

Carroll presents four attacks on the soundness of Beardsley’s argument.44 The first is that Beardsley is wrong in claiming that all literature is fictional, which is apparently a misunderstanding of Beardsley on Carroll’s part.45

Second, not all assertions in fictions are representations of illocutionary acts. For example, in many fictional works there are passages about scientific knowledge, history, or philosophical doctrines. It is inappropriate to say that in these cases the author is always pretending to assert.

Carroll is mistaken in considering only intentional representation. Recall that Beardsley’s reason for treating a fictional work as an illocutionary-act representation is that there are always infelicity conditions of illocutionary-act performance present in such a work. In that case, the literary utterance tends to end up with a representation. As said, this includes both cases of intentional and unintentional representation. It may be true that sometimes the author indeed aimed to assert instead of merely pretending to assert; however, the act eventually ends up with a representation, because there are always marks of fictionality that remove its illocutionary force.

The third objection has it that the thesis projection in a serious literary work, seen as a performance of illocutionary act, is not always pretended. It would be absurd to say that when Sartre wrote Nausea he was just pretending to suggest philosophical ideas.

Again, the real reason for treating Sartre’s work as representation is that his utterance satisfies the conditions of representation, not that he must have the intention to represent. To reject Beardsley’s claim that every fictional work is a representation, what one should do is reject the obtaining conditions he identifies, not reject the assumption that all authors must have the intention to represent in every case, which Beardsley also denies.

The fourth objection is that convention cannot settle all interpretative inquiries. Questions about character construction and its function in the overall design of the work can be answered only by appeal to the real author, rather than to the implicit speaker who can be understood only via conventions.

This is a big point of debate. Beardsley could maintain that appeal to authorial intent is still dispensable because textual evidence is sufficient for us to consider character construction and function. If we really need a notion of author, one based on textual evidence will still be an option for the anti-intentionalist. Defending Beardsley, Daniel O. Nathan is basically holding this view when he draws on the concept of the ideal author, whose nature is mainly based on textual evidence.46

Dickie’s and Carroll’s main concerns are with interpretation, while Joseph Margolis’s target is the speech act theory itself. Margolis criticizes a series of speech act theorists, including Beardsley. He specifies four confusions in the pretense theory of literature.47

First, verbal imitation is not itself a speech act but an adverbial qualification of it, better described as “mimickingly.” Therefore, Beardsley is wrong to say that the verbal imitation is a sort of speech act. Nevertheless, just as we do not say that a mime climbs a ladder mimickingly, so we do not say that a stage actress speaks mimickingly. We say that she imitates the speaking of the fictional character. To imitate or to mimic is verbal rather than adverbial in either case. If the stage actress is just speaking mimickingly, what kind of speech act is she performing? Margolis does not say. And verbal imitation still seems to be the answer.

Second, not all categories of poetry are fictional discourses, and we cannot even assume that all literary writers have the intention to pretend when composing. These criticisms are again misrepresentations of Beardsley’s view that I have already clarified.48

Third, based on the first and second objections, it would be better to understand literary distinctions in terms of styles or genres rather than as speech acts.

Surprisingly, Beardsley had tried to show that the concept of style can still be accommodated in the speech act model.49 He distinguishes between dominant (primary) and subordinate (secondary) illocutionary acts (either real or fictional). For example, Caesar’s famous triple locutionary act “I came, I saw, I conquered” not only makes three dominant illocutionary acts of assertion, but also makes a subordinate one of asserting that, for him, to arrive was to act. This latter act is subordinate to the other three. Subordinate illocutionary acts can be multiple, as it is possible that one single illocutionary act is accompanied by several implicit ones.

Drawing on this distinction, a stylistic feature is thus defined as a feature of any linguistic form that enables a subordinate illocutionary action. Difference in style means difference in meaning; sameness in meaning means sameness in style. The view that authors can express the same thing in different styles is faulted. To say something in a different way is basically to say that thing plus something else because a stylistic feature is tied to subordinate illocutionary acts.50

As for the concept of genre, Beardsley hints, but not clearly advocates, that a genre could be defined in virtue of the style commonly found in it.51 That is, a genre G can be defined by a set S of which certain linguistic features are members.

Margolis’s final objection is that the mimetic theory confuses imitating a speech act with speaking poetically, and with imagining someone speaking. Let us consider the first kind of confusion. Margolis claims that there is no reason to suppose that it is not the author herself speaking poetically. But the point is, as long as any marks of fictionality are discerned in one’s utterance, those marks qualify the utterance as a representation. That is how Beardsley draws a line between fictional works and nonfictional ones, and also between someone who is imitating a speech act and who is speaking poetically.

Margolis claims further that there is also no reason to suppose that the author is reporting anything about any speaker: “he is simply speaking, though not necessarily speaking simply.” This “simply speaking” oversimplifies speaking. If the speaker is not speaking as herself, she is then representing someone else speaking, that is, the implicit speaker in the story.

The second sort of confusion is a key point. According to Margolis, it is wrong to conflate “reporting what an imaginary speaker ‘has said’ (granting the fiction) and creating a poem in which an imaginary speaker speaks.” Instead of representing an illocutionary act, what an author really does in telling a story is to imagine a fictional world in which characters speak to each other. To imagine this world is a distinct act, but it is not a speech act. Margolis concludes that “the fictional use of language concerns the presuppositions that inform a set of speech acts, not those speech acts themselves.”

Now, writing a work (be it fictional or nonfictional) necessarily involves intentional text production, and in performing this very act the author intends to say something. In that case, it is hard to see why this act is not itself a speech act. The author of a fictional work creates or imagines a world by representing what a fictional narrator has said. If she just imagines a story in her mind, that does not involve any fictional use of language; once she outputs her imaginings into words, this very act is suitably perceived as a speech act. Put another way, in writing a fictional work, the author performs a speech act that informs us of an imagined world.

Margolis’s objection reflects a fundamental disagreement in the debate on literary interpretation: whether writing a literary work counts as an utterance or not. If the answer is positive, producing a literary work will no doubt be a speech act. I will not pursue the issue here.


Once we treat a fictional work as an illocutionary-act representation, our interpretative task becomes that of constructing the act likely to have been performed, based on textual clues.52 Reconstruction in this case does not make sense, for the illocutionary act in question is only pretended, one that falls short of being; there is no actual act to be reconstructed at all. Rather, our goal is to construct the act by relying on textual evidence to supply its missing conditions.

By contrast, the reconstructionist position can be described as follows:

An ordinary text … belongs to a set of circumstances, or an encompassing situational context. If we are in doubt about what it says—that is, about exactly what illocutionary actions were being performed in the act of composing it—our task is to reconstruct the situational context and examine the act-generating conditions that were present. Though we can always in principle distinguish between the question: Which illocutionary actions did the author perform? And the question: Which illocutionary actions did the author intend to perform, we are often mainly interested in the latter … if we take a poem as a text that is to be interpreted in this reconstructionist fashion, we will study the letters of the author and the memoirs of his friends.53

Once we take this stance, the interpretative interest typically shifts from textual meaning to authorial meaning; that is, we would be more interested in the illocutionary act the author intended to perform than in the one she did perform. This is because an illocutionary act is an act of communication, and typically the audience’s goal is to grasp what the utterer actually intends to say. In that case, research into biographical data and other types of external evidence becomes necessary.

A problem to be solved first is whether the intentionalism described above is really a brand of intentionalism as such. In the contemporary debate of literary interpretation, intentionalism subdivides into actual intentionalism and hypothetical intentionalism. An important feature of the former is that it allows authorial intent to be constitutive of textual meaning. For example, extreme intentionalism maintains that authorial meaning is identical to textual meaning.54 Modest intentionalism allows authorial intent to be a necessary component of successfully produced textual meaning.55 In the case of hypothetical intentionalism, authorial intent is better described as relevant to, rather than constitutive of, meaning, in the sense that meaning-determination involves hypothesizing authorial intent.56

The version described in the present context seems to be very different. Beardsley denies that a speaker’s intention is constitutive of, or is relevant to, meaning with respect to language in general, including illocutions.

Perhaps we can call the Beardsleyan intentionalism quasi-intentionalism, which denies that the author’s intention can affect meaning-determination in any way, while allowing an interest in retrieving it. This position qualifies as one sort of weak intentionalism as Beardsley does worry about it:

Textual meaning is not reducible to authorial meaning. But does it take precedence over it? We have not yet established the full authority of the text. For if the two meanings are not identical, then there are two possible interpretation-tasks … which of them is the proper function of the literary interpreter? No one can deny that there are many practical occasions on which our task is precisely to try to discover authorial meaning, or intention ….57

The problem at hand becomes whether Beardsley can consistently hold the following claims: (1) quasi-intentionalism is appropriate for interpreting illocutions; (2) nonfictional works are illocutions; (3) we should hold anti-intentionalism on interpreting literary works. Two considerations might save Beardsley from inconsistency.

First, prior to the speech act argument, Beardsley presents many others against intentionalism, all of which are directed at literary works in general, not just fictional works. Even if the speech act argument fails to ward off quasi-intentionalism, his other arguments might still do so.

Second, Beardsley actually has a response to quasi-intentionalism. With regard to literature (and actually all the arts), our main task is to dig out textual meaning rather than authorial meaning, even though in ordinary life the latter concerns us more. That is to say, literary interpretation is a special case in which the interpreter has to take a robust anti-intentionalist stance even on illocutions.

Adopting the above moves sacrifices (2). I want to argue that it is possible for Beardsley to accommodate quasi-intentionalism within his framework without violating the anti-intentionalist strictures he wants. The moral of my formulation is that even for the most unyielding anti-intentionalist some degree of intentionalism may prove helpful and perhaps even necessary.

Beardsley actually makes four concessions that allow for quasi-intentionalism. I will show them one by one before considering how they variously apply to the interpretation of fictional and nonfictional works and how they together form a sound interpretative strategy compatible with anti-intentionalism.

It is true that the initial aim of the intentional fallacy was to expose the fallacious use and irrelevance of biographical data (external evidence) in interpretation.58 Apparently, the word “fallacy” is not used here in its technical sense. It does not explicitly refer to a certain pattern of inference (although this inference pattern might be implicitly assumed); rather, it refers to some illegitimate action, that is, the appeal to authorial intent. This is what the dilemma argument in the notorious article “Intentional Fallacy” really shows: either an intention is successfully realized in the work or it fails; if an intention is successfully realized in the work, there is no need to consult external authorial intent; if it fails, it turns out to be something outside the work; therefore, either way appeal to external evidence of authorial intent is irrelevant to interpretation.59 This reading is supported by the fact that external evidence is seen as illegitimate.60

But do not forget that Beardsley legitimizes the use of intermediate evidence, which is about:

… the character of the author or about private or semi-private meanings attached to words or topics by an author or by a coterie of which he is a member. The meaning of words is the history of words, and the biography of an author, his use of a word, and the associations which the word had for him, are part of the word’s history and meaning.61

The use of intermediate evidence does not involve (full-fledged) intentionalism. It may be evidence of what the author intended; more importantly, it is also evidence of the meaning of her utterance. This concession consequently grants us some room for talking about biographical data and authorial intent, as it involves probing into not only the histories of the words used in the text but also the topics to which private associations are attached.

In later discussion Beardsley loosens the strictures more. He begins to allow full consultation of external evidence, with the proviso that the inference from external evidence to probable intentions be tested against textual or internal evidence.62 This move presupposes what Beardsley calls the Principle of Autonomy: “literary works are self-sufficient entities, whose properties are decisive in checking interpretations and judgments.”63 As meaning is an important property of a literary work, the work enjoys autonomy with respect to meaning. A literary work, construed as a text, acquires meaning in terms of the linguistic conventions of the words used in it, not affected by external factors and hence enjoying semantic autonomy.

The significance of licensing full appeal to external evidence lies in an apparent but often neglected fact accepted by Beardsley: the author’s intention, when successfully realized, codetermines meaning. This is indeed implied in the dilemma argument. Accurately speaking, a successful intention of the author can never alone determine meaning because convention is sufficient for doing the job. But since authorial intent still plays some role here, consultation of it seems natural, legitimate, even if dispensable. Stephen Davies puts this point well:

Wimsatt and Beardsley do not argue that it is fallacious to consult far-flung evidence of an author’s intentions in arriving at interpretations of a work, provided that what is uncovered is tested against what is manifestly in the work. We can easily overlook interpretationally relevant features of the work and recourse to data about its author’s intentions could be helpful in guarding against this danger. In other words, Wimsatt and Beardsley hold that an author’s intentions never alone determine the contents of a work, but this does not entail that reference to those intentions can’t help us to detect meanings that have been conveyed successfully to the work.64

This involves a move away from the initial sense of intentional fallacy. This can be seen from how Beardsley reidentifies the intentional fallacy: making an inference from the author’s intentions (premises) to textual meaning (conclusion).65 In other words, the real intentional fallacy is about the misidentification of textual meaning with authorial meaning, not about the appeal to external evidence.66

This concession results in licensing the first part of the intentionalist argument: inductive inferences from biographical data to intentions. This grants enough space for an interest in authorial meaning. If one is allowed to explore biographical data fully and infer authorial intent, what else can one do except reconstruct it? The admission of testing probable intentions against textual evidence is especially significant in the case of nonfictional works. This is because such works always refer to reality, and we would be very interested in the discrepancies between what the author intended to say about the real world as contrasted with what she ends up saying.

The third concession is more explicitly stated. To support the view that literary interpretation should aim for textual meaning rather than authorial meaning, Beardsley supplies two arguments.67 The first concerns the unavailability of authorial intent and the second is about aesthetic satisfaction resulting from understanding the work itself. Nevertheless, Beardsley admits that the first argument is inconclusive and the second “takes for granted quite a few assumptions about art and aesthetics.” Worried about “the stiffness and formality of my purported demonstration,” Beardsley makes the concession that modest importation of authorial intent is admissible.68 This is not to allow authorial intent to constitute textual meaning in any way; rather, the importation is taken to be an informative extension. Given that the boundaries of textual meaning tend to be vague, it is not always clear what is assumed in, or excluded from, the text. In other words, Beardsley here is not granting authorial intent the role of meaning-shaping but is moderately allowing the consideration of it to enrich criticisms where no obvious conflicts with textual evidence occur.

The fourth concession is significant, as it undermines semantic autonomy, the principle that textual features of a literary work are sufficient for determining meaning. Now recall the first part of Beardsley’s speech act argument: speaker’s intention is not constitutive of illocutionary meaning. Though Beardsley believes this sub-argument to be correct, he leaves room for doubting its conclusiveness. He suspects that some sort of intention might be constitutive of the illocutionary act being performed in some cases. For example, the intention to deceive might be requisite for lying.69 Beardsley hesitates about such autonomy in the case of illocutions. Once this point is acknowledged, the author’s intention will be partly constitutive of meaning: it will partly play a role in determining which illocutionary act the author performs.

Certainly this last concession has gone further than quasi-intentionalism, which does not reject semantic autonomy. And as the most important contention Beardsley wants to secure (and other debaters want to attack) is semantic autonomy, I do not think that we should interpret his doubt as certainty. But given the indecisiveness raised, I do think that it leaves some room for importation of authorial intent, in a different way from the second concession. The critic may be granted more room for speculating about textual meaning based on probable authorial intentions where textual clues are indecisive. But this will be speculation only, and by the anti-intentionalist strictures we should still maintain that the text in this case is ambiguous in nature.70

The four concessions outlined above form a qualified version of quasi-intentionalism (I will call it QQI hereafter). It is qualified because it does not grant a primary role of speaker’s meaning in criticism. Consideration of speaker’s meaning is dependent upon that of textual meaning, as the four concessions all assume textual meaning as the basis on which further thought on speaker’s meaning becomes possible.

Now, let us see how the above discussion links to fiction and nonfiction, and how it may help resolve the inconsistency exposed so far in Beardsley’s framework, along with the stiffness and formality with which he feels uneasy.

Beardsley’s main worry about quasi-intentionalism is that a critic should not make the reconstruction of authorial meaning dominant in her criticism. Though it is unclear how much a critic can deal with reconstruction explicitly, it can be reasonably assumed that Beardsley does not completely exclude talk about the author, so long as one makes a clear distinction between what the author intends to say and what is actually said. In Beardsley’s terms, this is the distinction between non-intentionalistic criticism and intentionalistic criticism. When a critic talks intentionalistically, she never uses the subject terms “the artist” and “the work” discriminately and tends to mix up evidence of intention with evidence of work meaning.71

Ideally, references to the author should be avoided, and the critic should all along focus on textual meaning. There seems to be no problem with this ideal applied in the case of the interpretation of fictional works. Given that authorial intent is ontologically detached from reference in the representational framework, we may renounce the talk about reconstruction. Analogically speaking, probably not many people would bother to inquire into what a mime intended to imitate in a performance, not to mention that her biography would shed little light on what she would like to imitate.

But in the case of nonfictional works, we have more license to allocate space for authorial intent, because these are based on illocutions instead of mere imitations. And this is legitimately allowed by QQI: the critic is free to explore private associations by looking into the author’s biography, to compare authorial meaning with textual meaning, to enrich textual meaning with biographical data, and to speculate about textual meaning based on such data—where these moves are relevant. Beardsley’s worry that authorial meaning would take precedence over textual meaning is not quite justified, because these concessions do not fully justify giving it that precedence. Their shared premise is that talk about authorial intent is dependent on what is already said in the text. It is not possible to engage the former without the latter. Certainly in practice it is hard to draw a clear line between moderation and excessiveness, but to argue that there is, hence, no such a distinction would commit a decision-point fallacy.72

I sum up for Beardsley three critical principles that emerge from what has been discussed so far. First, external evidence is in theory not essential to interpreting fictional works. As just mentioned, the ideal is to focus on the work itself rather than on its creator, which has been the central tenet of anti-intentionalism. Nevertheless, this does not entail an ultimate exclusion of authorial intent from interpreting fictional works, assuming that QQI is at work.

Second, reconstruction of authorial intent is encouraged, if not necessary, in the case of nonfictional works. Literary illocutions are still illocutions, and QQI satisfies our interest in digging out speaker’s meaning when interpreting them without losing ground to intentionalism as such. This leaves room for a flexible approach to criticism of nonfictional works.

Third, a critic should keep talking non-intentionalistically, that is, objectively, instead of intentionalistically, so that no confusion of authorial meaning with textual meaning occurs. This is a crucial caveat to bear in mind. If we are to put QQI in practice, the distinction has to be made all the time to avoid shading into intentionalism.

Beardsley’s biggest concern is what Carroll calls the constitutive question: what constitutes the meaning of the work?73 A significant part of the debate of literary interpretation and most of Beardsley’s arguments are directed to answering that question, including the speech act argument. Once the core of anti-intentionalism is secured, Beardsley need not worry about the residue of intentionalism implicated in the speech act theory because he can still make it compatible with the anti-intentionalist tenet.


Before offering the speech act account of literature, Beardsley is apparently a formalist and anti-pragmatist in regard to literary interpretation. He is only interested in the work itself and shows indifference to external factors embedded in the context of the work’s production. The work enjoys full autonomy and remains aloof from outside influences.

However, the introduction of the speech act theory to anti-intentionalism damages such autonomy and suspiciously results in contextualism, the view that the identity and meaning of a work partly depend upon the relations it holds to the contextual factors present at the time of its production. As claimed, to grasp the meaning of an illocution, one will have to consider the pragmatic context in which the text was produced in order to identify the act-generating conditions constitutive of meaning. In the case of nonfictional works, this is tantamount to saying that textual meaning can only be pinned down by appealing to contextual factors, excluding authorial intent. This move seriously undermines semantic autonomy.

One way to save Beardsley’s theory would be to hold a weaker version of the principle of autonomy with respect to nonfictional works and to claim that contextual considerations, excluding the author’s semantic intentions, do contribute to textual meaning. This appears to commit Beardsley to some form of contextualism, but one might think that even Beardsley’s conventionalism itself is already a weak form of contextualism: if a text is in a given language, that immediately makes its possible meanings context-dependent. A problem of this approach is that the exclusion of authorial intent from contextual factors seems dogmatic.74 But some theories of interpretation based on contextualism—for example, hypothetical intentionalism—do take this line,75 and it has been argued that Beardsley’s account differs from hypothetical intentionalism more in degree than in kind.76

Considerations of context lead to deep issues in the ontology of art: does the meaning and identity of a work change across time? Some philosophers argue that context continues to affect a work’s identity and that when context changes, the work crucially alters while reaming self-identical. For example, Shakespeare’s Macbeth has generated different meanings in different contexts while remaining self-identical. This view, dubbed by Davies the modern context theory, is appealing for those who deny the idea of a text as a stable center.77

The debate is still ongoing within philosophical aesthetics. And it is of interest to note that Beardsley paradoxically shows leanings to the modern context theory; he accepts that work meaning changes as convention changes in a different sociohistorical setting.78 This interesting fact is discussed and seen by Richard Shusterman as a way for Beardsley to leave room for continuing interpretation of a work.79

This also echoes the suggestion made earlier that tries to reconcile Beardsley’s view with a certain form of contextualism. The above discussion shows that Beardsley’s anti-intentionalism still has the potential to come to terms with the recent trends such as literary pragmatics and contextualism. A fuller account is yet to be formulated.


1. John L. Austin, How to Do Things with Words (Oxford: Clarendon Press, 1962).

2. Richard Ohmann, “Speech Acts and the Definition of Literature,” Philosophy & Rhetoric 4 (1971): 1–19. See also his “Speech, Literature, and the Space between,” New Literary History 4 (1972): 47–63.

3. Charles Altieri, “The Poem as Act: A Way to Reconcile Presentational and Mimetic Theories,” Iowa Review 6 (1975): 103–24.

4. John Searle, “The Logical Status of Fictional Discourse,” New Literary History 6 (1975): 319–32.

5. Mary Louise Pratt, Toward a Speech Act Theory of Literary Discourse (Bloomington, IN: Indiana University Press, 1977).

6. Stanley E. Fish, Is There a Text in This Class? The Authority of Interpretive Communities (Cambridge, MA: Harvard University Press, 1980).

7. There are other literary applications of speech act theory: Ohman offers one of the earliest accounts of verbal style based on the speech act model in “Instrumental Style: Notes on the Theory of Speech Action,” in Current Trends in Stylistics, eds. Braj B. Kachru et al. (Edmonton: Linguistic Research, 1972), 115–41; Ted Cohen analyses figurative speech in terms of speech act theory in “Figurative Speech and Figurative Acts,” Journal of Philosophy 72 (1975): 669–84; and Keir Elam applies speech act theory to explore dramatic language in Shakespeare’s Universe of Discourse: Language-Games in the Comedies (Cambridge: Cambridge University Press, 1984).

8. Joseph Margolis, “Literature and Speech Acts,” Philosophy and Literature 3 (1979): 39–52.

9. Roger D. Sell, Literature as Communication (Amsterdam: Benjamins, 2000), 48–64.

10. Joseph Hillis Miller, Speech Acts in Literature (Stanford, CA: Stanford University Press, 2001), and his Literature as Conduct: Speech Acts in Henry James (New York: Fordham University Press, 2005).

11. Monroe C. Beardsley, “Intentions and Interpretations: A Fallacy Revived,” in The Aesthetic Point of View: Selected Essays, eds. Michael J. Wreen et al. (Ithaca, NY: Cornell University Press, 1982), 196–7.

12. Monroe C. Beardsley, “Fiction as Representation,” Synthese 46 (1981): 294–5. See also his “Intentions and Interpretations,” 191.

13. Beardsley, “Fiction as Representation,” 296–7, and “Intentions and Interpretations,” 192–3.

14. Beardsley, “Fiction as Representation,” 297. Note that unintelligibility is not equal to ambiguity. The content of a representation can be ambiguous in the sense that it resembles more than one type of act. But when the type of action performed is completely unrecognizable, it is unintelligible.

15. Beardsley, “Intentions and Interpretations,” 191.

16. Ibid., 189–91.

17. Monroe C. Beardsley, The Possibility of Criticism (Detroit, MI: Wayne State University Press, 1970), 58.

18. A similar concept, that of “the implied author,” is later fully developed by Wayne C. Booth, The Rhetoric of Fiction (Chicago, IL: Chicago University Press, 1961).

19. Monroe C. Beardsley, Aesthetics, Problems in the Philosophy of Criticism, 2nd ed. (Indianapolis, IN: Hackett, 1981), 238.

20. There can be variations. For example, a story is told from the second person point of view, or from multiple perspectives. But the point is that in all such cases there is still someone who tells the story and qualifies as the implicit speaker.

21. Ngaio Marsh, Artists in Crime (London: Harper Collins, 2000), chapter 4.

22. I will omit “compound” hereafter.

23. Representation in this context means depiction, not portrayal, because the act imitated need not be one that has ever occurred. See Monroe C. Beardsley, “The Concept of Literature,” in Literary Theory and Structure: Essays in Honor of William K. Wimsatt, eds. Frank Brady et al. (New Haven, CT: Yale University Press, 1973), 31–2.

24. Beardsley initially speaks of the whole category of poetry in The Possibility of Criticism, but adopts a more reserved position later in “The Concept of Literature” and “Fiction and Representation.” Also notice that Beardsley does not claim that fictional narratives and lyrics exhaust the realm of fictional works. I suspect that nursery rhymes, for example, would be other possible candidates.

25. Yuanchong Xu, trans., Selected Poems of Li Bai (Changsha: Hunan People’s Publishing House, 2007), 35.

26. Ibid., 67. I omit the last two lines to serve the current purpose better.

27. Beardsley, “Fiction as Representation,” 304–5.

28. Beardsley, Aesthetics, 238–40.

29. Monroe C. Beardsley, “Aesthetic Intentions and Fictive Illocutions,” in What is Literature, ed. Paul Hernadi (Bloomington, IN: Indiana University Press, 1978), 175.

30. There are two worries here. First, many literary critics may be hesitant to refer to nonfiction as literature, because of other non-literary and non-aesthetic functions that can be attributed to the text. Second, the distinction between fiction and nonfiction is not always clear. Beardsley has a reply to both questions. The answer to the first question really depends upon one’s definition of literature. According to Beardsley, the literariness of a text will be defined by whether an artistic intention is at play. For the second question, Beardsley admits that the line is bound to be fuzzy. However, this does not mean that the distinction is useless, as there are more clear-cut examples. See Monroe C. Beardsley, “The Philosophy of Literature,” in Aesthetics: A Critical Anthology, eds. G. Dickie et al. (New York: St. Martin’s Press, 1977), 427–8.

31. Beardsley, “Aesthetic Intentions,” 174.

32. Beardsley, “The Concept of Literature,” 35.

33. Beardsley, “Fiction as Representation,” 305.

34. My presentation of Beardsley’s speech act argument is similar to Wreen’s. However, I think that he is mistaken in stating that Beardsley’s second sub-argument applies to all literary works. See Michael Wreen, “Beardsley’s Aesthetics,” in The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta (2014), Accessed March 16 2015

35. The term is used here in its broadest sense, including biographical information of various kinds and reports of authorial intention.

36. Beardsley, “Intentions and Interpretations,” 207.

37. George Dickie and W. Kent Wilson, “The Intentional Fallacy: Defending Beardsley,” The Journal of Aesthetics and Art Criticism 53 (1995): 235. See also George Dickie, “Beardsley, Monroe,” in Encyclopedia of Philosophy, 2nd ed, Vol. 1, ed. Donald M. Borchert (Detroit, MI: Macmillan Reference, 2005), 509.

38. Eric D. Hirsch, The Aims of Interpretation (Chicago, IL: University of Chicago Press, 1978), 71.

39. Christopher New, Philosophy of Literature: An Introduction (London: Routledge, 1999), 104–5. Irony is an especially popular topic in interpretation. See Daniel O. Nathan, “Irony and the Artist’s Intentions,” British Journal of Aesthetics 22 (1982): 245–56; “Irony, Metaphor, and the Problem of Intention,” in Intention & Interpretation, ed. Gary Iseminger (Philadelphia, PA: Temple University Press, 1992), 183–202. See also Mary Sirridge, “Artistic Intention and Critical Prerogative,” British Journal of Aesthetics 18 (1978): 137–54; and William Tolhurst, “On What a Text is,” British Journal of Aesthetics 19 (1979): 3–14. In “Intentions and Interpretations,” 203–7, Beardsley makes great efforts to show that irony is not a counterexample to anti-intentionalism.

40. Dickie and Wilson, “Defending Beardsley,” 236, 244.

41. Noël Carroll, “Criticism and Interpretation,” Sztuka i Filozofia: Art and Philosophy 42 (2013): 7–20.

42. Peter Lamarque, “Appreciation and Literary Interpretation,” in Is There a Single Right Interpretation? ed. Michael Krausz (University Park, TX: Pennsylvania State University Press, 2002), 285–306.

43. Beardsley, “Intentions and Interpretations,” 193–4.

44. Noël Carroll, “Art, Intention and Conversation,” in Intention & Interpretation, 97–131.

45. Carroll in Footnote 25 notices that Beardsley in “The Philosophy of Literature” recognizes that there are nonfictional works. The recognition is not considered because it undercuts the argument in “Intentions and Interpretations.” The problem is that the second article never makes the claim Carroll attributes. From my presentation of Beardsley’s theory it can be seen that Beardsley has been rather consistent in making a distinction between fiction and nonfiction.

46. Nathan, “Irony and the Artist’s Intentions,” and “Irony, Metaphor, and the Problem of Intention.”

47. Margolis, “Literature and Speech Acts,” 45–8. Though Margolis mentions Beardsley and Richard Ohmann when presenting his criticisms, he addresses the latter most of the time. However, since he classifies the two in the same camp, I see no problem in treating all of his attacks as directed at Beardsley as well.

48. A point to note is that Beardsley sees didactic poems as imitations of arguments: they are philosophy fiction and a subclass of lyric poems. See his The Possibility of Criticism, 59–60, and “Fiction as Representation,” 303.

49. Monroe C. Beardsley, “Verbal Style and Illocutionary Action,” in The Concept of Style, ed. Berel Lang (Philadelphia, PA: The University of Pennsylvania Press, 1979), 149–68.

50. Beardsley, “Fiction as Representation,” 306. Beardsley, however, acknowledges two exceptions in which different expressions result in sameness of meaning: (1) stipulation or definition; (2) elegant variation. See Beardsley, “Verbal Style,” 166.

51. Ibid., 150.

52. Beardsley, “Fiction as Representation,” 307–8.

53. Ibid., 307.

54. Eric D. Hirsch, Validity in Interpretation (New Haven, CT: Yale University Press, 1967). See also Steven Knapp and Walter B. Michaels, “Against Theory,” Critical Inquiry 8 (1982): 723–42.

55. Noël Carroll, “Interpretation and Intention: the Debate between Hypothetical and Actual Intentionalism,” Metaphilosophy 31 (2000): 75–95; Gary Iseminger, “An Intentional Demonstration?” in Intention & Interpretation, 76–96; Paisley Livingston, Art and intention: A Philosophical Study (Oxford: Clarendon Press, 2005); Sirridge, “Artistic Intention”; and Robert Stecker, Interpretation and Construction: Art, Speech, and the Law (Oxford: Blackwell, 2008).

56. Typically in Jerrold Levinson, “Intention and Interpretation: A Last Look,” in Intention & Interpretation, 221–56; Tolhurst, “On What a Text is.” In Carroll, “Criticism and Interpretation,” hypothetical intentionalism is characterized as letting authorial intent fix meaning.

57. Monroe C. Beardsley, “The Authority of the Text,” in Intention & Interpretation, 32–3.

58. Beardsley, “Intentions and Interpretations,” 188; Stephen Davies, The Philosophy of Art (Oxford: Blackwell, 2006), 117–18; Stein H. Olsen, “Biography in Literary Criticism,” in A Companion to the Philosophy of Literature, eds. Garry L. Hagberg et al. (Malden, MA: Wiley-Blackwell, 2010), 441. For a short comment of the divided readings of the intentional fallacy, see Stephen Davies, Definitions of Art (Ithaca, NY: Cornell University Press, 1991), 226–7.

59. William K. Wimsatt and Monroe C. Beardsley, “The Intentional Fallacy,” The Sewanee Review 54, (1946), 468–88.

60. Ibid., 478. Wreen corroborates this point when he says in the last section of “Beardsley’s Aesthetics” that “external evidence is verboten in criticism.”

61. Wimsatt and Beardsley, “The Intentional Fallacy,” 478.

62. Beardsley, Aesthetics, 26. See also his “Fiction as Representation,” 303.

63. Beardsley, “The Authority of the Text,” 24.

64. Davies, The Philosophy of Art, 118.

65. Monroe C. Beardsley, “On the Creation of Art,” Journal of Aesthetics and Art Criticism 23 (1965): 301.

66. Beardsley mentions that the intentional fallacy in his 1946 article is of the re-identified sense. I agree with Dickie that the re-identified sense of the fallacy is implicitly assumed there. See Dickie and Wilson, “Defending Beardsley,” 234. However, that is not what the dilemma argument really proves.

67. Beardsley, “The Authority of the Text,” 33–5.

68. Ibid., 36.

69. Beardsley, “Intentions and Interpretations,” 196.

70. Beardsley, “The Authority of the Text,” 31.

71. Beardsley, Aesthetics, 27–8.

72. That is, concluding that there are no gradations in a certain process just because a clear line cannot be drawn at any point in that process.

73. Carroll, “Criticism and Interpretation,”10.

74. Davies, The Philosophy of Art, 119.

75. Notably, they distinguish the author’s semantic intention, that is, an author’s intention to mean something by a text, from her categorical intention, the intention that the text be classified, taken, and approached in some specific or general way. See Levinson, “Intention and Interpretation,” 221–2.

76. Stephen Davies, “Beardsley and the Autonomy of the Work of Art,” The Journal of Aesthetics and Art Criticism 63 (2005): 179–83.

77. See Stephen Davies, Philosophical Perspectives on Art (Oxford: Oxford University Press, 2007), 146–65. A related view associated with the mutable nature of meaning is constructivism: interpretation imputes meaning, which a work lacks prior to interpretation. Therefore, Macbeth does not have a meaning before a critic constructs one for it, and that meaning changes when another is imputed to the work. See Robert Stecker, Aesthetics and the Philosophy of Art: An Introduction (Lanham: Rowman & Littlefield, 2010), 123–43.

78. Wimsatt and Beardsley, “The Intentional Fallacy,” 488. See also Beardsley, “The Authority of the Text,” 26.

79. Richard Shusterman, “Interpretation, Intention and Truth,” in Intention & Interpretation, 65–75.

About The Author

Szu-Yen Lin
University of Auckland
New Zealand

Szu-Yen Lin is a PhD candidate at the University of Auckland. His research areas include philosophy of literature, aesthetics and philosophy of art.

Article Metrics

Metrics Loading ...

Metrics powered by PLOS ALM

Related Content