One of the things that has been alleged to be the purpose of art is its cognitive function: art as a means to the acquisition of truth. Art has even been called the avenue to the highest knowledge available to humans and to a kind of knowledge impossible of attainment by any other means.

Knowledge in the most usual sense of that word takes the form of a proposition, knowing that so-and-so is the case. Thus, it can be learned from sense observation that the sun is setting, and this is knowledge. Is knowledge acquired in this same sense from acquaintance with works of art? There is no doubt that there are some propositions (statements) that can be made after acquaintance with works of art that could not be made before: for example, that this performance of Beethoven’s Eroica Symphony was 47 minutes long, that this painting predominates in green, that this piece of sculpture originated about 350 bce. The question is whether there is anything that can be called truth or knowledge (presumably knowledge is of truths, or true propositions) that can be found in works of art.

Literature is surely the most obvious candidate, for literature consists of words, and words are combined into sentences, and sentences (at least declarative sentences) are used to convey propositions—that is, to make assertions that are either true or false. And works of literature do certainly contain many true statements: a novel about the French Revolution conveys facts about the series of events; in a verse of the English scholar and poet A.E. Housman, it is said, “The tears of all that be / Help not the primal fault.” Since literature contains statements, it would be surprising indeed if at least some of them were not true.

But the relevance of this fact to literature as an art is extremely dubious. If an 18th-century novel gives a true picture of English country life of that time, this makes it useful to read as history, but does it also make it a better novel? Many, at any rate, would say that it does not: that a tenth-rate novel might give more facts about 18th-century life than a first-rate novel of the same century. For that matter, many of the propositions in a novel are, taken at face value, false; it is false, for example, that there was a foundling named Tom Jones who had an uncle named Squire Western. The thousands of pages of description in novels of fictional characters, ascribing to them thoughts and actions, are all false, since these characters never actually existed. (Some philosophers have preferred to say that propositions about fictional or nonexistent entities are neither true nor false.) Yet this fact in no way impugns their value as literature. Shakespeare, in The Winter’s Tale, sets part of the action on the seacoast of Bohemia, but the fact that Bohemia has no seacoast does not damage The Winter’s Tale as literature, though it would as geography. The fact that Milton used the outdated Ptolemaic astronomy does not make Paradise Lost less valuable, nor does the nonexistence of the lands described in Gulliver’s Travels (1726) in any way diminish Swift’s work. There is no doubt, then, that works of literature can contain true statements and false ones. But it is tempting to ask, What does their truth or falsity matter? Literature is not astronomy or geography or history or any branch of knowledge, particular or general.

Many would hold that the above statements are indeed irrelevant, as are any that encroach upon the domain of science, but, they would add, there are other assertions that matter a great deal: for example, the statements in which a worldview is presented in a poem or drama or novel. The main burden of the ancient Latin poet Lucretius’s De rerum natura (“On the Nature of Things”) is a presentation of the materialism of the Greek philosopher Democritus, and an embodiment of the worldview of medieval Roman Catholicism is the very warp and woof of Dante’s The Divine Comedy (written c. 1308–21)—and such considerations, it would be contended, are relevant to these works as literature.

In reply, however, it might be said that while it is true that these worldviews must be understood and taken into consideration in the reading of these poems and that they cannot be understood or appreciated without knowing them, the truth or falsity of these views still does not matter aesthetically. If Lucretius’s view is true, then Dante’s must be false, and vice versa, since they are incompatible, but, in order to appreciate the poem, it is not necessary to know which (if either) is true. Appreciating art, unlike taking a stand for or against a cause in life, does not require a yes or no to statements. It requires only that the viewers look and appreciate, that they experience as richly and fully as possible the feeling and attitudes involved in the worldview that is presented. Philosophers and scientists are concerned with whether the Democritean materialism of Lucretius is true; appreciators of art are concerned only to capture the feeling appropriate to the worldview in question.

Many statements in works of literature are not explicitly made at all but are implicit: Thomas Hardy never tells in his novels what his worldview is, but it emerges rather clearly before the reader is halfway through any of them. Probably the most important points made in works of literature that contain a central thesis are implicit rather than explicit. How, in that case, can it be determined what thesis it is that is implied? In a court of law, if someone says, “She didn’t say it exactly, she just implied it,” the judge would be likely to rule that this was insufficient evidence of slander, since the person did not actually say it. Still, many statements in daily life are not stated but implied—in the sense that they are intended. The trouble lies in proving that the speaker intended them, since no one else is in a comparable position to say what the speaker’s intentions were, and in the case of deceased authors there is no evidence of their intentions other than what they said. One is doubtless on safer ground, therefore, saying that many statements are implied in the sense that they are suggested (whether the speaker intended to do so or not) by the tone of voice and the juxtaposition of the words used. Thus, “They had children and got married” suggests, though it does not state, that they had the children before they were married; any normal user of the English language would tend to construe it thus. And it is surely no overstatement to say that Swift’s Gulliver’s Travels suggests that the author was misanthropic or that the novels of the French author Marcel Proust suggest a pessimistic view of love and other human relationships close to that of the German philosopher Arthur Schopenhauer. Serious readers of literature will become increasingly sensitive to what is suggested in the works they are reading.

But, once again, the importance of the suggested statements, even when they are true, in no way shows that they must be accepted as true by readers if they are to value them as works of art. Are sincere Roman Catholics who find Dante’s worldview congenial and Lucretius’s repellent committed to saying that Dante’s is the better poem? If so, they may be accused of confusing his moral and theological judgments with his aesthetic ones. Still, it should be noted that there are some critics who believe that if two works of literature are both equal in excellence on all counts yet one presents a true view of reality and the other fails to, the one presenting a true view is better—better even as a work of art—than is the other one.

There is, however, another way of talking about truth in literature that is not or is not as obviously connected to propositions. A characterization in a novel or drama is spoken of as being true to human nature, true to the way people actually speak or behave or feel. No matter that Becky Sharp—in the English novelist William Thackeray’s Vanity Fair (1847–48)—is a fictional character, it would be said, as long as she is depicted as a person of a certain type would behave, she is being depicted truly; truth in fiction does not mean truth of the statements (for the statements in Thackeray’s novel describing her are false) but truth to human nature.

But what exactly does “truth to human nature” mean? The criterion is as old as Aristotle, who wrote that poetry is more true than history because it presents universal truths whereas history gives only particular truths and that poetry (dramatic fiction) shows how a person of this or that kind probably or necessarily would behave (or think, or feel). This criterion, however, is too vague as it stands: what is probable or plausible behaviour in one person is not in another, and what is probable in one set of circumstances is not so in another. The test of truth to human nature would be roughly as follows: Would a person such as has been described thus far (in the novel or drama) behave (or think or feel or be motivated) in the way that the author depicts this character as behaving in the circumstances described? It is often very difficult to decide this question, because knowledge of human beings is insufficient or because the dramatist has not provided enough clues. Still, once readers or critics are convinced that the character described would not have behaved as depicted by the novelist, they may criticize the characterization (at least with regard to this bit of behaviour or motivation) as implausible. If a character who has been described as spending years working toward a certain goal is represented by the novelist as abandoning it once it is within sight, the reader will have considerable reservations about this delineation unless the author has depicted the character as being unstable or masochistic or in some way as being the kind of person who might in these circumstances do this kind of thing. It is true that there are people in the world who abandon their goals within sight of them after years of labour, but the conviction must be implanted that the character already presented by the novelist belongs to this classification or the behaviour will seem reasonless and unmotivated.

Is truth to human nature aesthetically relevant? That is, when present does it make the work of literature better and when absent or flawed does it make the work worse as literature? Here again there would be some difference of opinion, but a very large number of critics and aestheticians, in the tradition of Aristotle, would say that it matters aesthetically a great deal. Novelists do not have to be true to geography or history or astronomy, but they must be, as the 19th-century American author Nathaniel Hawthorne said of all literary artists, true to the human heart. A literary artist may tamper with all the other truths with impunity but not this one: the characters the artist creates must be convincing, and they will not be convincing if they are not depicted as having anger, love, jealousy, and other human emotions that real people have and in pretty much the contexts in which real people have them. If a novelist’s characters are not motivated in much the way that human beings are motivated, the reader will not even be able to understand them—they will be alien and unintelligible. Even when a writer—such as the British author Kenneth Grahame in The Wind in the Willows (1908)—depicts animals as central characters in novels, however much they may differ from human beings in external appearance, they must psychologically be presented as human beings—how else and in what other terms could their behaviour and their motivation be understood? Such, then, are the reasons for saying that whatever else literary artists do, their depictions must be truthful to human nature.

Can works of art other than literature possess truth to human nature? It would seem that in a limited degree they can. Motion pictures and operas and other mixed arts clearly can, but they employ words, and literature is a principal ingredient in them. But what of arts that employ no words at all? Painting and sculpture, not being temporal arts, cannot depict action, and action is all-important in the representation of human character. These arts, as noted earlier, contain depictions of persons (real or imaginary) only on a knife-edge of time. Still, sometimes something may be inferred even from a knife-edge. The late self-portraits of the 17th-century Dutch artist Rembrandt do seem to reveal an agonized yet sometimes serene inner spirit, suggesting that there are flashes of human insight to be found in depictions of human beings in visual art. As for musical art (music without the accompaniment of words), it contains nothing that could be called depiction, not even depiction on a knife-edge of time, and, if this is so, there can be no such thing here as true depiction or false depiction. Music may be expressive of human feelings, in the sense already described, but this is a far cry from saying that it contains depictions that are true to human nature.

Even if truth to human nature in the depiction of character is aesthetically relevant (which many would question), to say this is still far from saying that it is the only criterion for excellence in works of art, or even that this is the principal thing that art gives or its main excuse for being. To go so far would be to discount colour and form and expressiveness as criteria for excellence in art, and this virtually no one is willing to do. It would seem, then, that in no case is truth (even truth to human nature) necessary in works of art, seeing that entire genres of art, such as music, exist without it and that, even when it is present and when its presence increases the merit of a work of art (which again many would deny), it is only one virtue among many. Thus, the view that the purpose or function of art is to provide truth is quite surely mistaken; perhaps the person who wants truth and is indifferent to the presence of anything else had better turn to science or philosophy rather than to the arts.