Infinite Canons: A Few Axioms and Questions, and in Addition, a Proposed Definition

Todd M. Compton – June 13, 2016

© 2016 Todd Compton.


Recent years have seen the appearance of a number of books affirming the existence of definite, authoritative “canons” of music, film, art and literature. For example, in literature, there is Harold Bloom’s The Western Canon: The Books and School of the Ages (1994). In music we have David Dubal’s The Essential Canon of Classical Music (2003). Screenwriter/critic Paul Schrader, in an influential article, sought to “define and defend the film canon.”[1] The German literary critic Marcel Reich-Ranicki edited a multi-volume anthology of German literature entitled The Canon: German Literature (Der Kanon. Die deutsche Literatur) (2002-2006). To assess this use of the “canon” concept, a short history of canon will be helpful.

The word “canon” is from the Greek kann: to quote the standard ancient Greek dictionary, a “straight rod, bar, esp. to keep a thing straight”; and then, metaphorically, “rule, standard”; and then “in Art, model, standard.”[2] The sculptor Polykleitos created a sculpture of the human body called “The Canon” because its dimensions were to be followed by other sculptors. He also wrote a treatise called The Canon.[3] This metaphor was applied to literary works by the time of rhetorician and historian Dionysius of Halicarnassus (c. 60 BCE-c. 8 CE), who referred to Herodotus’s history as providing “the highest standard [kanōn] of the Ionic dialect, Thucydides of the Attic.”[4] The word “canon,” singular, was at first applied to individual works, or aspects of works.

The Alexandrian canon.

The first well-known Western literary canon was the “Alexandrian canon.” It was produced by two head librarians of the Alexandrian library, Aristophanes of Byzantium (c. 257- c. 180 BCE), who became library director about 195 BCE; and his student and successor, Aristarchus of Samothrace (c. 220 – 145 BCE). It was a canon of authors, not works, and it was divided into genres. It included nine lyric poets, for example, and three iambic poets. The rhetorician Quintilian (c. 35 – c. 100 CE), our primary source for this canon, described it as an ordo or numerus – essentially a list.[5] Unfortunately, there are no extant details telling us how these lists were used. Aristophanes and Aristarchus were both famous scholars, as well as librarians. Their canon could have been a syllabus for beginning students; a list for traveling book-buyers; or an outline of authors deemed worthy of commentaries or textual analysis to be written by employees of the library or students of the librarians. In any event, it shows the process of winnowing literature, isolating a limited number of authors as the “best.” Quintilian does tell us that the librarians included no living authors on their lists, thus excluding Apollonius of Rhodes.[6]

A late manuscript, tenth century CE, purports to contain the actual lists.[7] If it is close to the actual Alexandrian canon, it is an unnerving document, for most of the authors listed are close to forgotten now, and none of their complete works are extant. We know many of these poets only through fragments. We have only one complete poem by the great poetess Sappho, the only woman to show up in this canon.[8] The rest of her many poems are available only as fragments. Of the fourteen comic poets listed, there are only two who have complete plays extant, Aristophanes and Menander, and the latter, the greatest author in New Comedy, has only one nearly complete extant play, The Grouch (Dyskolos), which did not come down to us in multiple manuscripts; it was discovered on a papyrus in Egypt in 1952. In many ancient lists of the best authors, Menander ranked just behind Virgil and Homer.

However, the word kanōn was not used to describe the “Alexandrian canon” during the time of Aristophanes and Aristarchus. Modern scholars applied that term to it.[9]

The Alexandrian canon became very influential. It was not followed slavishly by Greco-Roman critics, as there is considerable variation in individual rosters of authors in different lists; but its organization by genres, for example, continued into Roman criticism.


One method of canonizing authors is by symbolically deifying them. In Jean Auguste Dominique Ingres, Apotheosis of Homer, 1827, the blind poet is crowned by an angel. He is surrounded by ancient and modern authors, from Pindar, Orpheus and Virgil to Moliere, Shakespeare and Tasso. Standing to the far right, in a brown robe, is Aristarchus of Samothrace, flanked by Aristotle and Alexander the Great. Courtesy  Wikimedia Commons.


Christian canons.

The first actual use of the word kanōn applied to a group of authoritative, exemplary works occurs in the first centuries of the Christian church. As the early Christians went through the process of judging which books of scripture were “best,” most inspired and most authentic, they explicitly used the word kanōn at least by the time of Origen (ca 184-ca 254 CE). He wrote, “No one should use for proof of doctrine books not included among the canonized scriptures [canonizatas scripturas].”[10]

In the early centuries of the church, the canon varied considerably, with certain books, such as Hebrews or Revelations, considered suspect and left out of many lists. However, the canonization process was finalized by a series of ecclesiastical conferences. In the Council of Laodicia, in 363 CE, the authorized New Testament books were referred to as “the canonical works,” ta kanonika. The third Council at Carthage in North Africa, held in 397 CE, listed the New Testament books as used presently. In a summary of the proceedings of this conference, we read:

It was also determined that besides the Canonical Scriptures [Scripturas canonicas], nothing be read in the Church under the title of divine Scriptures. The Canonical Scriptures [Canonicae Scripturae] are these: [The Old and New Testament books are listed]. Let this be made known also to our brother and fellow-priest, Boniface [of Rome], or to other bishops of those parts, for the purpose of confirming that Canon [isto canone], because we have received from our fathers that those books must be read in the Church. Let it be also allowed that the Passions of Martyrs be read when their festivals are kept.[11]

Despite this, the canon of scripture presently varies among the Christian churches. For example, the Catholic canon of scripture differs from Protestant scripture by the inclusion of the Apocrypha in the Old Testament. So in Christianity, we have canons as opposed to “the canon,” a Catholic canon and a Protestant canon. The Catholic church did not finalize its canon until the Council of Trent (1545-1563 CE), a convocation called as a response to the Protestant Reformation.

We should keep the chronological caveat in mind; a future church council might adjust the list of best works, adding some, removing others. For practical purposes, Luther rejected the Epistle of James, calling it an “epistle of straw” and “unworthy of the apostolic Spirit.”[12]

One of the early criteria for canonicity of the New Testament books was authorship by one of the early apostles. So when the apostolic authorship of books such as Hebrews and the Apocalypse was questioned (by Erasmus and many others), these books’ place in the canon was being questioned, in an absolute sense.[13] However, in a practical sense, these books had been accepted into the Christian canon, and have never been formally decanonized.[14]

Thus the canon of New Testament scripture, after a period of uncertainty, was finalized by meetings and votes of official church leaders. The canonizing was specific: for a limited community, the Christians, certain books were authoritative, the best, exemplary. 

While the process of winnowing the best secular writings had been a feature of the Greco-Roman traditions, and the medieval and Renaissance periods produced many lists of best works, the word canon was not applied to these lists until recent centuries. A parallel metaphor is secular masterpieces viewed as “classics,” works that are the best in their genre, examples for other works to aim at. Classic is from Latin classicus, “of the highest class of Roman citizens, of the first rank.” Thus, a classic is, according to Websters, “Of recognized value: serving as a standard of excellence. b Traditional, enduring.”[15] The great works of the Greeks and Romans were called classics because they were of the first rank, the best, “serving as a standard of excellence,” enduring because of their high quality, a concept directly parallel to the idea of canon as a group of exemplary works tested by time. “Classical music” was not music of a certain genre at first; originally, it simply meant the best music, music of the first rank.[16] 

Canons of literature.

The word “canon” came to be applied to catalogues of best literary works in 1768, as the classicist David Ruhnken described the lists of Aristophanes and Aristarchus as a canon.[17] This usage came to be accepted among classicists and beyond. As Gorak writes, “it became common, if sometimes controversial, to extend the application of canon to any list of valuable inherited works. . . To this day, the idea of a canon as a list of standard texts relating to a particular culture or area remains a vital one.”[18]

However, the “Alexandrian canon” was fundamentally different from the canon of New Testament scripture. First, it was not called a canon at the time it was produced, or in antiquity. Second, we do not know how the lists of the “Alexandrian canon” were used, though clearly they had something to do with the Alexandrian library or with scholarship.[19] Third, they were the product of one man, at first, revised by a second. Therefore, fourth, they were not finalized by meetings of leaders. Finally, they were not authoritative for a specific broad community, beyond the library of Alexandria, though they were influential. One could argue that the idea of a canon of literature has no validity, given these differences.[20]

In some ways, however, the usage was attractive. It did borrow the idea of a select group of best, exemplary works and apply it to the masterpieces of secular literature. After Ruhnken, other classicists adopted his reference to the Alexandrian canon, and gradually the concept became widespread. In 1827, Robert Pollok referred to “The lofty seat of canonized bards.”[21] By 1918, the literary critic Van Wyck Brooks could speak of “the accepted canon of American literature.”[22] In America, in 1927, Henry Canby, editor of the Saturday Review of Literature, wrote a column entitled “An American Canon.”[23] Five years later, Carl Van Doren felt there was a need for a “New Canon” of American literature.[24]

In 1936, French critic Valery Larbaud referred to “canonical English authors.”[25] British critic Q. D. Leavis broadened Brooks’s scope in 1943 and referred to “the canon of accepted Literature.”[26]

1948 saw the appearance of Ernst Curtius’s European Literature and the Latin Middle Ages, which included an influential chapter on canon formation in the Middle Ages.[27]

Current usage of “canon of literature” (or music or art) is reflected in contemporary dictionaries. Merriam-Webster defines canon, in the literary sense, as “a sanctioned or accepted group or body of related works <the canon of great literature>.” In a draft addition to the Oxford English Dictionary, a canon, in literary criticism, is:

A body of literary works traditionally regarded as the most important, significant, and worthy of study; those works of esp. Western literature considered to be established as being of the highest quality and most enduring value; the classics (now freq. in the canon). Also (usu. with qualifying word): such a body of literature in a particular language, or from a particular culture, period, genre, etc.[28]

Here, the OED introduces a chronological aspect, as works in a canon are “traditionally” regarded as best (and the definition also mentions works of “most enduring” value). This only leads to more questions, such as, how long a process of traditional acceptance is required? OED also states that canonical works “are considered to be established” as best. This is very wishy-washy language. Instead of works “established” as best, we have works “considered” to be established as best. Considered and established by whom?[29]

So a canon, by these definitions, is a list of great works that are “accepted,” in Brooks’s and Leavis’s usage. The OED thinks the works of the canon are “traditionally regarded” as the best. “Regarded” parallels “accepted.” The question now is: who “sanctions,” “accepts” or “regards” the group of works in a canon? Who decides what is best? How big is such a group, and what is its composition: librarians, academics, critics, writers, book reviewers, or just plain readers? Whatever group, large or small, “sanctions” or “regards,” can each member of the group reach absolute unanimity as decisions are made on a limited number of works that are “best”—classical, “canonical”?

In addition, “those works of esp. Western literature” is silly. There have been canons of Asian and African literature also, and holistic canons including all of world literature. A canon of Western literature is certainly allowable, but “the canon of accepted Literature,” in Leavis’s phrase, does not suggest Western literature only.

Critics such as Brooks and Van Doren, and the OED, seem to think that there is a pure, precise canon existing somewhere, like one of the Platonic forms, and assume that everyone simply understands and accepts this canon (whose components are conveniently not specified). The reality is that there is no one literary canon, no one American canon or Western canon, no one musical canon and no one film canon. The field of literature has no organized overall organization that could send its leaders to a convocation to vote on a list of authorized books. One critic cannot prescribe an authoritative canon of literature, by the definition offered by OED and the critics listed above. The reality is, in literature, art and music, we have an infinity of limited canons, sometimes agreeing on certain authors, but often disagreeing.

For example, one critic, or scholar, or writer, can certainly delineate the authors he feels should make up “the canon.” But the next day, another critic, or scholar, or writer can publish a rival list that disallows many authors on the previous list, and adds authors the previous list did not include. This is exactly what happened when John Ruskin assessed Sir John Lubbock’s list of “best hundred books” in 1886.[30] Lubbock originally gave a speech on reading the best literature, and then a number of lists deriving from it ensued. Asked to respond to Lubbock’s list, Ruskin wrote, “Putting my pen lightly through the needless—and blottesquely through the rubbish and poison of Sir John’s list—I leave enough for a life’s liberal reading—and choice for any true worker’s loyal reading.” The following graphic shows dramatically how two distinguished men of letters agreed on a canon of literature:

Ruskin on Lubbock


Proponents of “the canon” may argue that it is the product of all “right-thinking” critics and scholars in a certain field. But however many critics one may choose, and whichever critics one chooses, they will all disagree on exactly which works would make up “the canon.” And if you polled fifty leading critics or authors on the hundred greatest works of literature, and counted their votes as in an election, the resulting list of hundred great works would certainly be of interest, and express the “opinio communis” of that group. But if one were to poll another fifty leading critics or authors, their list would certainly vary from the first list in many points.

There is thus a harmonizing and discordant infinity of canons.

The usage of Brooks and other critics, and the definition of the OED, must then be rejected. There is no such thing as an “accepted” canon of literature, unless one specifies exactly which group “accepts”, and for what group that specific canon is “accepted.” And even in this case, the “acceptance” is only for a limited group. In addition, the chronological aspect of canonical endurance cannot be specified.

A proposed definition of canon.

I propose a more limited, less Platonic, definition of canon. A canon is a list of works considered best and exemplary, propounded by an individual or group, often for a communal purpose.[31] This will allow us to include lists of individuals, such as Aristophanes of Byzantium or Harold Bloom (though such canons will be called the Aristophanes canon or the Bloom canon). The term “personal canon” has started to come into general use.[32] By the OED definition, this is not valid (since the OED implies one collective canon, created over time); by my limited definition, it is.  

This definition will also include groups of canonizers, such as the convocations of early Christianity or the members of an English department creating a reading list for graduates or a poll of critics and/or writers, or polls organized by magazines, websites or book clubs. By my definition, no canon is absolute, though a canon can be accepted as binding for a specific purpose or community.

However, my definition will not allow us to refer to any work as “canonical” in an absolute sense, and there is no such thing as “the canon” of anything, unless we are referring to a specific limited canon. There is no such thing as “the accepted canon of American literature,” “the canon of accepted literature,” “the literary canon,” “the Western canon,” or “The Essential Canon of Classical Music.[33] The term “canon formation” also has no validity, unless we are talking about the formation of one specific canon created for a specific community or purpose. To speak of the formation of “the canon of English literature,” for example, presupposes that there is some pure, precise, timeless list of the favored works in the English tradition—accepted by all authors, critics, readers, librarians and teachers. Instead, we have many canons of English literature, with some shared elements and some contradicting elements.

Having made this short historical survey and offered a definition of canon, I now turn to some observations and questions on what canons are and should be.

Canons are often communal.

Canons are often created by groups of people for specific purposes. We have seen that the canon of New Testament books was finalized by a series of meetings of church leaders. Other examples of communal canons are the syllabus, a list of books to be read in a class, or reading lists for students in literature programs, which the students will be asked to know to prepare for written tests.

A canon is often a list with a social purpose. For that limited purpose and social situation, the canon is fixed. For a certain literature survey class, five specific books will be on the syllabus.

But none of these canons are absolute; they are limited by their specific social purpose. For a literature survey class of modern American fiction, two teachers may assign five entirely different books. And different religious groups will have different canons.

A canon is both inclusive and exclusive.

There is an unavoidable survival-of-the-fittest aspect to the process of reading, as no one can read everything, and so we have a natural desire to read the best works, given the limited time allotted to each human being in a lifetime. So reliable guidance on what the best books are is a valuable contribution. Recommending any book, or a group of books, by definition leaves millions of books off the list. Jonathan Swift once described this competitive aspect of literature as the Battle of the Books— “An ACCOUNT of a BATTEL between the Antient and Modern BOOKS in St. James’s Library.”[34]


Illustration from Jonathan’s Swift’s Battle of the Books.


This same survival-of-the-fittest aspect of literature is present in the process of formulating canons, deciding what is best in a genre; for every work included in a canon, there are hundreds, thousands, millions, an infinity of works not included. Many of these may be superb works; they are simply not the best. A canon is by definition a limitation.

In creating a small canon, say of ten books, if we love a hundred books, we are forced to exclude ninety books. Thus, creating a canon can be a painful process. On the other hand, the process of delineating a canon can be a joyful experience, as one seeks to internalize and celebrate the highest reaches of the human spirit. So engaging in the canonical process is always ambiguous—both destructive and creative, painful and joyful.

Some have criticized canons as exclusionary. They are by definition both inclusionary and exclusionary. If you pick the best, that which is not the absolute best is by definition not included. But experiencing the best of the best is an entirely worthwhile goal.

We should also not forget the value of many works that may be seen as of secondary stature—the less important books of great writers, or the inspirations of writers who may be of lesser stature than the titans, but who are still entirely worthwhile. For example, I would not place Jerome K. Jerome’s comic masterpiece Three Men in a Boat on my list of twenty greatest literary accomplishments in the history of the world. In one sense, it is second-rate. Nevertheless, I have read and reread it repeatedly in my life. By the standard of pleasurable rereading, it is higher on my personal canon than is many a weighty tome.[35]

How big is a canon?

A canon by definition limits to a certain number of the best works. So we may ask: How big is a canon? 1000 works? 100 works? 50 works? 25 works? 20 works? 10 works? 5 works?

There is no objective answer to this question. The answer in part depends on how the canon-receiver, or the canonical community, needs to use the canon. Thus:

The size of a canon depends to some extent on the purpose of the canon.

If you want to find a good list of books to read privately, you may want to start with ten or fifty or a hundred books, and continue adding to that number throughout your lifetime.

A canon of a thousand books, arranged alphabetically, thematically or chronologically, as in the Guardian’s article with the absurd title, “1000 Novels Everyone Must Read,”[36] might not serve such a reader well, as he or she may want to read the “best” five, ten or twenty books first, and a canon of a thousand gives no help for this (unless it is ranked).

And five and ten are not the lowest possible canonical numbers: T. S. Eliot created a canon of two “modern” writers: “Dante and Shakespeare divide the modern world between them, there is no third.”[37] In fact, Keats seemed to narrow that down to Shakespeare alone: “I never quite despair and I read Shakspeare—indeed I shall I think never read any other Book much [. . .] I am very near Agreeing with Hazlit that Shakspeare is enough for us.”[38]

If you are creating a canon to educate undergraduates on the high points of the western cultural tradition, in a class lasting four months, this has become, as it were, a social canon, part of the relationship between teacher and students. Because of the four month limitation, it might be a very selective canon—possibly four to eight full books, in addition to a selection of short works.

If you are a librarian opening up a new, small branch library, and selecting a core number of select books, a list of a thousand books might be useful. A thousand books is still a limited canon, compared to the millions of books published throughout history (and the huge quantity published every year).

One method of creating a canon is to collect short works in a book. The Bible is such an anthology—its name comes from Biblia, “Books”—and the history of the anthology reaches back to Greco-Roman times. The word comes from the Greek word anthologia, “collection of flowers,” and one of the earliest known anthologies, The Garland (Stephanos) by Meleagros of Gadara (1st century BCE), characterized each poet as a flower. This led to the well-known poetry anthology, The Greek Anthology.[39] If you are creating an anthology—a book-size canon of the best short works—this selection will be limited by the number of pages that is practical for the formation of the book as physical object, and the limitations given by your editor. The contents can contain only comparatively short works, short books created from scrolls, short novels, short stories, essays, poems, though a novel is sometimes included in a long anthology of fiction.

Is a canon composed of works or creators? Do we pick Hamlet or Shakespeare?

There is no objective answer to this question. Some authors have one overwhelming masterpiece, like Dante and the Divine Comedy, or Proust and In Search of Lost Time. However, if we decide to create a canon of works, can one play, say, Hamlet, stand beside the towering expanses of the Divine Comedy? Probably not. However, in a canon of creators, Shakespeare as represented by his entire corpus—including Hamlet, King Lear, Macbeth, Othello, Romeo and Juliet, Midsummer Night’s Dream, The Tempest, As You Like It, Twelfth Night, the historical plays, the sonnets—can arguably stand beside Dante.

Many canons, starting with Aristophanes of Byzantium, include only the creator. Andrew Sarris, in his influential book The American Cinema: Directors and Directions 1929–1968 (1968), created a personal canon of American directors.

If we create a canon of creators, not works, we will need to create an additional canon for each of those creators. In other words, what is the best of Shakespeare? Of Goethe? What five to twenty works of each of these would we choose?

Therefore, there are advantages to creating a canon of works, as such canons will include the creator. In addition, sometimes an author’s vast corpus can be intimidating, and much of it is far inferior to his or her best—if we select the work, we have by definition also created a useful subcanon of an author.

Canons can be limited by geography or chronology or some other special feature.

“Limited” canons are entirely valid, especially for readers interested in a particular focus. Canons of American novels, French short stories, or Nigerian music can be useful. The “Western” element in Bloom’s title, The Western Canon, represents a valid limitation, entirely open to the creators of any canon.

However, if you are looking for the best of the best, then “limited” canons give only a percentage of the “holistic” canon. They are incomplete.

Obviously, limited canons can be the product of ethnocentrism and narrowness. And limited canons, per se, can cut us off from some of the greatest works in literature. Leslie Schenk, in a review of Bloom’s The Western Canon, admires the book, but makes the point that a well-educated Japanese person is conversant in Japanese classics, Chinese classics and Western classics. In the west, we are usually conversant only with Western classics, at best.[40] In cinema, the idea of a Western canon would be laughable—ignoring the films of Akira Kurosawa, Yasujirō Ozu, Kenji Mizoguchi, Satyagit Ray, Abbas Kiarostami, Hou Hsiao-Hsien, Edward Yang and Zhang Yimou would separate us from some of the great cinematic experiences.

Poet, translator and critic Kenneth Rexroth, after an overview of translated Chinese fiction, asked,

What kind of novels are they? . . . they are great novels, very great novels. In fact I would say that The Dream of the Red Chamber and the Japanese Tale of Genji are the two greatest works of prose fiction in all the history of literature, and that all the others belong on anybody’s list of 100 Best Books. That they are not on the Hutchins-Adler list is an excellent indicator of the Western, Thomism-cum-Whiggery parochialism of Mr. Hutchins and Mr. Adler.[41] I am not trying to be odd or annoying. I am not saying something like “Sturge Moore is the greatest poet of the twentieth century.” I really do believe that these are the two best novels in the world. Furthermore, there are not many people who are familiar with them who do not agree with me.[42]

Thus, if we confine ourselves to Western literature, we may be missing the two greatest novels in the history of mankind.

If a holistic view of canon is important to you, you will include Japanese, Chinese, Indian, French, English, American, South American, African, Tibetan, works in your personal canon, and works from all time periods. The holistic canon may require some ambition and energy both on the part of canon provider and canon user. On the other hand, some of the marvels of literature, such as The Bible, Rumi, The Bhagavad-Gita, The Arabian Nights and The Dream of the Red Chamber, will be technically excluded by limited Western canons.[43]

However, as the above short list shows, translations of non-Western works are an interesting problem. Is a translation of a non-Western work a re-creation, thus a Western work? Few would disagree with the proposition that the King James Bible is a foundational work in English literature for the beauty of its prose, its poetry, and compelling storytelling, and it is absurd to consider a history or canon of English literature without it. The Rubaiyat of Omar Khayyam, as “translated” by Edward Fitzgerald, has been considered a work of English poetry; though it took great liberties with the Persian text, it certainly often reflects Omar Khayyam’s poem. Do we disallow it from our canons of English literature? Rexroth argues that Arthur Waley’s translations from the Chinese are themselves profound works of art.

By this argument, non-Western works can enter Western canons at the time of an good translation. If we use influence and impact as a criterion for inclusion in a canon, The Arabian Nights is certainly part of the Western tradition. Its impact on writers as diverse as Tolstoy, Proust, Borges and Dickens shows that it is technically impossible to entirely separate “Western” and “non-Western” canons.

As is well known, American poet and essayist Ralph Waldo Emerson was profoundly influenced by Hindu religio-philosophical works. On October 1, 1848, he wrote in his diary, “I owed . . . a magnificent day to the Bhagavat Geeta.— It was the first of books; it was as if an empire spake to us, nothing small or unworthy, but large, serene, consistent, the voice of an old intelligence which in another age and climate had pondered and thus disposed of the same questions that exercise us. Let us . . . cherish the venerable oracle.”[44] His famous poem “Brahma” is of course based on Hindu concepts, and some lines of this poem are almost direct quotes from the Bhagavat-Gita.

Tolstoy was profoundly influenced by Buddhist texts, the Tao Te Ching and the Hindu Vedas. For example in a list of “good, supreme art,” he mentioned the Iliad, the Odyssey, the stories of Isaac, Jacob, and Joseph, the Hebrew prophets, the psalms, from the Old Testament, the Gospel parables, the story of Sakya Muni (Buddha), and the hymns of the Vedas.[45] Except for the Iliad and Odyssey, all of this is non-Western.

No two canons are the same.

This is a verifiable, unalterable, inescapable fact. André Maurois wrote, “In literature as in love, we are astonished at what is chosen by others.”[46] And in the same vein, two anonymous maxims: “De gustibus non est disputandum.” (“You can’t argue about taste”) and “‘There’s no accounting for tastes,’ said the old lady as she kissed the cow.”

All canons will conflict with each other (with the exception of the “dittohead” phenomenon, in which one person simply accepts another’s canon unthinkingly and uncritically).

Let’s take Joyce, for example. He seems a star comfortably ensconced in the firmament of many western canons. Bloom devotes a chapter to him in The Western Canon, and believes he is a worthy successor/antagonist to his central “canonical” author, Shakespeare. In the 1998 Modern Library editors poll of twentieth-century novels in English, Joyce’s Ulysses is ranked at number one.[47] He has been acclaimed by many fellow writers, and is accepted as one of the dominant novelists of the twentieth century by the great majority of modern critics.

Yet Virginia Woolf, whose To the Lighthouse comes in at 15 in the Modern Library poll, and whose Orlando is on Bloom’s short list of 25 or so greatest works in “the western canon,” wrote, “I finished Ulysses, & think it a misfire. Genius it has I think; but of the inferior water. The book is diffuse. It is brackish. It is pretentious. It is underbred, not only in the obvious sense, but in the literary sense.” She regarded the book as “a memorable catastrophe—immense in daring, terrific in disaster.”[48]

Jorge Luis Borges, also on Bloom’s short list, spoke of “those two vast and—why not say it?—unreadable novels, Ulysses and Finnegan’s Wake.”[49] Rereadability is often cited as a canonical quality in literature; but if you can’t read a book once, that certainly removes it from consideration as rereadable. Borges referred to Ulysses as a “failure” because its characters never really come alive—“you don’t know them,” as you know the characters in books by Dickens and Stevenson.[50]

Nobel Prize winner Isaac Bashevis Singer, a writer with four books in Bloom’s appendices, said, “To me, Joyce’s Ulysses is almost boring. I don’t enjoy this kind of abstruse writing where style is dominant and the story only serves as a container of the style, a frame.”[51]

One could find as many or more quotes from distinguished writers who admired Ulysses.[52] One could also analyze why Woolf, Borges, and Singer did not warm up to it—Singer rejected it for its lack of “story.” Borges was not sympathetic to the genre of the novel (though he included some novels and novelists, such as Twain, Dickens and Conrad, in his personal canon), and criticized “padding” in long works of fiction.

But my point is simply that, if Woolf, Borges, and Singer deserve respect as judges of quality in literature, we have to admit that there is no consensus that Ulysses is the best novel in English in the twentieth century, or even that it is a good book.

Another example is Shakespeare. For Bloom, Shakespeare is the center of “the Western canon.” Certainly, this is plausible. In a recent international newspaper poll, readers were asked to choose the greatest author not in their country. Shakespeare easily won the poll, even with Britain unable to vote for him.[53] Yet two great writers, Tolstoy and George Bernard Shaw, emphatically excluded Shakespeare from their personal canons. Tolstoy despised Shakespeare, whom he regarded as merely an “ordinary” writer, with an absurdly inflated reputation. The novelist wrote an article on Shakespeare’s plays, “On Shakespeare and the Drama,” “to save people the necessity of pretending that they like them.” He expressed distaste for all the famous plays: Julius Caesar, Romeo and Juliet, Othello, King Lear, and Hamlet. “What a crude, immoral, vulgar and senseless work Hamlet is,” he wrote in 1896. “The whole thing is based on pagan vengeance; the only aim is to gather together as many effects as possible; there is no rhyme or reason about it. The author was so concerned with the effects that he didn’t even bother to give the main person any character.” As he looked at this play’s reputation, he wrote, “I never understood so clearly the utter helplessness of the crowd in making judgments, and how they can deceive themselves.”[54] Presumably, Tolstoy probably meant the “crowd” of intellectuals rather than the mass of common people. He is directly attacking the critical opinio communis.

In the essay on Shakespeare, he wrote,

After reading, one after the other, the plays considered to be his most beautiful—King Lear, Romeo and Juliet, Hamlet and Macbeth—not only did I derive no pleasure from them, but I felt an overpowering repugnance, a boundless tedium, and I wondered whether it was I who was mad, to find empty and offensive these works that are held by all cultivated people to be the summit of perfection.

George Bernard Shaw also rejected Shakespeare as even a good dramatist, let alone a great one. He wrote,

With the single exception of Homer, there is no eminent writer, not even Sir Walter Scott, whom I can despise so entirely as I despise Shakespeare when I measure my mind against his. The intensity of my impatience with him occasionally reaches such a pitch, that it would positively be a relief to me to dig him up and throw stones at him, knowing as I do how incapable he and his worshippers are of understanding any less obvious form of indignity.[55]

Shaw does mention Shakespeare’s gift for storytelling, his skill with language and characterization. Nevertheless, he wrote,

There are moments when one asks despairingly why our stage should ever have been cursed with this “immortal” pilferer of other men’s stories and ideas, with his monstrous rhetorical fustian, his unbearable platitudes, his pretentious reduction of the subtlest problems of life to commonplaces against which a Polytechnic debating club would revolt, his incredible unsuggestiveness, his sententious combination of ready reflection with complete intellectual sterility.

There are also positive choices in various canons that stray far from the opinio communis. One evening, Gabriel García Márquez, Bill Clinton, Carlos Fuentes and William Styron had dinner together, and each named his favorite book. Styron chose Huckleberry Finn; Clinton the Meditations of Marcus Aurelius, and Carlos Fuentes Faulkner’s Absalom, Absalom. All of these would have broad support among academics and literary critics. However, Márquez named, not a book by Joyce, or Hemingway, or Faulkner, but Dumas’s The Count of Monte Cristo.[56] Dumas is nowhere mentioned even in the appendix of Bloom’s Western Canon; yet one of the great writers of our century put Monte Cristo at the very top of his personal canon of books.[57]

At other times Márquez selected War and Peace and Moby Dick as the best novel in world history. Dracula was another favorite of his. Count Dracula, with Edmond Dantes (from Monte Cristo) and Gargantua, were his three favorite fictional characters.[58]

A Canon of Character – Gabriel Garcia Márquez’s three favorite fictional characters: Dracula, Edmond Dantes from Count of Monte Cristo, and Gargantua.


In 1891, Tolstoy prepared a list of books and stories that had influenced him significantly and these were ranked by influence that was “great,” “very great” and “enormous.”[59] In the “enormous” category, along with the story of Joseph from the Old Testament and the Sermon on the Mount, Rousseau’s Confessions, Dickens’s David Copperfield, and Hugo’s Les Misérables, is Gogol’s short story, “Viy.” This choice is extremely idiosyncratic, as “Viy” is a wild tale of witchcraft, not, seemingly, one of the great evocations of the human spirit. (Gogol’s much more famous and influential “The Overcoat” is only in the “great” influence category, for Tolstoy.)

Jorge Luis Borges often described Dante’s Divine Comedy as the greatest work of literature ever written; however, his favorite single book in his home library was a book of poetry—not Shakespeare, Keats, T. S. Eliot, Wallace Stevens or Yeats, but the collected poems of Rudyard Kipling. According to his friend Willis Barnstone, Borges virtually had the book memorized.[60] Few academic critics would rank Kipling’s poetry as great, both for political and aesthetic reasons; however, for Borges it was at the very center of his personal canon.

Canons change chronologically, from generation to generation, century to century.

Canons will vary over time. New works are constantly appearing, and some of these will deserve to be included in personal and collective canons. In addition, fashions in taste change. The canon of a typical nineteenth-century academic in Victorian England will be quite different from the canon of a twenty-first-century academic—aside from the obvious lack of twentieth century writers on the earlier list. Basic aesthetic and world views have changed. (This is one reason that the OED definition of canon lacks validity; inclusion on “the western canon” by “tradition,” over time, will never occur in a unified way.)

To take one example, the short story. In the nineteenth century, the great majority of short stories developed character through a series of events, as found in the work of Scott, Poe, Kipling and Stevenson. Canons of short story—lists as well as anthologies—were composed of such stories with well-defined beginning, middle and end. In addition, these lists included many fantasy and ghost stories, such as Gogol’s “The Overcoat” or Scott’s “Wandering Willie’s Tale.”

In our century, taste has inclined toward the short story without much event, capturing a mood, with an emphasis on style, and strict realism, as found in the work of Joyce, Hemingway and Carver.

It is easy to posit that our taste has matured. But what if the taste of the next century has substantially changed and “matured” beyond ours?

Broad canons tallied from votes change over time. For example, the early film polls in the 1950s selected films such as Eisenstein’s Battleship Potemkin or De Sica’s The Bicycle Thief as the greatest film in history. The influential Sight & Sound polls, held every decade, began in 1952 with The Bicycle Thief at number one. After the fifties, most film polls, including the Sight & Sounds polls, selected Orson Welles’s Citizen Kane as number one. However, the latest Sight & Sound poll, in 2012, saw Hitchcock’s Vertigo dethrone Kane.[61] Battleship Potemkin is now at number eleven, and The Bicycle Thief has dropped to thirty-three. In 2002 the National Society of Film Critics published a book listing and praising “100 Essential Films,” but Bicycle Thief was not included.[62]

Charlie Chaplin’s standing in the Sight & Sound polls has undergone a remarkable decline. In the first Sight & Sound poll, two Chaplin movies, City Lights and The Gold Rush, tied for second place. In the 2012 poll, City Lights barely squeaks into the top fifty, in a tie with two other movies, and The Gold Rush has dropped off the canonical cliff, at number 154.

Directors Poll, Brussels, 1952

Critics Poll, Sight & Sound, 1952

Critics Poll, Sight & Sound, 2012

The Battleship Potemkin (Bronenosets Potyomkin)

The Bicycle Thief (Ladri di biciclette)


The Gold Rush

City Lights

Citizen Kane

The Bicycle Thief (Ladri di biciclette)

The Gold Rush

Tokyo Story (Tokyo monogatari)

City Lights

The Battleship Potemkin (Bronenosets Potyomkin)

The Rules of the Game (La Règle du Jeu)

Grand Illusion (La Grande Illusion)


Sunrise: A Song of Two Humans

Le Million (The Million)

Louisiana Story

2001: A Space Odyssey

Brief Encounter


The Searchers


Le Jour Se Lève (Daybreak)

Man with a Movie Camera (Chelovek s Kinoapparatom)


The Passion of Joan of Arc (La Passion de Jeanne d’Arc)

The Passion of Joan of Arc (La Passion de Jeanne d’Arc)

Hallelujah! (tie)

Brief Encounter (tie)

8-1/2 (Otto e Mezzo)

Man of Aran (tie)

Le Million (The Million)  (tie)


The Threepenny Opera (Die Dreigroschenoper) (tie)

The Rules of the Game (La Règle du Jeu)  (tie)


The Impermanence of Critical Taste: Remarkably, the two 1952 polls agree on the first four movies (in slightly different order). Three of these are silent movies. The 2012 poll has three silent movies, but not the silent movies in the top four of the 1952 polls. Many of the films in the 1952 polls, such as René Clair’s Le Million, are now comparatively forgotten.


Sir Walter Scott offers a good example of a similar decline in the literary arena. He is now viewed as an author of adventure stories for boys, of historical importance perhaps, but not to be compared with Dickens, Thackeray and George Eliot.  In a 2007 poll of writers, not a single book by Scott appears in the top 286.[63] However, we have seen that when art critic John Ruskin “edited” Lubbock’s canon, he crossed out Thackeray and Eliot and kept Scott, writing after his name, “Every word.” A remarkable tribute from a critic whom one would expect would be judging by high aesthetic standards.[64]

The great poet and novelist Goethe was another critic who ranked Scott among the great writers. In an informal conversation, he said:

“We read far too many poor things, thus losing time, and gaining nothing. We should only read what we admire, as I did in my youth, and as I now experience with Sir Walter Scott. I have just begun ‘Rob Roy,’ and will read his best novels in succession. All is great—material, import, characters, execution; and then what infinite diligence in the preparatory studies! what truth of detail in the execution!”

And again, “‘Walter Scott,’ said he, ‘is a great genius; he has not his equal; and we need not wonder at the extraordinary effect he produces on the whole reading world. He gives me much to think of; and I discover in him a wholly new art, with laws of its own.’”[65]

In an influential article, Alastair Fowler has noted how literary genres change over time. As new genres become dominant, they will transform canons as a result.[66]

Some works are misunderstood or ignored at first, and are rediscovered by later generations. Other works invert this; they are valued by contemporary audiences and critics, then subsequent generations lose interest in them. But for an individual work, or an author, that process could invert again.

Haphazard factors of preservation over time often help define a canon.

The canon of Sophocles is limited to seven complete plays, though he wrote 123 of them. Were the seven survivors preserved because they were his best, most popular plays, and so more copies of them were made than of his lesser plays? Sometimes it seems clear that the works that had the most impact were copied (by hand) the most, were the most read, and survived in multiple manuscripts, thus ensuring their survival, and inclusion in canons.[67]

However, at other times there are unexplainable vagaries in the preservation of manuscripts. It is possible that the seven extant plays of Sophocles were preserved because at some point only one collection of scrolls escaped destruction.

Most would accept that Sappho is one of the great Greek literary artists; yet we have only a single short complete poem written by her. All of her books of poetry, and all of her complete poems (with that one exception), have been lost.

Many great works of Greco-Roman literature were preserved only as rhetorical teaching tools.

Books in dry climates have survived, while books in damp climates have not.

Tischendorf discovered one of the oldest copies of the New Testament in a basket of fuel at a monastery near Mount Sinai.

So there is an element of chance as well as of quality in the preservation of the best works.

Every medium of art has its idiosyncrasies. The movie, before the advent of the VCR and DVD, was dependent on public showings, and the only way to see older films was on TV or in revival theaters. Needless to say, the choice on TV, before cable and VCR, was meager (especially for foreign films), and many communities did not have revival theaters. Film stock was very flammable, and there are a number of lost early movies. Now, however, many classic films are widely available on DVD, on the internet, and on cable television. More people can enjoy great movies, and can judge which movies are best.

Canons change chronologically for individuals.

In the case of an individual, a personal canon may change throughout his or her lifetime. The personal canon of Leo Tolstoy when he wrote War and Peace in 1869 is much different than his personal canon when he wrote What Is Art?  in 1897.

Goethe was deeply influenced by the classical Greek authors, but different authors influenced him at different times.[68]

George Bernard Shaw, whom W. H. Auden called “probably the greatest music critic who ever lived,” often lambasted Brahms, and his scathing attacks on him are frequently quoted (such as his claim that Brahms’s German Requiem could be “patiently borne only by the corpse.”). However, he later stated that his attacks on Brahms had been misguided, admitted that he had not understood Brahms’s new, unfamiliar, musical idiom, and wrote, “I apologize.”[69]

Film critic Jonathan Rosenbaum has given top tens of his favorite movies at different times in his career and the listings are all different (only a few movies are constant).[70] As one novelist told me, in giving me his list of favorite books, “this will change tomorrow.”

Canons differ geographically.

If a poll is based in France, say, there will be a natural tendency for it to have a French bias. One is simply so familiar with and attached to one’s own national literature that it is entirely natural to show such a bias.

For example, the National Endowment of the Arts in Britain sponsored a survey of readers in England, called the Big Read. In the top ten, nine of the books were written by Britons. In a similar survey in Germany, titled Das große Lesen, the top twenty-six included a number of books by Germans, including Perfume by Patrick Süskind at number 4, Thomas Mann’s Buddenbrooks at number 6, and Goethe’s Faust at 15. In the British Big Read, Perfume is at 71, and Buddenbrooks and Faust are not found at all. There were also similarities between the lists—such as The Lord of the Rings at number one in both surveys—but the German survey had a significant German component that was absent from the English list.[71]

A poll of a hundred French writers produced a poll dominated by Proust’s In Search of Lost Time (33 votes, almost three times that of the second place entry, Ulysses). Also in the top ten of this poll were de La Fayette’s The Princess of Cleves (tied for third place), Baudelaire’s Flowers of Evil (tied for fifth place), and Flaubert’s A Sentimental Education (at ten).[72] With the exception of In Search of Lost Time, none of these books appear anywhere in an American-based poll of 125 writers conducted by J. Peder Zane.[73] Another poll of French writers had Shakespeare and the Bible in the first two places, but the rest of the top ten were books by French writers—Proust, Montaigne, Rabelais, Baudelaire, Pascal, Molière, Rousseau, and Stendhal.[74] (The American poll included only two French writers in its top ten, Flaubert and Proust.)[75] None of these polls, American or French, included Don Quixote in their top ten, despite the fact that it is sometimes regarded as the greatest novel ever written.[76] However, a poll of Spanish writers conducted by El Pais put it at number 1, just before In Search of Lost Time.[77] Other notable books in Spanish in this poll’s top twenty are Borges, Ficciones (at 10), Poeta en New York by Federico García Lorca (at 11), and Pedro Páramo by Juan Rulfo (at 16).

One could argue that every canon is geographically flawed; or one could celebrate the unique “geographical” perspectives of every canon. For example, the El Pais poll might lead you to read Rulfo’s marvelous Pedro Páramo, not surprisingly one of Gabriel García Márquez’s favorite books. He bought dozens of copies to give to his friends, and read it so frequently that he once said he could recite the whole book, forwards and backwards.[78] He felt that the few hundreds of pages that Rulfo had published would be as enduring as the pages written by Sophocles.

Three books in the top ten greatest books of all time in a French poll. None of these made even a low  appearance in a similar America-based poll.


We can try to surmount the contradictions of individual canons by creating polls, collective canons.

In other words a group of people may simply vote on a canon. The group may be small—an academic committee perhaps, or an editorial board of a newspaper or publishing company—or large, a poll of hundreds or even thousands of readers or listeners.

Polled canons—voting on best works—may serve a sort of practical purpose. Such polls provide a opinio communis, a snapshot of taste frozen in time and space, that can be valuable. However, they are not binding at all, because each collective canon differs from all other canons, individual or collective.

A large collective canon, with many voters, can be viewed as more democratic than a small collective canon, or an individual canon. In some ways it is more useful than a personal canon, and in other ways less useful.

If one accepts the validity of polls—and I find them fascinating, though not authoritative in any way—there is still no absolute standard for judging which polls are the best. A poll with a thousand voters is as valid as a poll of ten voters. A poll of French literature is as valid as a poll of World literature.

The methodology of polls can obviously be flawed, however. I like polls that are transparent and show the voters’ ballots, and thus supply a number of personal canons as well as the collective poll. One can see if the voters were critics, academics, librarians, writers, or simply readers. The lack of transparency in some polls seems suspicious, and supplying the ballots removes that suspicion. For example, at the end of every year the rock music magazine Rolling Stone proposes the year’s ten best albums. However, for two years running their highest albums have been very quirky. One wonders if the editor of the magazine merely selected the top ten, because you can’t imagine those entries coming from even a small poll of critics. Without the ballots, you cannot tell.

I like polls that include a wide variety of voters, from different countries. Polls in which a small group of critics select, as participants in their poll, a limited group of critics whose taste is very similar to their own, have much more limited interest, for me.

I also like polls that show the points awarded to each entry. The Norwegian Book Club poll treated this information like a state secret on which the fate of the free world depended. They generously identified their number one novel—Don Quixote—but gave us no rankings for the other 99 books.[79] If you think a poll is worth doing, then release the details about the poll, the ballots, and the number of votes for each book.

The methodology of the Sight & Sound film polls has been exemplary. The editors make a serious effort to recruit voters from around the world (though English speakers dominate). All the ballots are printed (in 2012, online). Granted the limitations of polls as guides to canon, this is how a poll should be run.

An individual canon can be valuable for its idiosyncrasy.

An individual’s list of twenty favorite works may include many well-known books, but it usually includes some that rarely appear on other individuals’ lists. Sometimes these out-of-the-way masterpieces can offer us our most valuable reading experiences.

For example, Frank O’Connor, the great Irish short story writer, chose as the supreme short story masterpiece, the finest short story in the world, “Old Portraits,” by Ivan Turgenev.[80] As a long-time reader of O’Connor, I sought it out, read it, and was grateful that he had pointed me to it. Yet in a survey of about fifty leading short story anthologies I happened to make, this short story was never included.

Borges was another idiosyncratic critic. In his personal canon are—not Hemingway, Joyce and Proust—but G. K. Chesterton, Kipling, Robert Louis Stevenson, Conrad and the early science fiction of H. G. Wells (among others). There is no question of Borges being “right” or “wrong” in his personal canon; but if you enjoy his writing, you might want to investigate the authors that influenced him most. He said that he had tried to read every word ever written by Chesterton, best known for his Father Brown mystery stories and his defenses of Catholicism. While some critics, such as Edmund Wilson, have rejected the mystery genre,[81] Borges edited anthologies of mystery stories, and a series of mystery books (the Seventh Circle, the title taken from Dante). He also edited an anthology of fantasy stories, and a series of fantasy books (the Library of Babel). In doing this, he was swimming against the current of most modern literary, academic taste.

A collective canon can be valuable for its lack of idiosyncrasy, or limited by it.

A collective canon expresses the opinio communis, a shared opinion, of a group of people. If the leaders of a church gather, and vote on which books will be considered part of a canon, that list will be more acceptable to a broad base of membership, in theory, than a list created by just one person.

If an English department is responsible for a survey class of English literature, it may decide which six books will be included as the texts in the survey. One could argue that the student will be better served to be exposed to works chosen collectively, than to the idiosyncrasies of an individual canon.

Or one could argue that if a student’s taste is close to the personal canon of a particular teacher, that student is served well, in a different way, by finding that teacher and exploring the off-the-beaten-path classics he or she selects.

On the other hand, one could argue that sometimes it is valuable to read works entirely outside our idiosyncratic bent.

Borges, in teaching, avoided the straitjacket of imposing his own personal favorites on students. He thought the idea of students reading assigned books they did not like went against all the logic and real experience of reading literature, which is a joyful experience of personal choice. Find the authors who speak to you, he said, and enjoy them thoroughly. Of course, he also enthusiastically praised his favorite writers in his classes.[82] In the same vein, Samuel Johnson said, “A man ought to read just as inclination leads him, for what he reads as a task will do him little good.”[83]

For practical reasons, we sometimes must work within the framework of some “social” canons, but in an absolute sense we, as individuals, are bound by no canons.

The use that is made of a canon is entirely dependent on collective relationship. For example, a teacher might select the five best American novels of the nineteenth century, in his or her judgment, for a class. The students will be influenced by that short list of novels. A book reviewer for a major newspaper or magazine, writing a column on a favorite book, might influence many thousands of readers. He or she may reprint it later in a book, reaching thousands more readers. Friends who read a great deal might be influential in their evaluations and recommendations of books, in their limited circles, and their friends will come to know these favorite books.

In a collective social situation, we may be required to work within the framework of someone’s canon temporarily for practical purposes—we may feel obligated to read the five novels in the American novel class, even if we start one and don’t care for it at all. But of course, we cannot be required to accept the teacher’s “canon” as correct, or applicable to ourselves in an absolute sense. (A student may be a Henry James person, not a Mark Twain person. The teacher may want to include both in a class so the student will be introduced to both.)

If a librarian accepts 100 books in a certain genre, for practical purposes, we will be more exposed to those books than any of the excluded books, if we go to that library frequently. (Of course, we can go to other libraries, or buy books, if necessary.)

Even less are we required to accept a canon—collective or personal—if we are not part of the collective relationship of the canon. A Buddhist does not need to read and live by the New Testament.

However, we may be enriched by all canons—whether produced by an individual, a committee, or a thousand voters in a poll. They will be worthwhile as possible guides in our own reading.

The most useful canons may be personal canons by critics or authors whose taste is similar to our own.

The most useful canon may be a personal canon from a friend or reviewer or teacher if we find a friend or reviewer or teacher who has taste similar to our own, but who has explored more widely in literature than we have. These will be much more useful to us than individual canons by critics whose taste differs fundamentally from our individual tastes. And it will be much more useful to us than polls of voters, as interesting as those sometimes are.

Sometimes a canon by an individual will be presented as a collective consensus.

As was noted earlier, in 1886, Percy Lubbock produced an influential list of the hundred best books.[84] When he reprinted the list in The Pleasures of Life in 1890, he wrote that in producing this list,

I had not presumed to form a list of my own, nor did I profess to give my own favourites. My attempt was to give those most generally recommended by previous writers on the subject. . . . I may observe that I drew up the list, not as that of the hundred best books, but, which is very different, of those which have been most frequently recommended as best worth reading.[85]

Thus, Lubbock writes, he is not expressing his own tastes at all in his list; he has taken an informal poll of sorts—through his education, friendships, and reading—and published the results. But we may ask: who are the voters in his poll? He has selected the “voters” he trusts. They are people who have taste similar to Lubbock’s taste. So we have gone full circle. He has created a combination of his own taste and what he feels most “right-thinking” critics would also choose, adding in a measure of works included for their cultural impact, not for their inherent value.

However, if a list intends to distill collective taste, why not actually create a real poll, a real voting situation? Either create your own individual canon, or organize a collective poll in a responsible way.

In our own generation, Harold Bloom’s influential The Western Canon is refreshingly idiosyncratic, in places. For example, he includes Tolstoy in his short list of twenty-six writers, which is entirely understandable, but he chooses as his entry by Tolstoy the short novel Hadji Murad. And in the long list in the appendix, his choice from Updike is not the Rabbit books, which have been overwhelmingly Updike’s most influential work, but the offbeat fantasy The Witches of Eastwick.

So we may at first conclude that The Western Canon is Bloom’s idiosyncratic canon, which in theory can be a very valuable contribution. But then we find him making statements such as this: “I am not as confident about this list as the first three.”[86] Why not? He can be entirely certain about his own individual taste. Or is he saying canonicity does depend on a present or future opinio communis? If so, how will we arrive at an absolute “common opinion”—when, in actuality, all canons will always conflict with all other canons, including past and future canons? And if he is saying that canon is dependent on present “common opinion,” he should include War and Peace and Anna Karenina in his shortlist of best works. 

Bloom continues on to say that he has included two writers, Robert Lowell and Philip Larkin, whom he personally believes are overrated.

Thus his canon is unfortunately confused; at times representing the real Bloom personal canon; at other times representing the purported absolute, opinio communis Western canon.

But now the reasoning gets circular: Bloom is representing the “common opinion” of whom? Of academics? Certainly not the type of academics whom he excoriates for debasing the canon by including works he disapproves of in the curriculum. So he is representing the subgroup of academics who think and read and judge as he does. But even here, how can he represent fellow academics conscientiously? A simple poll would do that better.

In the Molière section of Bloom’s short list of “canonical” authors, he writes, “By common consent, Molière’s masterpieces are The Misanthrope, Tartuffe, and the very ambivalent Don Juan.”[87] But wouldn’t some critics include four plays instead of three? Perhaps The Miser or The Imaginary Invalid? And some critics might very well disagree about Don Juan. In fact, Molière expert Brander Mathews writes that the play “demands discussion, if not as one of Molière’s masterpieces, at least as a striking product of his genius.”[88] So Bloom’s “by common consent” is not valid; there is no way for him to give convincing support for this, even though he is in theory expressing opinions not his own.

In fact, Bloom appears to be expressing his own opinion, but presenting it as the “common opinion.” If he is expressing his own individual view—the Bloom canon—then why make a seemingly authoritative generalization about what the “common consent” supposedly is?

So Bloom’s The Western Canon, aside from the title, is flawed because he cannot decide if he should reflect an opinio communis—which to do carefully and thoroughly, he could conduct a responsible and intelligent poll—or his own idiosyncratic taste. (Of course, if he conducted a poll, he would have to decide whether it would be a fairly limited poll (the Modern Library route), or a broad-based poll. And would he select only critics who mirror his own views? Would he include a segment of non-academic voters, the man or woman in the street?)

Bloom sometimes seems to suggest that he is making a prophecy of the future opinio communis. But canons, and the opinio communis, change every generation, and from country to country. Which future poll and geographical poll is he hoping to approximate?

How do we select the arbiter or arbiters for a canon?

“Critics” can be accepted as arbiters.

The word “critic” derives from the Greek “kritikós,” meaning one who judges or selects (from Greek, krínō, in the standard ancient Greek dictionary, “separate, put asunder, distinguish,” then “pick out, choose,” then “ judge of, estimate” and “decide in favour of, prefer, choose”)—in other words, the critic evaluates, dividing the best works from lesser works.[89] Quintilian referred to Aristophanes and Aristarchus as “judges of the poets,” poetarum iudices (10.1.54). The product of this kind of evaluation is a list of best works. The critic, of course, does much else—from establishing a text to interpreting it in many possible ways or putting it in historical perspective—but the essential meaning of the word “critic” points to the basic task he or she has of evaluating aesthetic creations, conducting the sifting process that produces limited groups of best works.

Some critics have downplayed or rejected this aspect of criticism. Northrop Frye, notably, seemed to regard the list of “masterpieces of literature,” the “existing monuments of literature” (which are “the materials of literary criticism”) as obvious; and since they were clear to any good critic, he felt that it was not necessary to discuss their comparative standings. The critic must “snip off and throw away” evaluative tendencies. Frye referred to such critical evaluation as “debaucheries of judiciousness.”[90] But in one sense Frye was abdicating the most basic task of a critic, given the word’s root meaning. The list of “masterpieces of literature” is not obvious. If one were to teach a class introducing students to the novel, and there was time enough to cover five novels, which novels would different critics or teachers pick? Even within the canon of a single author, some critics would choose Anna Karenina over War and Peace, and some would pick David Copperfield over Bleak House. If one were to select one American novel for this class, would it be Moby Dick, The Scarlet Letter, Huckleberry Finn, The Great Gatsby, or The Sound and the Fury? Then we have the problem of “obvious masterpieces” that are rejected by some competent critics.[91]

In addition, sometimes authors view critics as natural enemies. For example, Hemingway stated that he had “always regarded critics as the eunuchs of literature.”[92] Mark Twain wrote, “The trade of critic, in literature, music, and the drama, is the most degraded of all trades.” P. G. Wodehouse asked, “Has anybody ever seen a drama critic in the daytime? Of course not. They come out after dark, up to no good.”

But the opinions of critics, academic or journalistic, are valuable for many reasons. Many major authors are also critics and/or academics. Poets such as S. T. Coleridge, Edgar Allen Poe and Matthew Arnold wrote important criticism. In addition, the views of critics are useful simply because they have read many books; the well-read critic is an important resource (though critics’ personal favorites are only their own).

Anthologies, often edited by academic critics, are used in many classrooms, and so they function as important academic canons of short works.

Authors can be accepted as arbiters.

One could argue: Who better to judge art than a person who has created superior art? Novelists, composers, and artists know their craft, and thus would be the best judges of other practitioners of their craft (in theory).

This is a convincing argument. Creators—novelists, composers, poets—are important arbiters of the canons of their particular genres. Sometimes creators outline canons by writing formal essays or books, sometimes informally by reported conversations or interviews, sometimes by pointing to “precursors.” Dante and Chaucer, for example, as they wrote their own poems, included lists of the best former poets.[93]

However, there are important limitations that must be placed on the role of artist as source of canon.

First, a particular artist may have very narrow taste, instead of broad, catholic taste. Novelists may prefer books that are fairly similar to their own. They may reject good or great works that are significantly different from their own masterpieces.

Also, there may be competition between living writers, and so a writer may be the least reliable source on a rival. Personal animosity can cloud judgment, when authors know each other. Tolstoy and Turgenev came close to fistfights and duels. In other cases, a writer may overpraise a close personal friend.

Brahms and Tchaikovsky, two of my favorite composers, are another example. On October 21, 1886, Tchaikovsky confided to his diary, “I have played over the music of that scoundrel Brahms. What a giftless bastard! It angers me that this conceited mediocrity is regarded as a genius.”[94] Tchaikovksy, known for his hyper-emotional romantic music, did have a favorite composer—the comparatively restrained composer who defined the “classical” era, Mozart, whom the Russian composer referred to as a “musical Christ.”

Librarians can be accepted as arbiters.

When a librarian decides which books should be in a library, he or she is in essence creating a canon—choosing a list of best books, accepting some, rejecting others. Librarians have limited money and limited shelf space. Which limited number of books should be chosen? Reviews of books in such magazines as Library Journal may have a great practical impact, as librarians fill their shelves. Librarians as arbiters have a long tradition in western literature, beginning with Aristophanes and Aristarchus. Borges is an example in our own century, though he was also a writer of fiction, a poet and a critic.

The Library of Alexandra. A library is by definition a canon, a selection of books, and librarians are by definition arbiters of canon.
Illustration by  O. Von Corven, courtesy Wikimedia Commons.

Teachers can be accepted as arbiters.

There is a powerful impetus toward creating a canon in simply deciding what books to teach in a course, from the few books chosen for high school English classes, to graduate courses at universities. Many literature programs at universities have reading lists, which the students are expected to cover before taking exams. So academics have an important motivation to create canons; they have to.[95]

Nevertheless, academic canons have a tendency to become powerful, then invariable, then accepted as absolute. A collective academic canon can be intolerant of personal canons, “new” canons, new genres, new works. For example, in the Middle Ages, “classic” works were written in Latin; works in the vernacular languages were not considered to be worthy of consideration as respectable literature.

In American literature, Twain’s Huckleberry Finn was at first viewed by some readers as a boy’s book written in a non-literary style. According to one newspaper report,

The Concord (Mass.) Public Library committee has decided to exclude Mark Twain’s latest book from the library. One member of the committee says that, while he does not wish to call it immoral, he thinks it contains but little humor, and that of a very coarse type. He regards it as the veriest trash. The librarian and the other members of the committee entertain similar views, characterizing it as rough, coarse and inelegant, dealing with a series of experiences not elevating, the whole book being more suited to the slums than to intelligent, respectable people.[96]

This incident also shows librarians acting as arbiters of canon, deciding which books to include on (and exclude from) their shelves.

Friends can be accepted as arbiters.

On a personal level, we can talk to friends who have similar (or different) tastes in art, books, or music. They will be useful creators of personal canons, though their impact will be limited.

Readers can be accepted as arbiters.

A poll can be created from readers of a magazine or a newspaper, or (moving to the film genre) from the users of a major website such as Internet Movie Database.

We may select books for their cultural impact, the size of their readership.

The supreme democratic canonical criterion is a book’s sales. Some critics reject the “best-selling” criterion, even making it an argument against a work’s high worth, but it is actually valid if you are judging a book by historical impact. The best-selling books of the decade or century have had undeniable historical impact. If we accept the importance of cultural influence, then we must agree that a book having a wide readership should be considered when we create a canon.

In addition, we have seen that if an author does not have a large readership, his or her books may simply be lost, and he or she drops out of the canon. For example, we have only fragments of many of the authors included on the Alexandrian canon. During the middle ages, Homer was known as an ideal, but good, full Latin translations of the Iliad and Odyssey were not available. Homer had in effect become decanonized.

So a large popular readership enhances an author’s chances of survival, and his place on any practical canon.

We can certainly argue, however, that a book may have great cultural impact, but lack lasting aesthetic value.

Or course, duration of impact is an important factor, as we may consider a book’s impact over time—if it was a best seller briefly, after which it was mostly forgotten, its impact is obviously less than the impact of a book that was a bestseller on its appearance, and has continued to be read (and bought) for generations afterward.

Or a great book may never be a bestseller, but may continue to have a dedicated readership over time.

Harriet Beecher Stowe’s Uncle Tom’s Cabin had enormous historical impact, yet few readers or critics today would rank it next to Huckleberry Finn or Moby Dick. Nevertheless, it places high in at least one great writer’s canon: that of Tolstoy.[97] In fact, Tolstoy argued that the greatest works of art must reach the common man, not just upper-class elites.[98]

However, many factors argue against “the best-selling” criterion, taken alone.

chronology must be taken into account.

Many books are best-sellers for a short time, then are forgotten. They are not best-sellers in the broad scope of history. Other books are not best-sellers when first published, but gain a limited solid readership, and then the readership grows through decades or even centuries.

John Ruskin divided books into two classes, by a chronological criterion: “All books are divisible into two classes: the books of the hour, and the books of all time.”[99]

Some books become best-sellers because editors select them, then promote them.

There is a large element of chance in which books get published and promoted, and often the best books don’t receive the best publication and promotion. There are striking examples of great books rejected by many publishers before finding an editor willing to take a chance on them. What if the author had not sent it to that one last publisher?

The first volume of Proust’s In Search of Lost Time is a case in point.  Graham Greene regarded Proust as the “greatest novelist of the 20th century”, and W. Somerset Maugham called In Search of Lost Time the “greatest fiction to date.” Yet the first volume of the work, Swann’s Way, was rejected by a number of publishers. The distinguished novelist André Gide served as a reader for one of the publishing houses, and recommended against publication. Proust himself finally published it privately. The book subsequently had a major impact, but what if the author hadn’t had enough money to finance the book publication himself?

Thus some editors have gone out on a limb to publish idiosyncratic, high-quality work not obviously suitable to a mass market; while other editors focus on derivative, shallow work that looks financially reliable. Some best-sellers do not represent a democratic choice; they represent the choice of one editor, who, if he is powerful, can get the book a readership.

Many great books have had an enormous contemporary impact, were and are best-sellers.

W. Somerset Maugham pointed out that, of the ten books he selected as the ten greatest novels, seven had been best-sellers at their time of publication:

I have dwelt upon this matter in order to point out how stupid it is of a certain class of critics, and unhappily also of a portion of the public that regards itself as belonging to the intelligentsia, to condemn a book because it is a bestseller. It is inept to suppose that a book that vast numbers of people want to read, and so buy, is necessarily worse than a book that very few people want to read, and so don’t buy.[100]

Of course, he went on to agree that not all best-sellers are good books! “It [A bestseller] may be a very bad one. A book may become a bestseller because it deals with a subject that at the time happens to interest the public, and so notwithstanding the great faults it may have is widely read.” And there are many other reasons that bad books may become bestsellers.

We may select books for their importance and influence within a tradition.

If a book may be influential, and much copied within a tradition, it is, by definition, canonical, in the sense of acting as a model. For example, a book may be first in a genre, the first novel, the first short story, and thus influence all its epigones in the genre. It obviously has importance for that reason.

But it may be a lesser book than its best epigones, from the standpoint of lasting value, even though it is more influential than better books. Thus it is a model for certain technical aspects of genre, but not a model in the sense of being an example of the highest quality, from an absolute standpoint.

Intertextual impact may cause us to revise our judgments of books, or cause us to read unheralded books.

As we have seen, at one stage in his life, Tolstoy did not feel he could make a list of the “best books,” for various reasons, but he was willing to create a list of the books that had influenced him the most.

Canons—different lists of what is the best—are the product of the readers’ standards for selection.

We can be amused (or shocked!) at differences in canon, and in tastes. However, when Shaw does not include Shakespeare in his canon, this is the result of his view of what elements in art create the best art. For example, “word music,” which he agreed that Shakespeare excelled at, was less important to him than what the work said philosophically and politically, and he felt that Shakespeare’s strengths did not include ideological depth.

Therefore, in creating canons, and understanding variations in canons, we must decide which standards we (and others) use in selecting the “best.”

Bloom and others have spoken of “canonical virtues” in a literary work, “canonical qualities.” Actually, if a canon is a list that seeks to gather the best in an artistic genre, then we simply need to decide what works of art are best, and why. A work is not really canonical in a general way; it is simply great, exemplary. No matter how great, it will never gain entrance into “the canon” of anything, unless we’re talking about a specific canon. On the other hand, it may enter many canons.

Ideas on what is great in literature, music, film and art are extremely varied. It is useful and interesting to understand what aesthetic factors led to particular canons.

The works in the canon of Quintilian, and many other Roman canons, were selected for their value as texts for developing the student’s talents for public speaking. It was a purely rhetorical criterion—literature for lawyers or politicians.

Many medieval canons were based on Christian ideology and morality.

Most modern critics and academics would choose masterpieces based on aesthetic quality. Harold Bloom selected his canon on the basis of pure aesthetic experience. Similar, but subtly different, is Borges’s statement that he read his favorite authors solely for pleasure.[101]

But critics and authors will disagree on which aesthetic standards are most important or pleasurable. When introducing a canon, therefore, it is helpful to define the criteria we are using. Then readers may disagree with our criteria, but will agree that our choices have some logic, given the criteria. Or reader may agree with our criteria, and our canon may be especially useful to them.

There are many possible pitfalls in the business of judging what is authentically the best in art. On the one hand, we must avoid the shallow and evanescent; but on the other hand, sometimes we rank as “best” that which we do not authentically enjoy. Or one might argue that the greatest works of art are not enjoyable in a shallow way, but are challenging, rather than easily enjoyed. Or, on the other hand, one might posit that the greatest works of art, like War and Peace, Don Quixote, David Copperfield or 100 Years of Solitude, are vastly enjoyable on one level, while deeply moving and challenging on another. By this yardstick, works that are merely challenging, but not vastly enjoyable, will never authentically join the company of highest works of art.

In fiction (and to some extent, in poetry, which often tells a story, as in epic poetry or Coleridge’s “The Rime of the Ancient Mariner,” or even Frost’s “The Road Less Traveled”), stylistic skill; depth of characterization; skill in story-telling; social significance; moral vision, philosophical importance and depth are all part of the aesthetic experience.

Obviously, a reader’s complex responses to all of these criteria are personal and subjective. One reader or critic may choose the neologisms of Joyce’s Finnegan’s Wake as the height of stylistic achievement, and as thoroughly enjoyable. Another reader or critic may choose the surface clarity and emotional depths of Sappho or Chekhov as the highest model of stylistic skill. Another reader or critic may feel that character as defined by a strong story is central to the best fiction, and so may leave Finnegan’s Wake off his or her list entirely.

In fiction, we could argue that all these—stylistic superiority; depth of characterization; skill in story-telling; social significance; moral vision, philosophical importance and depth—are present in all great fiction, and we cannot separate them from each other. The best novels and short stories orchestrate them in the most skillful ways. As Robert Louis Stevenson wrote, in defense of fictional “romance,” “In the highest achievements of the art of words, the dramatic and the pictorial, the moral and romantic interest, rise and fall together by a common and organic law. Situation is animated with passion, passion clothed upon with situation. Neither exists for itself, but each inheres indissolubly with the other. This is high art; and not only the highest art possible in words, but the highest art of all, since it combines the greatest mass and diversity of the elements of truth and pleasure.”[102]

However, some artistic creations of great writes are certainly flawed, and the flaws must be balanced against great strengths. Some writers are especially good at some aspects of the aesthetic mix; and some in fact feel that certain aspects are less important. In modern short fiction, eventful story has often been dropped in favor of style evoking mood, characterization, setting. Some have argued that this has been a false step in our modern tradition; others have regarded it as a step forward. Faulkner criticizes Balzac for writing badly, but still admires the great characters in his novels.[103]

If we, as readers, feel that one element of fiction is important, if an author de-emphasizes it, we will be less inclined to read that author and put that author’s works on our personal canon; and we will obviously receive less value from a canon reflecting that de-emphasis.

Style as a canonical criterion.

Some critics feel that the primary criterion for great literature is style—beautiful or elegant language, how something is written—as opposed to what is written.[104] But in fiction, critics or authors who see style not as the heart of a work, but as a supportive element in aesthetic pleasure, will evaluate the best in literature differently.

Isaac Bashevis Singer excluded Joyce from his personal canon because “style is dominant and the story only serves as a container of the style, a frame.”[105] Style that is dominant is exactly the characteristic that would cause other authors or critics to include Joyce in their canons.

One could argue that how one says something—no matter how elegantly you say it—is less important than what one says. Or you could argue that a story or poem badly told or written is by definition bad.

Story as a canonical criterion.

Authors such as Isaac Bashevis Singer, Frank O’Connor, and Flannery O’Connor feel that an involving, compelling story, with a structure of significant events, is central to the best in fiction.[106] Telling stories extends back through time to oral folktales, back to the beginnings of language, cavemen telling stories around the campfire at night. Stories are key components of the earliest historical books, the earliest religious texts, early poems such as The Odyssey and Iliad.

Singer proposed children as the best literary critics, and wrote:

The young reader demands a real story, with a beginning, a middle, and an end, the way stories have been told for thousands of years. In our epoch, when storytelling has become a forgotten art and has been replaced by amateurish sociology and hackneyed psychology, the child is still the independent reader who relies on nothing but his own taste. Names and authorities mean nothing to him. Long after literature for adults has gone to pieces, books for children will constitute the last vestige of storytelling, logic, faith in the family, in God, and in real humanism.[107]


Novelist Isaac Bashevis Singer proposed children as the only reliable literary critics because they demand “a real story, with a beginning, a middle, and an end,” and because children as readers lack hypocrisy, relying on their own taste, not on the authority of critics.

Characterization as a canonical criterion.

One might argue that creation of great characters is central to great fiction. William Faulkner, for example, judged books by the strength of their characters:

The character, the book is the thing and not who wrote it. . . the people I know and love are Don Quixote, and Sarah Gamp,[108] and some of Conrad’s people, a lot of Dickens’s people, Balzac’s people, but not Balzac especially, because I think some of Balzac’s writing is bad writing. Some of Conrad’s writing is bad writing, but some of Conrad’s people that he created are marvelous and endured.[109]

Of course, many of these criteria are intertwined. Proponents of the story criterion could argue that a compelling story is always based on involving characters who are revealed best by the events of the story.[110]

Moral vision as a canonical criterion.

Harold Bloom has argued that “moral insight” is not a criterion for evaluating the greatest literary art. Only aesthetic achievement is.[111] 

However, throughout history, canons have often been created based on works’ moral or religious value. For an example of a religious criterion for canonical inclusion, Virgil was allowed on many medieval canons because some Christians felt that the Fourth Eclogue was a prophecy of the coming of Christ. The satirist Juvenal survived on many lists of texts for Christian students because learned and moralizing monks felt his satire, attacking wickedness of many sorts, exhibited a high morality.[112]

We may regard this perspective as medieval, outdated. However, for two of our great modern novelists, Tolstoy and Dostoyevsky, ethics were a central feature of the greatest art. Tolstoy’s first condition for a “true work of art” is “a correct, that is, a moral relation of the author to his subject.”[113] He also felt that the greatest art must be ethical in a religious sense: “The best works of art of our times transmit religious feelings urging towards the union and the brotherhood of man.”[114]

Novelist Arnold Bennett wrote, “Scratch a serious novelist and you will find a preacher with a moral message. I doubt there is any exception to this rule.”[115] (Not surprisingly, four novels by Dostoyevsky and three by Tolstoy begin Bennett’s list of twelve “finest” novels in world history. Four more slots are taken by Turgenev, followed by Gogol’s Dead Souls.)[116]

If you argue that the characterization and story of Conrad’s Lord Jim does not deal with a central moral issue, and that this issue is not a central concern of the novel, you are clearly not reading the book very perceptively.

Bloom argues that great art subverts all values and morality, which is one of his less convincing arguments. What he seems to be saying is that great art sometimes offers a higher morality that subverts widely-accepted, lesser social morality.[117]

It is certainly true that bad art can be moralistic. Flannery O’Connor wrote,

I find that most people know what a story is until they sit down to write one. Then they find themselves writing a sketch with an essay woven through it, or an essay with a sketch woven through it, or an editorial with a character in it, or a case history with a moral, or some other mongrel thing.[118]

Yet morality, in the best fiction, is woven into the story and the drama. O’Connor also wrote, “When you write fiction you are speaking with character and action, not about character and action. The writer’s moral sense must coincide with his dramatic sense.”[119]

Characterization and ethics are closely connected. The hero of epic poetry and the novel is heroic because of his or her actions, guided by his or her ethical outlook. As Turgenev said of Don Quixote:

The massiveness of Don Quixote’s moral structure (it must not be forgotten that this distracted knight errant is the most moral creature on earth) imparts a particular gravity and stateliness to whatever he may say or do. In a word, his ethical character gives an uprightness to his whole figure despite the preposterous situations and the humiliations into which he is incessantly tumbling.[120]

Borges included Conrad and Shaw in his personal canon because of their focus on courage and heroism, moral concerns. He said, “I am essentially ethical. I always think of things in terms of right and wrong.”[121] He stated that his interest in ethics was partially an influence from Robert Louis Stevenson, and referred to himself as an amateur Protestant.[122]

Italo Calvino also described Stevenson as great not just because of “the clean, light clarity of his style, but also because of the moral nucleus of all his narratives.”[123]

Stevenson himself, in an essay on books that had influenced him, wrote, “the course of our education is answered best by those poems and romances where we breathe a magnanimous atmosphere of thought and meet generous and pious characters. Shakespeare has served me best.” So characterization is a central aspect of his criteria for the best in literature, but the characters also must be “generous and pious,” exhibiting moral, even religious qualities. For Stevenson, these great characters in fiction or drama are profoundly lovable, and even improve the reader: “Few living friends have had upon me an influence so strong for good as Hamlet or Rosalind.”[124]

Maugham seems to come down on both sides of this issue. In the introduction to The Art of Fiction, he wrote, “I think it an abuse to use the novel as a pulpit or a platform . . . What it all comes down to is the question whether the novel is a form of art or not. Is its aim to instruct or to please? If its aim is to instruct, then it is not a form of art. For the aim of art is to please.”[125]


The fact remains that to describe a novelist as a mere storyteller is to dismiss him with contumely. I venture to suggest that there is no such creature. By the incidents he chooses to relate, the characters he selects and his attitude towards them, the author offers you a criticism of life. It may not be a very original one, or very profound, but it is there; and consequently, though he may not know it, he is in his own modest way a moralist.[126]

Matthew Arnold felt that morality was a key underpinning of the best poetry. He wrote,


It is important, therefore, to hold fast to this: that poetry is at bottom a criticism of life; that the greatness of a poet lies in his powerful and beautiful application of ideas to life—to the question, How to live. Morals are often treated in a narrow and false fashion, they are bound up with systems of thought and belief which have had their day, they are fallen into the hands of pedants and professional dealers, they grow tiresome to some of us. We find attraction, at times, even in a poetry of revolt against them. . . . A poetry of revolt against moral ideas is a poetry of revolt against life; a poetry of indifference towards moral ideas is a poetry of indifference towards life.[127]

Regarding ethics as a quality of great art is out of fashion now, as Bloom’s summary dismissal shows, but some modern authors and critics have continued to emphasize the moral tradition in literature.[128]

Philosophical depth as a canonical criterion.

Henry James excluded Dickens from his personal canon because he felt that Dickens lacked “philosophy,” and that this was always a component of the best fiction:

Mr. Dickens is a great observer and a great humorist, but he is nothing of a philosopher. Some people may hereupon say, so much the better; we say, so much the worse. For a novelist very soon has need of a little philosophy. In treating of Micawber, and Boffin, and Pickwick, et hoc genus omne, he can, indeed, dispense with it, for this—we say it with all deference—is not serious writing. But when he comes to tell the story of a passion, a story like that of Headstone and Wrayburn, he becomes a moralist as well as an artist. He must know man as well as men, and to know man is to be a philosopher.[129]

Other authors disagree. For example, Faulkner said, “The writer of fiction is not too interested in ideas, . . . any more than he is in facts. His interest lies in people.”[130] Therefore, Dickens was included in Faulkner’s personal canon.

Isaac Bashevis Singer rejected authors who substituted philosophical treatises for real, involving, storytelling. “I feel that modern writers began to write essays about literature instead of telling stories. . . . they write disguised essays which they call stories. And I think The Magic Mountain [by Thomas Mann] is such a disguised essay.”[131]

seriousness as a canonical criterion.

In his essay “The Study of Poetry” (1880), Matthew Arnold took “high seriousness” as a key quality that distinguished the greatest poets, such as Milton, Shakespeare, Dante, and Homer, from lesser poets, such as Chaucer, Dryden, Pope and Burns (who are still good or great poets, simply not in the first rank). Chaucer, for all his historical importance, “lacks the high seriousness of the great classics.” Arnold selected eleven short passages from the greatest poets as “touchstones” by which we can “test” other poets.

Arnold applies the “high seriousness” test to poetry; one wonders if this same criterion works with fiction, or music. Are comic novels (or plays) disbarred from Arnold’s select company of greatest works of art? Would Tom Jones, Alice in Wonderland, Pickwick Papers, The Importance of Being Earnest and Midsummer Night’s Dream be considered lower art? In music, an integral part of Beethoven’s symphonies was the scherzo, the “joke,” and the final movements of most symphonies and concertos are rollicking, quick rondos, much lighter in tone than the first two movements. Is “low entertainment” as important to the totality of great art as is “high seriousness”? The Canadian novelist, Robert Davies, who always had leanings toward comedy, put Tolstoy on a lower rung of greatness because he had no humor. “Tolstoy is full of compassion, but I mistrust it because he has no humour.” And he describes Dostoevsky as “drowned in his own seriousness.”[132]

Politics as a canonical criterion.

In Bloom’s perspective, “aesthetics” has everything to do with canon formation, politics has nothing to do with it. It is easy to point to much bad art, music, fiction, poetry that is politically informed and aesthetically bankrupt—such as bad Marxist art, which can be simplistic and numbingly didactic propaganda.

However, we have seen that many authors and critics regard ethics as central to great art; and that art often deals with the ethics of the individual in relation to society, as early as Antigone and Plato’s Apology. Politics is an important aspect of human existence. Therefore, one could argue that political insight, however subtly expressed, is often an aspect of great art.

This is exactly the argument made by Gabriel García Márquez, whose socialist, anti-colonialist tendencies are well known. In a dialogue with Peruvian novelist Mario Vargas Llosa, Márquez said,

I know of no good literature that serves to praise established values. I have always found in good literature the tendency to destroy the established, the accepted, and to contribute to the creation of new societies, in the end, to better the lives of men. . . . When I sit down to write a book it is because I am interested in telling a good story. One that will appeal. ...  [However, if a writer has a firm, sincere ideological position], this position will nurture the tale and it is from this moment on that the story can have the subversive forces of which I speak. I do not think it is deliberate, but it is inevitable.[133] 

He also said, in an interview, “The arts are always at the service of politics, of some ideology, of the vision the writer or the artist has of the world. But the arts should never be at the service of a government.”[134] And also,

The problems of our societies are mainly political. And the commitment of a writer is with the reality of all of society, not just with a small part of it. If not, he is as bad as the politicians who disregard a large part of our reality. That is why authors, painters, writers in Latin America get politically involved. I am surprised by the little resonance authors have in the United States and in Europe. Politics is made there only by the politicians.[135]

Of course, all bad art produced as simplistic propaganda shows that political viewpoint alone does not make fiction great. Bloom’s emphasis on the necessity for aesthetic accomplishment in great art cannot be simply rejected. Nevertheless, to Márquez himself, what a novel said about politics was obviously a central aspect of its artistic impact. To write a great novel on a political issue requires all of the characteristics of great art—great storytelling (for Singer), great characterization (for Faulkner), philosophy (for Henry James), stylistic sophistication (for Joyce and Woolf). Nevertheless, the politics is still there and is still important. What the author is saying is still important, not merely how he is saying it.

Some influential writers, such as George Orwell or Flannery O’Connor, have stated that they were using aesthetic skill as a means to an end: to convey political or religious perspectives. However, we might add that they convey these perspectives in profound ways.[136] Orwell, in his essay “Why I Write,” listed, among four categories for his motivations, “Political purpose.” But another major motivation was “Aesthetic enthusiasm.” Great art allows us to experience the human element in political or religious themes. In addition, there are elements of ambiguity (or as O’Connor would say, mystery) in great art that are not found in oversimplified political or religious propaganda.

Paradoxically, in literature, no political position can be accepted as morally absolute. Borges worshipped both Kipling and Chesterton as great writers, but their politics were diametrically opposed: Kipling supported England as colonial power, while Chesterton was a little Englander. An author can be limited in some political positions, and still be a great artist.

If we allow ethics as an important criterion for great fiction, a protagonist working for or against his or her social and political milieu may be a central aspect of a story. For example, one of the most moving moments in Huckleberry Finn is when Huck decides not to turn Jim the slave over to the law, to be punished and literally sold down the river, and believes himself consigned to hell for that choice. That crucial moment is impossible unless Huck is seen against the cultural, political and religious backdrop of southern slave-holding states before the Civil War.

One could argue that Twain’s moral vision is not what makes Huckleberry Finn great. Instead it is his stylistic and narrative innovations, the poetry and characterization of the book. Yet such a view overlooks the organic, multifaceted nature of art. Without the moral vision, one could argue, the poetry, narrative and characterization are simply dry technical achievements. Calvino writes that Huck Finn is “wonderful for its language, the richness of life and adventure it contains, its sense of nature, its involvement with the social problems of its time and place.”[137] I would add humor. All of these contribute to Huck Finn’s achievement as a masterpiece.



Huck and Jim “on the raft.” Illustration from Mark Twain’s Huckleberry Finn. Huck’s moral struggle with slavery is one of the striking examples of fiction driven by political perspective in American literature.


Márquez argues that politics is always part of the totality of human experience, and this part of human experience must be reflected in great art. He often cites Oedipus the King and Antigone as two of his favorite literary works; in both of these plays, political themes are central.

Re-readability as a canonical criterion.

One feature of great books is their ability to draw us back to re-read them, both for delight in their characters and stories, and for the complexity of their worlds and world-views. Calvino, in his essay, “Why Read the Classics,” wrote, “A classic is a book which with each rereading offers as much of a sense of discovery as the first reading.”[138]

Fun as a canonical criterion.

Frank O’Connor, in an interview, confessed:

I really got into this row, big, at the novel conference at Harvard, when I had a couple of people talking about the various types of novel—analyzing them—and then we had a novelist get up and speak about the responsibilities of the novelist. I was with Anthony West on the stage, and I was gradually getting into hysterics. It’s never happened to me before in public; I was giggling, I couldn’t stop myself. And, “All right,” I said at the end of it, “if there are any of my students here I’d like them to remember that writing is fun.” That’s the reason you do it, because you enjoy it, and you read it because you enjoy it. You don’t read it because of the serious moral responsibility to read, and you don’t write it because it’s a serious moral responsibility. You do it for exactly the same reason that you paint pictures or play with the kids.[139]

This, of course, seems to contradict Tolstoy’s basic view of literature. Perhaps O’Connor or Tolstoy would agree that fun, delight in aesthetic experience, can be part of a great novel that also has a moral tendency. Tolstoy included Dickens’s comic novel Pickwick Papers and the often-comic David Copperfield among the great works of literature, though in his later life he put them at a slightly lower level than Dickens’s two Christmas books, A Christmas Carol and The Chimes, and A Tale of Two Cities, with their profoundly Christian ethics.[140]

Isaac Bashevis Singer said, “literature is basically entertaining.”[141]

hypocrisy in the process of canon formation

Mark Twain is often quoted as saying, “A classic is something that everybody wants to have read and nobody wants to read.”[142] While he actually quoted someone else saying this, the principle is correct. Often we put books that we don’t authentically enjoy on canons, books that we never want to reread, and that we read only out of duty in the first place. However, we think that such books should be on the shelves of libraries, that students should be required to read them. They may be books that were very influential, so have great historical importance. They may be books that are seen as profound, but are not especially readable.

When spy novelist John Le Carré was asked to provide a list of recommended books for the webzine Salon, he began by writing,

I always loathe answering this question because it’s such a temptation to show off. How impressed you would be if, for instance, I told you I liked nothing better than to curl up with Bishop Ulfilas’s translation of the Bible into Gothic. And how often, I suspect, do all of us—writers and readers alike—urge upon our luckless friends and children books we have never managed to wade through ourselves. First things first then. No library, however humble, is complete without its well-thumbed copy of Right Ho, Jeeves by P.G. Wodehouse, which contains the immortal scene of Gussie Fink-Nottle, drunk to the gills, presenting the prizes to the delighted scholars of Market Snodsbury Grammar School, built around 1416. . . . That’s just to remind yourself what fun reading is.[143]

Are canons by nature overbalanced to the past? Is a canon of “dead white males” (emphasizing here the deadness, rather than white maleness) valid?

In a sense, yes. It is impossible to predict which contemporary works will join the “collective” or “social” canons of the future. The dust has not yet settled. (And the dust will never settle completely.)

While some “collective” canons have been attacked as overbalanced toward “dead white males,” the “dead” component of “collective” canons is actually quite reasonable. Many authors of canons have intentionally not included living writers—Aristophanes of Byzantium and Percy Lubbock are examples. Future “collective” canons cannot be predicted.

There is always a tension between an accepted canon and new works. Those who have created a canon, especially when it has been formalized in printed or organizational form, may have a vested interest in not rethinking and updating the canon. We have seen that Jonathan Swift envisioned a “Battle of the Books” between ancient and modern books.

New works are always being considered for entry into a precisely limited canon, which necessarily drives previously accepted works out. The brutal survival of the aesthetically fittest remains in play, which might mean that a new candidate for the canon will cause a former component of the canon to be rejected, in certain practical situations. If in a certain class we can read only five works in the term, bringing in a new novel will force us to expel one of the previous five. So those who are emotionally attached to an old canon may become intensely opposed to new books that some may feel rise to even higher standards. Paradoxically, the better the new book, the more intensely the proponents of the “accepted canon” might be opposed to it.

Two examples from film and music show how canons can skew remarkably to older works. In the Sight & Sound film polls, a few older movies have dominated for decades. New or recent movies cannot compete with films such as Citizen Kane (1941), Rules of the Game (1939), 8-1/2 (1963), Tokyo Story (1953), Sunrise (1927) and Battleship Potemkin (1925). In the 2012 top ten, the most recent movie is Kubrick’s 2001: A Space Odyssey (1968). This leaves us with a film canon in which nearly half a century has been ignored. And in the top twenty, the most recent film is Coppola’s Apocalypse Now (1979). This canon has understandably been described as stodgy. Within the dynamics of how Sight & Sound conducts its polls (which they do admirably, as I noted earlier), new movies have not been able to enter the inner circle of favored films.[144] Often individual critics make a point of including more recent movies in their lists, but there has not been a consensus on which recent films (that is, films after 1968!) are the best.

The distinguished composer and music critic, Virgil Thomson, vociferously attacked a moribund classical music canon that had narrowed to “Fifty Pieces,” older works from the European tradition, which were played over and over (which is reminiscent of the top 40 format of popular music radio stations), while worthy new music by American composers was ignored.[145] He wrote from the perspective of a composer seeing the danger in a canon that was too small, too limited geographically and chronologically, and was for practical purposes closed to new works. Someone like Bloom, however, writes from the viewpoint of a critic defending his personal canon from new, unworthy rivals.

How can we judge which contemporary works will enter future canons?

Some masterworks are recognized as such upon publication.

On the other hand, one generation’s masterpiece may be the next generation’s most-loathed—or most-forgotten—book. One generation’s unknown artist may become the next generation’s most-prized artist.

Starting with the Aristophanes/Aristarchus canon, and continuing, for example, in the Lubbock canon, living writers were often left out of lists of the best, because it was felt that their works had not proven themselves over time—an example of the flawed idea that such a factor is a necessary part of an absolute canon acceptable to all. If we reject the idea of canon as perfect Platonic form, living writers can and should figure in contemporary canons. For example, Dante put himself in his personal canon while he was living and writing. Traveling with his great model, Virgil, he is led to a place where he sees the greatest poets of antiquity, Homer, Horace, Ovid and Lucan. These poets confer with Virgil, “And far more of honor yet they did me, for they made me of their band, so that I was the sixth amid so much wisdom.”[146]

Virgil and Dante meeting Homer, Horace, Ovid and Lucan in the First Circle of Hell (Limbo).  William Blake, Dante Hell IV. Courtesy Wikimedia Commons.


Authors asked to participate in polls generally leave themselves out of their own ballots, but here Dante “votes for himself,” just as politicians vote for themselves in elections. Horace was another poet who “placed” himself in his own canon, and hoped he would appear in future canons (Ode 1.1).

Canons can provide guidance on living writers, though obviously, we should agree that canons will change over time, and every contemporary canon will be different. Experiencing great writers as part of their contemporary community is one of the great pleasures of literature and art.

However, often canons work to minimize and discourage new works, which are seen as a threat to the sanctity of past, settled canons. In addition, academia leans toward the “common opinion” rather than the idiosyncratic viewpoint. If we are trying to teach with independent thought and judgment, not merely rubber-stamping the judgments of others, new works (and genres) should be entering the curriculum periodically. For example, Neil Gaiman’s acclaimed and influential Sandman series appeared first as comic books. Academics and literary critics did not typically pay attention to this genre.

Do prizes and awards help create canons?

Prizes define canon in much popular culture. For example, the Oscars, presented yearly by the Academy of Motion Picture Arts and Sciences, are far and away the most influential and well known awards for film.

In literature, we have the Booker Prize, the Pulitzer Prize and National Book Award; the Prix Goncourt, the Nobel Prize for Literature. Influential literary prizes extend as far back as the Golden Age of Greek drama, as Aeschylus, Sophocles and Euripides vied for highly-sought-after dramatic prizes at the Athenian religious festivals of the Dionysia and Lenaea.[147]

Such contemporary prizes and awards can show the early impact of a new work, but they are far from infallible. Despite the fact that prizes are influential and interesting, they are as limited in value as any poll. They are contemporary reactions and subject to all the limitations and drawbacks of contemporary reactions to art. Some great artists were not recognized during their own lifetime—for example, the great Impressionists were rejected or ignored by the French Academy of Art for many years. Van Gogh only sold one painting during his lifetime. On the other hand, some artists who have later been viewed as second rate have received major contemporary awards.

Awards are also geographically limited; for example, the National Book Award considers only American books, and the Booker Prize honors only books from the United Kingdom, Ireland and Zimbabwe. The chief Oscar award, Best Movie, is really only the Best English-language Movie. Best foreign movie is a separate, and a much less influential, award.

Perhaps the most famous example of a contemporary under-evaluation of a work of art is Sophocles’s Oedipus the King, undoubtedly one of the most influential works in Western civilization; its impact has even gone further than aesthetic circles as it served as a basis for Freudian psychology. Yet it could not even win the First Prize at the Greater Dionysia in Athens at its premiere in 427 BCE.[148]

As we have seen, Citizen Kane is usually selected as the greatest movie ever by critics, filmmakers and academics, but it did not win the award for best movie of the year from the Academy of Motion Pictures in 1941. The Academy Awards, at times, have coincided with “critical” taste; at other times, they have been famous for giving the best picture award to slick, bloated crowd-pleasers.

Famous Losers

Sophocles (Oedipus Tyrannus). Roman copy of Greek original. Courtesy Wikimedia Commons.

“Lev Nikolayevich Tolstoy shoeless,” 1901, by Repin. Courtesy Wikimedia Commons.

Orson Welles (Citizen Kane)
Photographed by Carl Van Vechten, March 1, 1937. Courtesy Wikimedia Commons.


In addition, it is customary now for movie studios to invest a great deal of money into campaigns lobbying for their films. A film may win an Oscar in part because a movie studio has mounted an expensive lobbying campaign for it. Another film may lose, in part, because its studio is not as deep-pocketed, or because it doesn’t play the game as skillfully.

Mark Twain, Leo Tolstoy, Henrik Ibsen, Anton Chekhov, Thomas Hardy and Jorge Luis Borges did not receive the Nobel Prize for literature, while dozens of arguably lesser writers did. Tolstoy, often acclaimed as the greatest novelist who ever lived, lost to Christian Matthias Theodor Mommsen, German classical scholar and historian, in 1902. The previous year, the fairly obscure poet and essayist Sully Prudhomme had won the first Nobel prize for literature.[149] As a historian of the Nobel Prize has written, the literary Nobel prize has been given to “a long list of mediocrities,” as well as to the occasional worthy recipient, such as Yeats, Faulkner, Neruda, Márquez, Mann or Shaw.[150]

According to one source, Menander, the dominant author of Greek “New Comedy,” used to continually lose dramatic contests to his lesser rival, Philemon, because of “intrigue, influence and faction.”[151] In addition, a wealthy producer (the choregus) greatly increased a dramatist’s chances of success, as the producer could hire the best actors and singers/dancers, and could buy the best costumes and stage props. As Haigh and Pickard-Cambridge write, “Even in modem times an inferior play, if well mounted and acted, is more impressive than a good play badly performed.”[152]

Sometimes judges were swayed by the reactions of the audience, and Plato laments that this could be the case.[153]

Of course, there have been times when works and authors that have come to be accepted as great have received contemporary awards. Aeschylus and Sophocles received many first place prizes. But such awards are certainly not infallible criteria for inclusion in future canons.

What of the “white male” component of “the Western canon”?

Henry A. Giroux, a professor of education at Miami University of Ohio, writes:

In the most general sense, Bloom and Hirsch represent the latest cultural offensive by the new elitists to rewrite the past and construct the present from the perspective of the privileged and the powerful. They disdain the democratic implications of pluralism and argue for a form of cultural uniformity in which difference is consigned to the margins of history or to the museum of the disadvantaged.[154]

As we have seen, there is a tendency for canons to become closed, and to lean to the past. Often new rivals to important canons are seen as the rabble of the common man (read minorities, women, blacks, Asians) striving to gain entrance by force into an inner circle of aristocratic art. Often new genres, popular with the masses, are not even considered as worth mentioning beside older, accepted canons. The great Chinese novels, written in the vernacular, not classical Chinese, were not at first viewed as worthy art; only later did these novels receive their own canon.[155] In our day, popular music and “classical” music are seen as entirely different genres. “Classical music” is canonical, and popular music will not be considered for entrance in the same canon.

However, as philosopher John Searle reasonably writes, discussing the characteristics of a good basic education:

First, the student should have enough knowledge of his or her cultural tradition to know how it got to be the way it is. This involves both political and social history, on the one hand, as well as the mastery of some of the great philosophical and literary texts of the culture on the other. It involves reading not only texts that are of great value, like those of Plato, but many less valuable that have been influential, such as the works of Marx. For the United States, the dominant tradition is, and for the foreseeable future, will remain the European tradition. The United States is, after all, a product of the European Enlightenment. However, you do not understand your own tradition if you do not see it in relation to others. Works from other cultural traditions need to be studied as well.

If these two streams, both the political-social and the philosophical-literary, are well organized and well taught, the claims of the various minorities should have their place. Intelligently taught social and political histories of Europe and the United States, for example, should recognize the history of all of the major components of European and American society, including those that have been treated unjustly. It is important, however, to get rid of the ridiculous notion that there is something embarrassing or lamentable about the fact that most of the prominent political and intellectual leaders of our culture over the past two thousand years or so have been white males. This is just a historical fact whose causes should be explored and understood. To deny it or attempt to suppress the works of such thinkers is not simply racism, it is unintelligent.[156]

Most importantly, we have seen that references to “the canon” of any broad genre are invalid. So there is no single, absolute “Western canon” or “literary canon.” While Bloom is thus fundamentally mistaken in this concept, critics of “the Western canon,” to the extent that they accept that there is one single absolute Western canon, are also mistaken. Instead there are many conflicting canons. The best way of opposing canons perceived as flawed is by participating in the canon dialogue energetically, in all possible ways. In the realm of personal canons, books, essays, anthologies, syllabi, teaching, talking; and in the realm of collective, social canons, by voting for curricula, departmental reading lists, and in polls.

In conclusion

If we are part of a group using a canon of some sort—a church, an English class, a patron of a library—we are gifted (or stuck) with a gathering of works that has been selected by the group’s leaders. Granted, they have made the selection seriously, often after a substantial winnowing process, and the works selected may very well speak to us as masterpieces.

Nevertheless, the most useful canons for us, as individuals, always searching for another good or great book to read, may be personal canons of persons whose taste aligns with our own. Polls, such as the 2002 Norwegian Book Club poll, can offer us an interesting and useful opinio communis; but it will not reflect our individual tastes. André Maurois has written,

We must choose our literary nourishment well. Each mind requires its own particular food. Let us learn which authors are our authors. They may be very different from those of our friends. In literature as in love, we are astonished at what is chosen by others. Let us cling to what suits us; we are the best judges of that.[157]

Polls are indeed faceless guides to literary exploration.

We will probably be better served by choosing personal canons carefully. If you, like Jorge Luis Borges, Italo Calvino and Vladmir Nabokov, are a confirmed admirer of Robert Louis Stevenson, then you will read Stevenson’s essay, “Books Which Have Influenced Me” and other similar essays.[158] Of course, then you will be very much off the beaten path in canon exploration, trying out books such as Dumas’s Vicomte de Bragelonne, the third volume in the Three Musketeers saga (“Perhaps my dearest and best friend outside of Shakespeare is D’Artagnan, the elderly D’Artagnan of the Vicomte de Bragelonne”), George Meredith’s The Egotist (“I have read The Egoist five or six times myself, and I mean to read it again”), John Bunyan’s Pilgrim’s Progress (“a book that breathes of every beautiful and valuable emotion”), Walter Scott’s Redgauntlet and Guy Mannering, non-fiction such as George Borrow’s The Bible in Spain, and Hazlitt’s essay, “On the Spirit of Obligations” (an essay on ethics that was, Stevenson wrote, “a turning-point in my life”). Most of these won’t be found on the Norwegian Book Club poll. Stevenson’s essay is an idiosyncratic personal canon, not an opinio communis. But in reading it, you will have, not just a bare list, but a description of and a judgment on a book, as you are experiencing the author’s inimitable and humane prose.

[1] Harold Bloom, The Western Canon: The Books and School of the Ages (New York: Harcourt Brace, 1994). David Dubal, The Essential Canon of Classical Music (NY: North Point Press 2003). Paul Schrader, “Canon Fodder: As the Sun Finally Sets on the Century of Cinema, by What Criteria Do We Determine its Masterworks?” Film Comment (September/October 2006): 33-49. Other references to “the” canon in literature: Trevor Ross, The Making of the English Literary Canon (Montreal: McGill-Queen’s University Press, 1998); Jonathan Kramnick, Making the English Canon: Print Capitalism and the Cultural Past, 1700-1770 (Cambridge: Cambridge University Press, 1998); Lillian S. Robinson, “Treason Our Text: Feminist Challenges to the Literary Canon,” Tulsa Studies in Women’s Literature 2.1 (Spring, 1983): 83-98. In music: Marcia J. Citron, Gender and the Musical Canon (Urbana: University of Illinois Press, 2000); Timothy J. Dowd, Kathleen Liddle, Kim Lupo, Anne Borden, “Organizing the Musical Canon: the Repertoires of Major U.S. Symphony Orchestras, 1842 to 1969,” Poetics 30 (2002): 35–61; Klaus Pietschmann and Melanie Wald-Fuhrmann, eds., Der Kanon der Musik: Theorie und Geschichte. Ein Handbuch (Munich: Edition text + kritik, 2002). For the concept of canon in art, Anna Brzyski, ed., Partisan Canons (Durham, North Carolina: Duke University Press, 2007).

[2] Henry George Liddell, Robert Scott, Henry Stuart Jones and Roderick McKenzie, A Greek-English Lexicon (Oxford: Oxford University Press, 1996). From kanna, kanê, a reed, cane. In Latin, canon means “A marking or measuring line; hence, a rule, canon, model.” In ecclesiastical Latin, “a catalogue of sacred writings, as admitted by the rule”; “the Canon.” Charlton T. Lewis, Charles Short, Latin Dictionary Founded on Andrew’s Edition of Freund’s Latin Dictionary (Oxford: Oxford University Press, 1879). See also Bruce M. Metzger, The Canon of the New Testament: Its Origin, Development and Significance (Oxford: Clarendon Press, 1987), 289-93.

[3] Richard Tobin, “The Canon of Polykleitos,” American Journal of Archaeology 79.4 (Oct., 1975): 307-321.

[4] Letter to Pompeius, 3. George A. Kennedy, “The Origin of the Concept of a Canon and Its Application to the Greek and Roman Classics,” in Canon vs Culture: Reflections on the Current Debate, edited by Jan Gorak (New York: Garland Publishing, 2001), 105-16, at 106.

[5] Quintilian, The Orator’s Education (Institutio Oratoria), 1.4.3, 10.1.54, 59, 61. In this work, Quintilian supplies his own list of authors an orator should read. Rudolf Pfeiffer, History of Classical Scholarship from the Beginnings to the End of the Hellenistic Age (Oxford: Clarendon Press, 1968), 203-7. Gregory Nagy, “Early Greek Views of Poets and Poetry,” in The Cambridge History of Literary Criticism, vol. 1, edited by George A. Kennedy (Cambridge: Cambridge University Press, 1989), 1–77, at 1, 46, notes that the process of judging “best” literary works had started long before Aristophanes and Aristarchus, in judged poetry and drama competitions. See also Pfeiffer 204 n. 4.

[6] 10.1.54.

[7] John Edwin Sandys, A History of Classical Scholarship: from the Sixth Century B.C. to the End of the Middle Ages, 3rd ed. (Cambridge: Cambridge University Press, 1921), 131.

[8] The poetess Corinna was added to the nine in some listings. Commentary of Melampus or Diomedes on Dionysius of Thrace (p. 21, Hilgard), as quoted in David A. Campbell, ed. and tr., Greek Lyric, Volume IV: Bacchylides, Corinna, and Others, Loeb Classical Library (Cambridge MA: Harvard University Press, 1992), 24-25. See also 18-23; Bekker, Anecdota Graeca, 2:751. Though Corinna continued to have a high reputation among Roman authors and critics, there are major problems with the dating of her life and poetry.

[9] Notably David Ruhnken in 1768, see below. For ancient canons, see also Margalit Finkelberg and Gedaliahu A. G. Stroumsa, eds., Homer, the Bible, and Beyond: Literary and Religious Canons in the Ancient World (Leiden: Brill, 2003).

[10] Commentary on Matthew, 28, as quoted in B. F. Westcott, A General Survey of the History of the Canon of the New Testament, 7th ed. (Cambridge: Cambridge University Press, 1896), 518. Eusebius (Church History 6.36) dates this work to the author’s later life, about 246-48 CE. For the New Testament canon, see also Metzger, The Canon of the New Testament.

[11] As quoted in Westcott, A General Survey, 448, 550-51.

[12] Luther, Preface to the New Testament (1522).

[13] Metzger, The Canon of the New Testament, 240-41, 253.

[14] For the process of decanonization, see A Van Debeek and Karel Van Der Toorn, eds., Canonization and Decanonization (Leiden: Brill, 1998).

[15] Websters Dictionary.

[16] For the idea of the classic in literature, Pfeiffer, History of Classical Scholarship, 206-7; T. S. Eliot., What Is a Classic? (London: Faber & Faber, 1945) (Eliot’s answer to his own question here was Virgil); Frank Kermode, The Classic (NY: Viking Press, 1975); Hans Ulrich Gumbert, “Phoenix or Ashes: From Canon to Classic,” trans. R. Norton, New Literary History 20.1 (1988): 141-163; Italo Calvino, Why Read the Classics?, tr. Martin McLaughlin (London: Cape, 1999); Tansu Acik, “What is a Classic According to T. S. Eliot and H.-G. Gadamer?” The International Journal of the Humanities 8.8 (2010): 53-63.

[17] Pfeiffer, History of Classical Scholarship, 207; Jan Gorak, The Making of the Modern Canon: Genesis and Crisis of a Literary Idea (London and Atlantic Highlands, NJ: Athlone Press, 1991), 51; Seth L. Schein, “Our Debt to Greece and Rome: Canons, Class and Ideology,” in A Companion to Classical Receptions, edited by Lorna Hardwick, Christopher Stray (Malden, MA: Blackwell, 2008), 75-85, 82.

[18] Gorak, The Making of the Modern Canon, 51.

[19] Pfeiffer suggests that as a result of being “chosen in” to one of these lists, an author would be more widely copied to allow for reading in schools and by the public; scholars would annotate and write about the author. These are convincing suggestions, but such developments would happen only at Alexandria, in an institutional way. Pfeiffer, History of Classical Scholarship, 208.

[20] And in fact, I believe “canon” can be validly applied to literature only with the limited definition I propose below.

[21] Reprinted in The Course of Time, Book IV (Hartford: Silas Andrus and Son, 1848), p. 120.

[22] Van Wyck Brooks, “On Creating a Usable Past,” The Dial 64.7 (11 April 1918): 337-41 at 338. Gorak, The Making of the Modern Canon, 66.

[23] As quoted in Gorak, The Making of the Modern Canon, 64.

[24] “Toward a New Canon,” The Nation 134.3484 (Apr. 13, 1932): 429-430. Sometimes critics admit that “the canon” can change—which begs the question, which of the many chronological, geographical and contradictory manifestations of canon is “the canon”? For example, Jay B. Hubbell, Who Are the Major American Writers? A Study of the Changing Literary Canon (Durham NC: Duke University Press, 1972). This valuable book is flawed by frequent references to “the” American canon.

[25] Gorak, The Making of the Modern Canon, 86.

[26] “The Discipline of Letters,” (1943), repr. in F. R. Leavis, ed., A Selection from Scrutiny (Cambridge: Cambridge University Press), 1:7-21, at 17; Gorak, The Making of the Modern Canon, 80.

[27] This book was a major influence on Harold Bloom. “An Interview with...   Harold Bloom on Literary Criticism” on the Five Books website, at (accessed 2-27-15).

[28] (accessed 4/16/15).

[29] The first example OED cites is from 1929, “the canon of literature.”

[30] See The Best Hundred Books, By the Best Judges. New and revised edition. Pall Mall Gazette “Extra” No. 24 (London: Pall Mall Gazette Office, 1886). Ruskin’s response is on page 7 of this publication. Though he removed at least half of Lubbock’s list, he only added a few books in replacement: Pindar, the first two books of Livy, and Clouds, Birds, and Plutus of Aristophanes, to replace the comic poet’s Knights.

[31] Lists of best works, such as top ten lists, have been denigrated as trivial. However, novelist-critic Umberto Eco has recently argued that lists, far from being non-serious, are one of the basic building blocks of civilization. Umberto Eco, The Infinity of Lists (London: MacLehose, 2009).

[32] For example, Alastair Fowler, “Genre and the Literary Canon,” New Literary History 11 (1979): 97-119, 98; Jonathan Rosenbaum, Essential Cinema: On the Necessity of Film Canons (Baltimore: Johns Hopkns University Press, 2004), 407; Joan Lipman Brown, Confronting Our Canons: Spanish and Latin American Studies in the 21st Century (Lewisburg, PA: Bucknell University Press, 2014), 46. Borges had a tendency to use the term “personal” to refer to his selections of “best” works, as in his A Personal Anthology (Antología Personal) (1961) and the series of books he was editing at the time of his death, the Personal Library (Biblioteca Personal).

[33] With all due respect to Mr. Bloom and Mr. Dubal, though their books are based on fundamentally flawed ideas, these books are still valuable overviews of genres. The Western Canon is not the Western Canon, but the Bloom canon is useful.

[34] Published as the prologue to The Tale of the Tub in 1704.

[35] It also ranks high in the personal canon of horror writer Clive Barker. When asked what three books he would take to the afterlife with him, he said, “And, I need a really really funny book, because Heaven can get dull once in a while, Three Men In A Boat, by Jerome K. Jerome, one of the funniest books I’ve ever read.” His other two books were Peter Pan and Moby Dick. Interview, with Lost Souls, May 18, 2001 at (accessed 3/1/15). “Which books would you take with you to the afterlife” is a variation on the widespread “desert island” books theme, which has produced many short canons.

[36] Guardian, 23 January 2009, at

[37] T.S. Eliot, “Dante” (1929), from Selected Essays (London, Faber and Faber, 1932).

[38] John Keats, letter to Benjamin Robert Haydon, May 11, 1817.

[39] Alan Cameron, The Greek Anthology: From Meleager to Planudes (Oxford: Clarendon Press, 1993). For the more recent history of anthologies, Anne Ferry, Tradition and the Individual Poem: An Inquiry into Anthologies (Stanford: Stanford University Press, 2001).

[40] Leslie Schenk, “The Western Canon,” World Literature Today, 70.2 (spring, 1996): 325-328.

[41] Mortimer Adler and Robert Hutchins, who edited the Great Books of the Western World series, fifty-four volumes, in the 1950s. In fairness to Adler and Hutchins, the list was labeled Western. However, Dwight Macdonald, “The Book-of-the-Millennium Club,” The New Yorker (November 29, 1952), 171, points out that this canon leaned toward British literature, leaving out such major French authors as Molière, Corneille, Racine, Stendhal, Balzac, and Flaubert.

[42] “The Chinese Classic Novel in Translation: The Art of Magnanimity,” published in Rexroth’s Bird in the Bush: Obvious Essays (NY: New Directions, 1959), available at (accessed 3-12-15). Admittedly, Rexroth was a Sinophile, and translated much Chinese and Japanese poetry himself. He created a far-ranging canon of sorts in his series of essays, “Classics Revisited,” which was republished in Classics Revisited (Chicago: Quadrangle, 1968) and More Classics Revisited (NY: New Directions, 1989).

[43] Bloom includes some non-Western works (though nothing Chinese) in his appendices “because of their influence on the Western canon.” Western Canon, 561. Here, Bloom’s logic doesn’t seem clear. He puts these works in his appendices (which would seem to place them in the Western canon); but he also says that they only “influenced” the “Western canon,” which would seem to set them apart from the real “Western canon.”

[44] Journals of Ralph Waldo Emerson: with Annotations, edited by Edward Waldo Emerson and Waldo Emerson Forbes (Boston: Houghton Mifflin Company, 1913), VII:511. The Bhagavad-Gita also profoundly influenced Thoreau. Philip Cafaro, Thoreau’s Living Ethics: Walden and the Pursuit of Virtue (Athens, GA: University of Georgia Press, 2004), 71. In Walden, he wrote, “In the morning I bathe my intellect in the stupendous and cosmogonal philosophy of the Bhagvat Geeta, since whose composition years of the gods have elapsed, and in comparison with which our modern world and its literature seem puny and trivial.”

[45] Leo Tolstoy, What Is Art?, tr. Aylmer Maude (NY: Thomas Y. Crowell & Co., 1899), 104-5.

[46] André Maurois, “The Art of Reading,” (1939) in The Art of Living, trans. James Whitall (NY: Harpers, 1940). Compare Andrew Lang: “In literature, as in love, one can only speak for himself.” Adventures Among Books (Longmans, Green and Co., 1912 ), 4. Maurois and Lang thus reject the idea of an ironclad authoritarian canon.

[47] “100 Best Novels,” at (accessed 3-1-15). One problem with this poll is that it was organized by a publishing house, which obviously could be a conflict in interest. It would have helped if the ballots had been released (see below on the methodology of polls). Ten members of the editorial board voted, Daniel J. Boorstin, A. S. Byatt, Christopher Cerf, Shelby Foote, Vartan Gregorian, Edmund Morris, John Richardson, Arthur Schlesinger Jr., William Styron and Gore Vidal.

[48] Virginia Woolf Diary, 6 September 1922, in Virginia Woolf, A Moment’s Liberty: the Shorter Diary‎, edited by Anne Olivier Bell (San Diego: Harcourt Brace Jovanovich, 1990), 148. “How It Strikes a Contemporary,” in Virginia Woolf, The Common Reader (New York: Harcourt, Brace and Co., 1925), 240. For complexities in Woolf’s generally negative response to Ulysses, see James Heffernan, “Woolf’s Reading of Joyce’s Ulysses, 1918-1920,” and “Woolf’s Reading of Joyce’s Ulysses, 1922-1941,” at's_Reading_of_Joyce's_Ulysses,_1918-1920 and,_1922-1941 (accessed Feb. 26, 2016). Heffernan curiously contends that Woolf had not read Ulysses all the way through, despite explicit statements in her diaries and elsewhere that she had.

[49] “Blindness,” in Seven Nights (New York: New Directions Pub. Corp., 1984), 118-19. The early Borges admired Ulysses, but not the later. Woolf agreed that the book was not an easy read: “I rather agree that Joyce is underrated, but never did any book so bore me.” Letter to Gerald Brenan, December 1923, Letters 3:80, as cited Heffernan, “Woolf’s Reading of Joyce’s Ulysses, 1922-1941.”

[50] Richard Burgin, Conversations with Jorge Luis Borges (NY: Holt, Rinehart and Winston, 1969), 56.

[51] Isaac Bashevis Singer and Richard Burgin, Conversation with Isaac Bashevis Singer (Garden City, NY: Doubleday, 1985), 96. See also and Grace Farrell, ed., Isaac Bashevis Singer: Conversations (Jackson: University Press of Mississippi, 1992), 146-60, 74.

[52] For example, Michael Chabon, “What to Make of Finnegans Wake?” New York Review of Books, July 12, 2012, at (accessed 3/5/15). Joyce Carol Oates, “Jocoserious Joyce,” Critical Inquiry 2:4 (Summer, 1976): 677-688.

[53] See Douwe W. Fokkema, “The Canon as Instrument for Problem Solving,” in Sensus Communis: Contemporary Trends in Comparative Literature; Festschrift für Henry Remak, edited by János Riesz, Peter Boerner, and Bernhard Scholz (Tübingen: Gunter Narr Verlag, 1986), 245-54, at 246 (citing Fritz J. Raddatz, “Die Europäische Literature-Gemeinschaft,” in Die Zeit, June 8, 1984, p. 45).

[54] Reginald F. Christian, Tolstoy’s Letters, 1880-1910 (New York: Scribner, 1978), vol. 2, p. 533.

[55] “Blaming the Bard,” Review of Cymbeline, Sept. 26, 1896, Saturday Review, see Bernard Shaw, Shaw on Shakespeare: an Anthology of Bernard Shaw’s Writings on the Plays and Production of Shakespeare, ed. by Edwin Wilson (NY: E. P. Dutton, 1961), 50.

[56] Gabriel García Márquez, “The Mysteries of Bill Clinton,” translated by Alastair Reid, Salon (Feb. 1, 1999), at (accessed  3/3/15). 

[57] Márquez apparently ranked this and books such as Stevenson’s Treasure Island so highly because veneration for story was an important part of his aesthetic outlook. In addition, he also admired Dumas’s characterization. George Bernard Shaw referred to Dumas’s novels as a summit of art . . . you get nothing above Dumas on his own mountain.Dramatic Opinions and Essays with an Apology (NY: Brentano’s, 1922), 2:308.

[58] Gabriel García Márquez, Fragrance of Guava: Plinio Apuleio Mendoza in Conversation with Gabriel Garcia Márquez, tr. Ann Wright (London: Verso, 1983), 117.

[59] Letter written on Oct. 25, 1891, in Christian, Tolstoy’s Letters, 2:484-86.

[60] Willis Barnstone, With Borges on an Ordinary Evening (Urbana : University of Illinois Press, 1992), 174-76.

[61] (accessed 3/5/15).

[62] Jay Carr, ed., The A List: the National Society of Film Critics’ 100 Essential Films (Cambridge, MA: Da Capo Press, 2002). This books describes its contents as “a list of the 100 movies you need to know to be film literate.” It also leaves out Fellini’s 8-1/2, Renoir’s Grand Illusion, Chaplin’s The Gold Rush and City Lights, Griffith’s Intolerance, Antonioni’s L’Avventura, and Tarkofsky’s Andrei Roublev. Such omissions would be entirely understandable in a personal canon, less so in a book that purports to include the hundred most important movies in an authoritative way. Film critic Pauline Kael often picked Intolerance as the greatest movie ever made, and Orson Welles often awarded Grand Illusion the same distinction.

[63] J. Peder Zane, The Top Ten: Writers Pick Their Favorite Books (NY: W.W. Norton, 2007). See also (accessed 4-14-15).

[64] Lubbock had also included Charles Kingsley and Edward Bulwer-Lytton in the “Modern Fiction” section, odd choices from our perspective. A little later, in 1897, British critic Clement Shorter picked “A Hundred Books for a Village Library,” and included six books by Scott. The Bookman (Oct. 1897), 13-14. He also included John Gibson Lockhart’s unabridged life of Scott (1838). He found room for only two books by Dickens, The Pickwick Papers and David Copperfield, on his list.

[65] Johann Peter Eckermann, Conversations with Goethe (New York: M. Walter Dunne, 1901), 528, 334-35. There are many other enthusiastic references to Scott in this book. Balzac also praised the “creative genius” of Scott and listed six of his books as masterpieces. The Letters of Honoré de Balzac to Madame Hanska . . . (Boston: Little, Brown, and Company, 1900), 471-72.

[66] Fowler, “Genre and the Literary Canon.”

[67] For an introduction, see L. D. Reynolds and Nigel Guy Wilson, Scribes and Scholars: a Guide to the Transmission of Greek and Latin Literature, 4th ed. (Oxford: Oxford University Press, 2013).

[68] William Jacob Keller, Goethe’s Estimate of the Greek and Latin Writers: As Revealed by his Works, Letters, Diaries, and Conversations, Bulletin of the University of Wisconsin, No. 786. Philology and Literature Series. Vol. 6. No. 1 (Madison, WI: University of Wisconsin, 1916).

[69] Bernard Shaw, The Great Composers: Reviews and Bombardments, ed. Louis Crompton (Berkeley: University of California Press, 1978), 148.

[70] Rosenbaum’s first top ten and last have no films in common: For the 1982 Sight & Sound poll, the critic supplied this top ten: Playtime, Gertrud, Breathless (A Bout de Souffle), The Eclipse, The Story of the Late Chrysanthemums, City Lights, Celine et Julie vont en Bateau, Night at the Crossroads (La Nuit de Carrefour), Providence, and Too Early/Too Late (Trop tôt, trop tard). Thirty years later, in 2012, he chose these ten movies for the Sight & Sound poll: Cuadecuc-Vampir, Greed, Histoire(s) du cinéma, I Was Born, But…, Ivan, Rear Window, Sátántangó, Spione, The Wind Will Carry Us, and The World.

[71] Summarized in Wikipedia at and

[72] Nathalie Crom, “Les 10 livres préférés de 100 écrivains francophones,” March 14, 2009, at,40420.php (accessed 8-22-15).

[73] Zane, The Top Ten. See also (accessed 4-14-15).

[74] Raymond Queneau, Pour une Bibliothèque Idéale (Paris: Gallimard, 1956), as cited at (accessed 6-13-16).

[75] The top ten: Tolstoy, Anna Karenina at number 1; then Flaubert, Madame Bovary; Tolstoy, War and Peace; Nabokov, Lolita; Twain, Huckleberry Finn; Shakespeare, Hamlet; Fitzgerald, The Great Gatsby; Proust, In Search of Lost Time; Chekhov, Stories; and Eliot, Middlemarch.

[76] Among the Quixote’s fans were Flaubert (who would hold a communal reading of the Quixote every Sunday afternoon at his home), Dostoevsky (“There is nothing deeper and more powerful in the whole world than this piece of fiction. It is still the final and the greatest expression of human thought.”), Turgenev (see his essay “Hamlet and Don Quixote,” 1860, contrasting these two fictional archetypes; Turgenev’s sympathies are definitely on the side of the idealistic knight); and Dickens (whose first great book was modeled on the Quixote, with a noble, idealistic protagonist and a Cockney Sancho Panza). Dostoevsky, A Writer´s Diary (Chicago: Northwestern University Press, 1997), 22:92, as quoted in Kenneth Lantz, The Dostoevsky Encyclopedia (Westport, CT: Greenwood Publishing Group, 2004), 51.

[77] “Cien escritores en español eligen los 100 libros que cambiaron su vida,” El Pais (August 2008), as reported in (accessed Feb. 17, 2016).

[78] Gabriel García Márquez, “Asombro por Juan Rulfo,” a talk given September 18, 2003, at (accessed Feb. 17, 2016).

[79] (accessed 3-5-15). The world-wide scope of this poll is admirable, however. It includes such books as the Epic of Gilgamesh, the 1001 Nights, The Mahabharata and Tale of the Genji.

[80] Frank O’Connor, The Lonely Voice: A Study of the Short Story (London: MacMillan, 1962), 145, 57-58.

[81] “Why Do People Read Mystery Stories?” New Yorker (October 14, 1944), at (accessed 4-23-15), “Who Cares Who Killed Roger Ackroyd? : A Second Report on Detective Fiction,” The New Yorker (June 20, 1945) and “Mr. Holmes, They Were the Footprints of a Gigantic Hound,” The New Yorker (Februrary 17, 1945). This is an example of a critic attempting to put an entire popular genre beyond the pale of respectable literature.

[82] “One should think of reading as a form of happiness, as a form of joy, and I think that compulsory reading is wrong. You might as well talk of compulsory love or compulsory happiness.” Borges said that he had always taught that “if a book bores you, lay it aside. It hasn’t been written for you.” Willis Barnstone, ed., Borges at Eighty: Conversations (Bloomington: Indiana University Press, 1982), 113. Professor Borges: A Course on English Literature, edited by Martín Arias and Martín Hadis, tr. Katherine Silver (NY: New Directors Books, 2013), 252.

[83] Boswell’s Life of Johnson, edited by George Birkbeck Hill, 6 volumes (NY: Harper & Bros., 1904), 1:496. Johnson here also recommended that a young person read five hours a day.

[84] See above, also W. B. Carnochan, “Where Did Great Books Come From Anyway?” Stanford Humanities Review 6.1 (1998), at (accessed 3/5/15).

[85] Percy Lubbock, The Pleasures of Life (London: MacMillan & Co., 1897), vii-x. The essay with the list, “The Choice of Books,” is at 70-93. In addition, Lubbock states that he included some works (the Analects of Confucius and the Sheking, Chinese Odes) purely for their impact on civilization, even though he did not “greatly admire” them. This is the canonical criterion of cultural impact, see below.

[86] Bloom, The Western Canon, 548.

[87] Bloom, The Western Canon, 160.

[88] Brander Mathews, Molière: his Life and his Works (NY: Charles Scribner’s Sons, 1910), 183.

[89] Strabo refers to Philetas of Cos (died ca 285 BCE) as “poet and critic [kritikós].” Strabo 14.2.19 C657. Pfeiffer, History of Classical Scholarship, 89. Gregory Nagy, “The Library of Pergamon as a Classical Model,” online edition (2011), at (accessed Feb. 17, 2016).

[90] Barbara Herrnstein Smith, “Contingencies of Value,” Critical Inquiry 10.1, Canons (Sep. 1983): 1-35, 3-5. While I believe that this rejection of enlightened literary selection is unhealthy, it is true that some critics might be interested in other aspects of literary scholarship. And the “selective” aspect of criticism can be done badly, shallowly, arrogantly.

[91] See above, on contested views of Joyce and Shakespeare, Dumas, and Gogol’s “Viy.” Gorak speaks of the “Visionary Canon” of Frye, a canon “without value judgements, without fixed shape or boundaries, unattached to particular work or authors.” But this would seem to use the word “canon” in an altogether non-standard way. The Making of the Modern Canon, 120-52, esp. 122.

[92] As quoted in Neal B. Houston, “Hemingway: The Obsession with Henry James, 1924-1954,” Rocky Mountain Review of Language and Literature 39.1 (1985): 33-46, 34.

[93] For Dante, see below. Chaucer, The House of Fame, 1429-1512.

[94] For the full story, see the Tchaikovsky Research page on Brahms, (accessed 9/15/15).

[95] Robert Black, in a book on Renaissance education, looks at canon from the perspective of education and schools. See his chapter 4, “Latin authors in medieval and Renaissance Italian schools: the story of a canon,” in Humanism and Education in Medieval and Renaissance Italy: Tradition and Innovation in Latin Schools from the Twelfth to the Fifteenth Century (Cambridge: Cambridge University Press, 2001), 173-274. For modern pedagogical canons, see John Alberti, ed., The Canon in the Classroom: Pedagogical Implications of Canon Revision in American Literature (New York and London: Garland, 1995); Barbara Mujica, “Teaching Literature: Canon, Controversy, and the Literary Anthology,” Hispania 80.2 (May, 1997): 203-215; Arthur N. Applebee, “Stability and Change in the High-School Canon,” The English Journal 81.5 (Sep., 1992): 27-32.

[96] The Boston Evening Transcript, 17 March 1885, as quoted in (accessed 3-19-15). Critics generally failed to come to Huckleberry Finn’s defense at this time. Arthur Lawrence Vogelback, “The Publication and Reception of Huckleberry Finn in America,” American Literature 11.3 (Nov., 1939): 260-272, at 267, 269.

[97] Aylmer Maude, ed. and tr., Tolstoy on Art (London: Humphrey Milford Oxford University Press, 1924), 288.

[98] Tolstoy on Art, 224. Tolstoy was scathing on the idea that only “Uebermenschen” could appreciate the greatest art, and that such art was often too obscure for the lesser, common man, the “vulgar herd.” Ibid., 196.

[99] Sesame and Lilies: Three Lectures (1865), lecture 1, “Of Kings’ Treasuries,” p. 8.

[100] W. Somerset Maugham Selects the World’s Ten Greatest Novels (Fawcett Publications, 1962), 199. The ten novels he chose were Tolstoy, War and Peace, Balzac, Old Man Goriot, Fielding, Tom Jones, Austen, Pride and Prejudice, Stendhal, The Red and the Black, Brontë, Wuthering Heights, Flaubert, Madame Bovary, Dickens, David Copperfield, Dostoyevsky, The Brothers Karamazov, and Melville, Moby Dick.

[101] Borges described his method of judging literature to be “pure aesthetic pleasure,” see his essay “Literary Pleasure” (1927). Somerset Maugham wrote; “I have always felt that reading should be a pleasure. Of course to get anything out of it you must give it your full attention, but to a healthy understanding there is nothing disagreeable in the activity of the intellect. It is however the business of an author to make your perusal of his work enjoyable.” Somerset Maugham’s Introduction to Modern English and American Literature (NY: The New Home Library, 1943). John Carey, Pure Pleasure: a Guide to the Twentieth Century’s Most Enjoyable Books (London: Faber and Faber, 2000).

[102] “A Gossip on Romance,” in Memories and Portraits (1887).

[103] See below, the characterization section.

[104] Bloom, The Western Canon, 29, seems to see style as the primary aesthetic factor for entering “The Western Canon.” One “breaks into the canon only by aesthetic strength, which is constituted primarily of an amalgam: mastery of figurative language, originality, cognitive power, knowledge, exuberance of diction.” The first and last elements in this list, its alpha and omega, specifically relate to style.

[105] See above.

[106] Flannery O’Connor, Mystery and Manners, 77. Frank O’Connor, interview, with Anthony Whittier, “The Art of Fiction No. 19,” Paris Review No. 17 (Autumn-Winter 1957), at (accessed 3/15/15). He talked about the “bony structure” needed for fiction, and how Chekhov has it, but that the imitators of Chekhov often lack it. See also Gabriel García Márquez’s views of Dumas, above.

[107] “Are Children the Ultimate Literary Critics?”, in Singer, Stories for Children (NY: Farrar, Straus, Giroux, 1984). Singer was particularly scathing about the idea that “writers who require elaborate commentaries and countless footnotes are the true creative geniuses of our time.” Though we often think of Borges as somewhat avant-garde, he felt that the short-story form had these “indispensable elements”: “economy and a clearly stated beginning, middle and end.” “An Autobiographical Essay,” in The Aleph and Other Stories, 1933-1969 (NY: E.P. Dutton, 1970), 238. This is one of the reasons he had a strong inclination for Stevenson, Kipling, Wells, and mystery fiction, Poe, Chesterton, Doyle, Eden Philpotts, Ellery Queen. “The Detective Story” (1978), in Borges, Selected Non-Fictions, ed. Eliot Weinberger (New York : Viking, 1999), 499 and Borges Oral (Buenos Aires: Emecé Editores, 1979). Our chaotic age “tends to eliminate character, plot; everything is very vague.” Yet despite this, “one thing has humbly maintained the classic virtues: the detective story. For a detective story cannot be understood without a beginning, middle, and end.”

[108] The grotesquely comic nurse from Dickens’s Martin Chuzzlewit.

[109] M. Thomas Inge, ed., Conversations with William Faulkner (Jackson, MI: University Press of Mississippi, 1999), 185.

[110] Flannery O’Connor, Mystery and Manners, 90.

[111] The Western Canon, 29-30.

[112] Curtius, European Literature, 445. Colin Burrow, “Virgils, from Dante to Milton,” in The Cambridge Companion to Virgil, edited by Charles Martindale (Cambridge: Cambridge University Press, 1997),79-90, at 79. Amy Richlin, “School Texts of Persius and Juvenal,” in A Companion to Persius and Juvenal , edited by Susanna Braund and Josiah Osgood (Malden, MA: Blackwell Publishing, 2012), 465-85, at 467. “School texts repeatedly justify their existence on moral grounds.” Alexander Neckam (or Nequam) (1157 – 1217 CE) advised the student to “store up the moral sayings of Juvenal in the depths of his heart and strive above all to avoid natural sin.” As quoted in ibid., 469. Curtius, European Literature, 18, 445, 590. Virgil’s Aeneid was also regarded as a moral guide, and Dante almost portrayed Virgil as a poet who leads to Christian conversion.

[113] Maude, Tolstoy on Art, 47.

[114] What Is Art, in Maude, Tolstoy on Art, 310.

[115] “The Progress of the Novel” (1929), in Arnold Bennett, The Author’s Craft and other Critical Writings, edited by Samuel Hynes (Omaha: University of Nebraska Press, 1968), 90.

[116] Arnold Bennett, “Books and Persons: The Twelve Finest Novels,” Evening Standard, March 17, 1927, p. 5, repr. in Andrew Mylett, ed., Arnold Bennett: The ‘Evening Standard’ Years. Books and Persons 1926- 1931 (London: Chatto & Windus, 1974), 32-34. The novels on his list are: The Brothers Karamazov, The Idiot, The House of the Dead, Crime and Punishment, Anna Karenina, War and Peace, Resurrection, Torrents of Spring, Virgin Soil, On the Eve, Fathers and Sons, and Dead Souls. Bennett felt that English novelists, such as Dickens, ranked nowhere near the great Russians. Paradoxically, the two highest-ranked novelists on Bennett’s list, Dostoevsky and Tolstoy, were both significantly influenced by Dickens. Tolstoy once said, “If you sift the world’s prose literature, Dickens will remain. Sift Dickens, David Copperfield will remain.” Tatyana Tolstoy, Tolstoy Remembered (London: Michael Joseph, 1977), 168, as cited in Philip Rogers, “A Tolstoyan Reading of David Copperfield,” Comparative Literature 42.1 (Winter, 1990): 1-28, 1. In 1852, Tolstoy wrote in his diary, “Read David Copperfield [in Russian]— a delight.” He subsequently worked through the lengthy novel in English, with dictionary in hand.

[117] The Western Canon, 29.

[118] Mystery and Manners, 66.

[119] Ibid., 76.

[120] “Hamlet and Don Quixote,” (1860) as translated by Moshe Spiegel in Chicago Review 17.4 (1965): 92-109.

[121] Willis Barnstone, With Borges on an Ordinary Evening (Urbana : University of Illinois Press, 1992), 104.

[122] Introduction to In Praise of Darkness, in Selected Poems, 265.

[123] Fantastic Tales, Visionary and Everyday (Houghton Mifflin Harcourt Trade & Reference Publishers, 2014), 461, xiv-xvi.

[124] “Books Which Have Influenced Me,” first published in British Weekly, May 13, 1887, reprinted in Stevenson’s posthumous collection, Essays in the Art of Writing (1905).

[125] W. Somerset Maugham, The Art of Fiction ((NY: Doubleday, 1955), 18-19.

[126] Ibid., 29.

[127] “Wordsworth” (1879).

[128] A modern example is Wayne Booth, The Company We Keep: An Ethics of Fiction (Berkeley: University of California, 1988).

[129] Review of Our Mutual Friend, in The Nation (December 21, 1865): 786-7. All the characters listed are from Our Mutual Friend, except Micawber from David Copperfield and Pickwick from Pickwick Papers. See Henry James, Theory of Fiction, edited by James E. Miller, Jr. (University of Nebraska Press, 1972), 214. James also described Dickens as “the greatest of superficial novelists . . . We are aware that this definition confines him to an inferior rank in the department of letters which he adorns; but we accept this consequence of our proposition. It were, in our opinion, an offence against humanity to place Mr. Dickens among the greatest novelists. . . . He has added nothing to our understanding of human character.” Review of Our Mutual Friend.

[130] Inge, Conversations with William Faulkner, 157.

[131] Farrell, Isaac Bashevis Singer: Conversations, 96.

[132] Robertson Davies, The Merry Heart ([NY]: Viking, 1997), 21.

[133] Márquez, in conversation with Mario Vargas Llosa. As quoted in José David Saldívar, The Dialectics of our America: Genealogy, Cultural Critique, and Literary History (Durham: Duke University Press, 1991), 37-38.

[134] Marlise Simons, “Gabriel Márquez On Love, Plagues And Politics,” New York Times, February 21, 1988. (accessed 5-17-2015)

[135] Marlise Simons, “A Talk with Gabriel Garcia Márquez,” December 5, 1982, New York Times Book Review.

[136] O’Connor, Mystery and Manners, 109, 115, 118.

[137] Calvino, Why Read the Classics, 229.

[138] Calvino, Why Read the Classics, 3, 5.

[139] Frank O’Connor, Paris Review interview.

[140] Maude, Tolstoy on Art, 84, 288, 290.

[141] Singer and Burgin, Conversations with Isaac Bashevis Singer, 30.

[142] Mark Twain, “The Disappearance of Literature: Address at the Dinner of the Nineteenth Century Club at Sherry’s, New York,” speech, Nov. 20, 1900.

[143] John Le Carré, “Personal Best,” Sep 30, 1996, at Salon website, (accessed 4/3/15). He went on to recommend Ford Madox Ford’s The Good Soldier, Love In The Time of Cholera and Chronicle of a Death Foretold by Gabriel García Márquez, Tolstoy’s “Death of Ivan Ilyich” and Anna Karenina, the novels of Dickens and Balzac, Herzen’s memoirs, Gibbon’s Decline and Fall of the Roman Empire, and Josephus’s History of the Jews.

[144] I am referring to the critics poll; there is also a directors poll.

[145] Virgil Thomson, The State of Music (1939). Joseph Kerman, Write All These Down: Essays on Music (Berkeley: University of California Press, 1994), 40.

[146] Inferno, 4.98-102, tr. Charles Eliot Norton.

[147] Arthur Elam Haigh and Arthur Pickard-Cambridge, The Attic Theatre: A Description of the Stage and Theatre of the Athenians, and of the Dramatic Performances at Athens, 3rd ed. (Oxford, Clarendon Press, 1907), 31-40; Philip Whaley Harsh, A Handbook of Classical Drama (Palo Alto, CA: Stanford University Press, 1944), 4, 441 n 5; Ian C. Storey and Arlene Allan, A Guide to Ancient Greek Drama (Malden, MA : Blackwell Pub., 2005), chapter 1.

[148] Haigh and Pickard-Cambridge, The Attic Theatre, 35.

[149] Naboth Hedin, “Winning the Nobel Prize,” Atlantic (October 1950), at Burton Feldman, The Nobel Prize: A History of Genius, Controversy, and Prestige (NY: Arcade Pub., 2000), 69-70. Tolstoy was not even nominated in 1901, as the Permanent Secretary of the selection committee disagreed with his political views. Later, Tolstoy unofficially let the Nobel organization know that he would turn down the award if it was given to him.

[150] Feldman, The Nobel Prize, 69.

[151] Aulus Gellius, Attic Nights, 17.4.1, tr. Edmonds, p. 5. See also Haigh and Pickard-Cambridge, The Attic Theatre, 35-36, listing cases in which artistic judges were intimidated or bribed; Ioannis M. Konstantakos, “Rara coronato plausere theatra Menandro? Menander's Success in His Lifetime,” Quaderni Urbinati di Cultura Classica, New Series, 88.1 (2008): 79-106, especially 93.

[152] Haigh and Pickard-Cambridge, The Attic Theatre, 37-38.

[153] Plato, Laws 2.659A-C; 3.700C-701A: “in place of an aristocracy in music there sprang up a kind of base theatrocracy.” Haigh and Pickard-Cambridge, The Attic Theatre, 37-38.

[154] As quoted in “The Storm Over the University,” The New York Review of Books, December 6, 1990, available at (accessed 5/17/15).

[155] Chih-tsing Hsia, The Classic Chinese Novel: A Critical Introduction (New York: Columbia University Press, 1968).

[156] Searles, “The Storm Over the University.”.

[157] André Maurois, “The Art of Reading,” (1939), partially quoted above. Borges’s rejection of compulsory reading in literary education has been quoted above. This is a fundamental rejection of the idea that there is one literary canon which everyone must accept.

[158] First published in British Weekly, May 13, 1887, reprinted in R. L. Stevenson, Essays in the Art of Writing (1905).