Canonicity

Author(s): Wendell V. Harris

Source: PMLA, Vol. 106, No. 1 (Jan., 1991), pp. 110-121
Published by: Modern Language Association
Stable URL: https://www.jstor.org/stable/462827
Accessed: 13-12-2019 20:18 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Modern Language Association is collaborating with JSTOR to digitize, preserve and extend
access to PMLA

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
 Wendell V. Harris

Canonicity

WENDELL V. HARRIS is HE CANONICAL FACTS about the canons of Engli

professor of English at Penn American literature are, first, that there are no canon
State University. His most never have been; second, that there have necessarily alwa
canons; and third, that canons are made up of readings, not o
recent books are The Om-
embodied texts. What is contradictory in that statement resu
nipresent Debate: Empiricism
play on different connotations of the word canon-a critical s
and Transcendentalism in
that is constantly, though often more subtly, in use. As with m
Nineteenth-Century English other critical term, the first step in understanding canon is t
its meanings. The "canon question" then proves much more c
Prose (Northern Illinois UP,
1981) and Interpretive Acts:than
In contemporary ideological criticism admits.

Search of Meaning (Clarendon-

The Inappropriateness of the Biblical Parallel
Oxford UP, 1988). He is cur-
rently completing a dictionary
The well-known core meaning of the Greek kanon is "rule" o
of primary literary concepts.
sure" and, by extrapolation, "correct" or "authoritative." As R
Pfeiffer has pointed out, the first application of the word to
of authors-by David Ruhnken in 1768-was catachrestic (
more nearly precise word than selection was so much needed t
quickly became almost indispensable, despite its entanglemen
concepts of authority and rule not necessarily relevant to
canons. Not surprisingly, the normative sense of the term h
alongside its elective sense: selections suggest norms, and nor
gest an appeal to some sort of authority. However, the criteria fo
ing literary texts are derived not from authority but from c
functions.
The normative sense of canon has been strongly reinforced
application of the term to the accepted books of the Bible, thou
is no agreement on the original force of the word even in this
tion. l The processes by which specific collections of Jewish an
tian writings became closed canons in the first century BC an
fourth century AD, respectively, are not only too complicated f
summary here but, more important, largely irrelevant to the
of the literary canon. In fact, considerable confusion has resul
the seductive apparent parallel between the creation and closin
biblical canons and the formation of lists of literary works th

110

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
 Wendell V Harris 11I

Ruhnken, have been called "canons." Chapter 14 whose essay on the subject has become an almost
of Ernst Robert Curtius's European Literature and obligatory citation, is among those who use the
the Latin Middle Ages, a chapter frequently cited biblical canon as a model: "The desire to have a
in discussions of literary canons, considers early canon, more or less unchanging, and to protect it
catalogs of authors, aspects of the ever-renewed against the charges of inauthenticity or low value
conflict between ancients and moderns, canon (as the Church protected Hebrews, for example,
formation in the church, the medieval canon, and against Luther) is an aspect of the necessary con-
finally modern canons. The section on the church servatism of a learned institution" (77). The point
and the Bible seems to have encouraged Curtius is unarguable, but despite the assumption in "the
to use canon for the authorized literary lists, whose profession of English" that some texts are better
variations from period to period and country to and some interpretations more reasonable than
country he learnedly surveys. But the section on others, admission into the profession hardly re-
modern canon formation essentially considers the quires the candidate to accept any list of texts as
relations among the works that Italy, France, Ger- uniquely necessary to the academic equivalent of
many, and Spain defined as "classical"-an issue salvation.3 The most conservative of our col-
that, for Curtius, depends not on degrees of leagues do not demand that a candidate take, say,
authoritative sponsorship but largely on the degree Johnson, Coleridge, Arnold, and Eliot as "articles
of romanticism in the literature of each country. of faith" and abjure Wharton, Dallas, Gosse, and
It has not been sufficiently noted that this chap- Fish. Nor have universities wielded full authority
ter is titled "Classicism," not "The Canon," and over the canon since they separated themselves
that it primarily concerns the question of how from the medieval church. Until then, of course,
works somehow come to be regarded as "classic." the following excerpt from a bishop's condemna-
Though there have been many lists of approved tion of heresy in the records of the University of
authors through the centuries and across Europe, Paris was not untypical:
they have differed widely; not even Aristotle has
been canonical in the sense of the biblical texts.
Let the body of master Amaury be removed from
This is not to deny that there were traditions-
the cemetery and cast into unconsecrated ground, and
usefully traced by Bruce Kimball-in which the same be excommunicated by all the churches of
authors were regarded as "standard" because they the entire province.... The writings of David of
seemed to inculcate the right moral and intellec- Dinant are to be brought to the bishop of Paris be-
tual principles or to demonstrate a mastery of fore the Nativity and burned.
accurate thinking.2 But the catalogs identifying es- Neither the books of Aristotle on natural philos-
pecially valuable works not only varied consider- ophy nor their commentaries are to be read at Paris
ably, they did not fence others out. The texts that in public or secret, and this we forbid under penalty
one ought to have read differed from others in de- of excommunication. He in whose possession the
writings of David of Dinant are found after Nativity
gree; they were not absolutely distinct like the
shall be considered a heretic. (Thorndike 26-27)
biblical books, which very quickly came to be
regarded as different in kind. Obviously the very
entelechy of the process of biblical canonizing was The most cogent portion of Kermode's essay is
toward closure, whereas literary canons have al- his suggestion that certain texts somehow become
ways implicitly allowed for at least the possibility "licensed" for exegesis and are thereafter subject
of adding new or revalued works. to "interminable" explication (83). What Kermode
Though the sense of "unquestionably and so aptly describes, however, is canon formation
uniquely authoritative" that belongs to the bibli-
not through a work's acceptance into a severely
cal canon (and to the theologically derived en- limited set of authoritative texts but through its in-
dorsements and prohibitions churches enforce on troduction into an ongoing critical colloquy. The
members of their faith) continually colors the de- analogy with a colloquy or conversation functions
bate over the modern literary canon, the analogy in several ways. In a given time and place there are
is more dramatic than helpful. Frank Kermode, events and topics that everyone presumably knows

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
112 Canonicity

about. Some of these have been of continuous in- A Multiplicity of Canons and the Pressures
terest, or at least have had a place in a society's cul-on Them
tural discourse from generation to generation;
others have recently gained attention and will
quickly fade. What proves interesting in general Alastair Fowler's discrimination of six kinds of
conversation will depend on what the conversa-canons has met with general acceptance. The
tionalists are at that time accustomed to discuss- potential canon "comprises the entire written
ing. There are topics that sustain no more than acorpus, together with all surviving oral literature."
few moments' talk, but the cleverness with whichThe accessible canon is that portion of the poten-
a subject is introduced has a great deal to do withtial canon available at a given time. Lists of
whether it is taken up. While there are conversa-authors and texts-as in anthologies, syllabi, and
tional unfortunates who wrap everything they sayreviewers' choices-are selective canons. What
in dullness, others do the opposite, using wit, an Fowler calls the official canon is, I take it, a blend-
ability to see unexpected significances, and in-ing of such lists. What individual readers "know
triguing modes of argument to lend at least mo- and value" are personal canons. And, finally, the
mentary interest to whatever topic they initiate.critical canon is made up of those works or parts
Precisely the same distinctions operate within the of works that are repeatedly treated in critical ar-
critical colloquy. ticles and books (98-99). As useful as these dis-
On this analogy, the historical resonance of atinctions are, it is equally useful to recognize the
text (the degree to which it explicitly relates to variety of the principles defining them, the loose-
other texts), the possible multiplication of its sig-ness of the resulting definitions, and the need for
nificances (the degree to which it is multivalent),4additional classifications. Fowler's potential
the skill with which it is brought into the criticalcanon, for instance, is defined by total inclusivity,
colloquy (the degree to which it finds fortunate but the degree to which what it includes goes be-
sponsorship), and the congruence between its pos- yond traditional definitions of literature will de-
sible significances and critics' current preoccupa-pend on one's critical allegiances. The accessible
tions (the degree to which it proves malleable)-allcanon is also inclusive, but only for a single loca-
these interact to determine how much interest the tion; perhaps further, it varies with the sophisti-
text can sustain over how long a period. Instead cation of each reader. Personal canons seem made
of stamping works with authority, literary canonsup of an indeterminate interaction between all the
propose entries into a culture's critical colloquy. works individuals have read and those they prefer
This colloquy is nothing but a corner of that "un-to some degree or other. Since official and criti-
ending conversation" Kenneth Burke so memora- cal canons are precipitated out of the mass of se-
bly describes: lective canons, the only canons produced by
systematic choice are the innumerable and heter-
ogeneous selective ones.
Imagine that you enter a parlor. You come late. When Additional distinctions suggest themselves as
you arrive, others have long preceded you, and theyuseful, though like Fowler's they rest on no sys-
are engaged in a heated discussion, a discussion too tematic taxonomic principle and therefore over-
heated for them to pause and tell you exactly what itlap at various points. The term canon as applied
is about. In fact, the discussion had already begunto a closed, uniquely authoritative body of texts,
long before any of them got there, so that no one
such as the Bible, fits nowhere in his six
present is qualified to retrace for you all the steps that
classifications; it represents a seventh kind
had gone before. You listen for a while, until you de-
(canon7). If we take Fowler's official canon to
cide that you have caught the tenor of the argument;
mean
then you put in your oar. Someone answers; you an- something like all the authors and titles in
swer him. . . However, the discussion is intermina-whatever reasonably comprehensive literary
ble. The hour grows late, you must depart. And youhistories are standard at a given time and if we ac-
do depart, with the discussion vigorously in progress.cept his definition of the critical canon as the texts
(94-96) most written about at that time, the list of works

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
 Wendell V Hamris 113

commonly taught in high school and undergradu- Further perspective comes from recognizing
ate classes will be not only much shorter than the that, until the Renaissance, selective canons in lit-
official canon but also unlikely to correspond ex- erature were generally of little importance, that
actly to the critical. Thus there is theoretical space selective canons of European vernacular literature
for a pedagogical canon (canon8). blossomed only in the eighteenth century (see Cur-
What of the numerous authors who are given tius 264-72), and that selective canons of English
special recognition in selection after selection over and American literature are more recent still. The
centuries or at least decades? Or those contem- only nonecclesiastical canons that have carried
even local authority have been the lists of required
porary authors who have high visibility? In a hap-
hazard way these tend to be grouped together-in readings for specific educational endeavors. In the
the 1990s a person interested in literature presum-
medieval period, given the aleatory availability of
ably knows not only Ovid, Milton, and Arnold but Greek and Latin texts and the practice of study-
Ozick, Morrison, and Ashbery. But the glacially ing the ancient poets largely as part of the process
changing core is a kind of diachronic canonof learning the classical languages and acquiring
(canon9), to be distinguished from a rapidly rhetorical skills, the literary works prescribed vary
changing periphery that could be called the nonce considerably. Knowledge of literature was in any
canon (canon,0), only a minuscule part of whichcase wholly ancillary for centuries, having little
will eventually become part of the diachronic directly to do with the requirements for the
canon.5 I leave open the question whether the bachelor-of-arts degree, which from the twelfth
academy's hegemony is such that there are no century to the sixteenth were almost wholly in
longer truly popular canons rivaling the diachron- logic.7
ic and nonce canons of the literary professoriat.6In contrast, the breadth of learning that a
What makes it easy to think of a continuing Renaissance humanist like Erasmus expected of
monolithic canon, to confuse the diachronic and teachers was less a canon than a summons to
something of the nonce canon with the biblical universal knowledge: at one point or another, De
analogue, is the smoothness with which certain Ratione Studii mentions Pliny, Macrobius, Gel-
new writers enter the diachronic canon and certain lius, Athenaeus, Plato, Aristotle, Theophrastus,
texts and authors move within it from the accepted Plotinus, Origen, Chrysostom, Basil, Jerome,
center to the doubtful periphery (and occasionally Homer, Hesiod, Ovid, Boccaccio, Pomponius
back again). That smoothness is born of an inter- Mela, Ptolemy, Strabo, Lucian, Demosthenes,
woven set of processes. What a generation is Herodotus, Cicero, Quintilian, Horace, Caesar,
taught depends on the tastes and interests of the Sallust, Aristophanes, Euripides, Terence, and
previous generation and on the anthologies and Donatus. The formalization of curricula after the
texts created in response to the demands that is- Renaissance, of course, tended to produce a selec-
sue from those tastes and interests. To the selec- tive canon in each university. At least in the United
tion that it has inherited, each generation addsStates and England, the prescribed literary and hu-
those works given visibility by either fortunate manistic texts remained wholly classical until at
sponsorship or malleability to current interests. least the latter half of the nineteenth century. Thus
What one generation transmits to the next, how-the Dartmouth "Course of Study" for 1852-53
ever, can hardly be the sum of these two-there is lists the following more or less literary readings
only so much time in any degree curriculum, onlyover the four years: Livy, Homer's Iliad,
so much time for anyone to read. Something has Coleridge's Introductions to the Greek Classic
to give. Still, as has frequently been noticed, Poets, Ovid, Horace, Felton's Selections from the
authors once a part of the diachronic canon gener-Greek Historians, Aeschylus's Prometheus, Taci-
ally retain at least a minimal cachet; they may betus, Sophocles's Ajax, Demosthenes, Cicero,
relegated to a canonical attic but rarely to the trashPlato's Gorgias, Juvenal, and Terence's Andria.
can. Perhaps, therefore, the diachronic canon isVernacular literature did not enter the university
actually divided into two subcanons, a canonicalcurriculum until the nineteenth century. English
haven and a canonical limbo. became a school at Oxford only in 1893, a tripos

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
114 Canonicity

at Cambridge only in 1917, and a major at most omits about seventy-five of Stedman's poets, while
American universities still later; thus academic adding a few American writers. Christopher
canons of English literature developed at about the Ricks's New OxfordBook of Victorian Verse, pub-
same time as those of American literature and
lished seventy-five years after Quiller-Couch's,
have gone through parallel twentieth-century
offers only 113 poets. Granted that neither the edi-
revisions. tors' intentions nor the publishing exigencies can
Comparison of anthologies-a popular aca- have been quite the same for these volumes, a two-
demic pastime at present-is informative in several
thirds reduction in the number of poets is nonethe-
ways. Alan C. Golding's "History of American
less a dramatic winnowing.
Poetry Anthologies" surveys the range of criteriaEditors of classroom texts have found the pres-
used in anthologies from Elihu Hubbard Smith'ssures for reduction of the number of authors even
1793 American Poems, Selected and Original to greater. Assigning three hundred, two hundred, or
the 1975 Norton Anthology of American Litera-even fifty poets to an undergraduate class would
ture. As Golding points out, an anthologist like
be a daunting and probably an unfruitful under-
Smith, gathering poems from periodicals and taking. Thus a true textbook like George Benja-
newspapers,8 seeks to preserve, to increase the
min Woods's Victorian Poetry (1930) offers only
Fowlerian accessible canon. But what an editor sixty-nine poets for instructors to choose from.
deems worthy of preservation is already a selec- That number was cut to fifty-four in the 1955 edi-
tion. The criteria Golding finds in the earlier tion
an- of Victorian Poetry and to forty-seven for the
thologies are the promotion of political values, the
1965 edition. (The three editions do not greatly dif-
celebration of an American sense of nationhood, fer in the number of lines of poetry they contain;
and the provision of moral inspiration; after theadditional poems by the poets who have been re-
middle of the nineteenth century the primary con-tained take up the space.) Bowyer and Brooks's
sideration is the maintenance of a formal and con-Victorian Age: Prose, Poetry, and Drama (1954)
servative tradition; this is followed by the urgeoffers
to poems by forty-nine writers; Bloom and
undermine genteel values, and today the most im- Trilling's Victorian Prose and Poetry (1973) by just
portant challengers are feminist, ethnic, and po-twenty-one (including only one poet, John David-
litical (primarily Marxist) concerns. son, who does not appear in Bowyer and Brooks).
A glance at Victorian poetry collections is par-
Among the forces driving textbook anthologists
to restrict their selections is what can be called "the
ticularly useful, since the selection processes are
recent. E. C. Stedman's Victorian Anthology
principle of academic recirculation." Academics
(1895), apparently intended for the general reader
tend to teach what they have been taught, what is
as well as the student, offers poems by 329 poets,
easily available in print, what others are writing
including George Darley, Barry Cornwall (Bryan
interestingly about, and what they themselves are
Procter), C. J. Wells, William Maginn, William
writing about; what is easily available in print
James Linton, Sara Coleridge, Mary Howitt, Eliza
tends to be what is being taught and written about;
Cook, Roden Noel, Cosmo Monkhouse, Dinah
what is written about tends to be what one is teach-
Craik, Gerald Griffin, Robert Gilfillan, and Eu-
ing or others are writing about.
gene Lee-Hamilton. Given the volume's date and
anticipated readership, Stedman was altogether
Selective Canons: Criteria and Functions
reasonable in casting a broad net "to make a truth-
ful exhibit of the course of song during the last
Sorting the criteria used in drawing up selective
sixty years, as shown by the poets of Great Brit-
canons requires as much attention as sorting the
definitions of the term canon itself. The criteria
ain in the best of their shorter productions" (ix).
Who could know which poets of the preceding also tend to overlap, and it is difficult to imagine
sixty years would maintain readers' interest? a selection truly being made on one alone. Any
Seventeen years later another anthology not in-
editor of a collection titled "Writing by American
tended specifically for the classroom, Arthur Women, 1990" would clearly have to apply criteria
Quiller-Couch's Oxford Book of Victorian Verse,
beyond those stated in the title. Moreover, un-

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
 Wendell V. Harris 115

recognized assumptions underlie both explicit question by looking for hallmarks rather than
criteria and unacknowledged intentions. Thus the functions; perhaps he intuitively knew that if he
New Critical argument that poems cannot be looked for functions he would find all too many
paraphrased was developed into the claim that po- for his purposes.
etry has no propositional meaning; the ultimate The personal and literary-historical estimates
implication of that position is the futility of criti- are but two among countless others. While an ex-
cal discussion. Similarly, as R. S. Crane points haustive list of the functions of selective canons
out, the tension valued in New Criticism all too would probably not be either possible or useful,
easily passed over into the automatic ascription of it is easy to suggest a sufficient range of examples
universal oppositions to any text the critic valued.9 to caution against reductive generalizations.
In explicitly seeking "the best that is known and Providing Models, Ideals, and Inspiration.
thought," Arnold intended either to impose the Though the furnishing of examples is one of the
uniform moral and social values of his own class oldest functions of selection, there are evidently
(a common interpretation today) or (as it is pos- models of many different kinds. The Alexandrians
sible to argue) to set in motion a constant processchose texts demonstrating the best grammatical
of revaluation, behind which lay assumptions usage, while the Ciceronian and Quintilian con-
grounded in a belief in the necessity of a hierar- cepts of the orator-leader required texts embody-
chical society (now the usual view) or (as it is also ing various social virtues. Models of belief and
possible to argue) the necessity of breaking conduct are of course constantly shifting. What
through the class structure. 0 serves as a model of morality for most readers in
To some extent one can avoid the problem of one period may at another time be regarded as a
differentiating specific criteria and basic assump- model of self-righteousness; one person's inspira-
tions in a critic's or theorist's work by investigating tional clarion is another's intolerable cant. Gol-
the functions a particular selection was apparently ding describes Rufus Griswold's Poets and Poetry
intended to perform. Barbara Herrnstein Smith of America as based on "the conviction that
argues that all evaluations of literary texts are ac- American poetry should be represented by speci-
tually judgments of how well the texts in question mens of the utmost moral purity, that poetry's
satisfy the changing needs of individuals and so- function is inspirational" (288), but the challenge
cieties, that is, how well they fulfill particular func- to the poetry that Griswold assumed filled that
tions. To analyze the criteria on which a selection function-and the rather similar poetry antholo-
appears to have been made, critics must seek suchgized by Bryant, Emerson, and Whittier-came
functions, always keeping in mind that they recog-primarily from Whitman, whose champions could
nize these through processes that reflect their own equally describe him as moral and inspirational.
changing needs. Moralizing and inspiring are currently rather out
Arnold having been mentioned, it is appropri- of fashion as honorific adjectives for literature,
ate to begin with his influential effort to dis-but the functions they designate are still fully oper-
entangle the evaluative criteria-the kinds of ative. Marxist and feminist arguments are no less
"estimates," in his terms-on which selections are appeals to assumed moral values than are Pope's,
made. The two against which Arnold warns canWordsworth's, or Holmes's; the working-class
easily be defined by their functions. The "per-writers sponsored by Paul Lauter in "Caste, Class,
sonal" estimate attempts to achieve congruenceand Canon" offer no less a commentary on how
with individual needs and experiences (an aim that the world should be than do the writers honored
presumably represents the evaluative component by the New Critics.
of Fowler's personal canon). The "historical" es- Transmitting the Heritage of Thought. Another
timate, by which Arnold means the literary- canonical function is the provision of what is
historical, seeks to provide signposts to mark forksregarded as the basic cultural knowledge necessary
and turns in the historical development of genres, to interpret past texts, see current issues in histor-
new subject matter, and formal features. In urg- ical perspective, and orient oneself to the aesthetic
ing the "real" estimate, however, Arnold begs theachievements, social and political changes, and

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
116 Canonicity

philosophical debates that have gone on for cen- enough in the 1802 Preface to the Lyrica
turies. The Philology of Martianus Capella's fifth- Arnold was creating a place for his ow
century Marriage of Philology and Mercury is the even in condemning his "Empedocles o
bride who brings this ability to Mercury (elo- Writing about the Victorian canon, G.
quence). At one level the goal is "cultural literacy" Stange states that
in the specific sense of the ability to read texts writ-
ten by writers who have assumed such literacy-
as most serious writers have. 1 Harry Levin's more the principal agents of canon formation are
themselves who alter the poetic tradition by
ambitious description of knowledge as "our most
ing some accepted "classics," giving authority
valued patrimony, our collective memory," ex-
tain earlier art that has special meaning for
presses the use of such knowledge.
redefining in the interests of their own pra
the nature and responsibilities of poetic l
Higher education, across the centuries, has con- (159)
stituted a continuous dialogue between the minds of
ancestors and contemporaries. If we, the latter, know
any more than the former, it is because we have Alan Golding reminds us that "Bryant, Emerson,
learned so much from them. As T. S. Eliot remarked, and Whittier all agree that the six most important
"They are that which we know." Naturally we may re- poets in America are themselves, Longfellow,
act against them, and the reactions would not prove Lowell, and Holmes" (292). Hugh Kenner com-
unproductive if they pointed towards a dialectical ments, "The Modernist canon has been made
synthesis. (362)
. .chiefly . . . by the canonized themselves,
who were apt to be aware of a collective enterprise,
In fact, what appear to be
and repeatedly efforts
acknowledged one another" to(374); ove
present canon are and, of course, Eliot's
often championship of Donne
endeavors to ex
enlarge our patrimony and
was no less strategic enrich the
than sincere.
memory," that is, communal
Legitimating Theory. The New Critics knowl
provide
awareness. a dramatic instance of the influence of critical the-
Creating Common Frames of Reference. It is
ory on selection: while the function of their expli-
possible to argue not that any particular
cation wascanon is
presumably to exhibit meaning as fully
justified but rather that some canon as is necessary
possible, their selection of works to be expli-
to provide common reference points. If had
cated it the
is true
implicit function of exhibiting the
that all interpretation of texts depends
poweron a com-
of their approach. Deconstructionists, it
munity's sharing interpretive strategies,
hardlyit may
needs to be
be said, prefer texts with almost
equally true that, as Howard Felperin argues,
invisible seams"the
that can be pried open to suggest
institutional study of [literature] isgaping
inconceivable
contradictions, while neo-Marxists, includ-
without a canon. Without a canon, a ing
corpus
most ofor
thecy-
new historicists, are partial to texts
nosure of exemplary texts, there that
cancanbebe no in-
shown to reveal unsuspected workings
terpretive community, no more than of there
politicalcan be Practiced New Critics, decon-
power.
a faith-community without a gospel" (46). This
structionists, and Marxists can, of course, read al-
conception of the function of the canon
most anydoes
text not
in a way that supports their own
directly offer criteria for text selection, but itbut
allegiances, tends
the texts each group is most likely
to favor a limited selection from the
todiachronic
select are those for which it can provide the
canon.
fullest, most dramatic, and most convincing
Logrolling. Writers have gained entrance
readings. into
the nonce canon not only by the power of their
Historicizing. Literary texts have been so tradi-
writing (for "power" one may read tionally
"appeal to ex-
thought to cast light on the periods in
tant societal or critical interests") but by
which theytheir ac-
were written-and historical and con-
tive espousal of texts or criteria congenial
temporary to their
events to affect the proper interpre-
own aims. Wordsworth does this tation
transparently
of texts-that arguments over the relation

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
 Wendell V. Harms 117

of literature to history have primarily involved of literature into conventional 'periods' and
questions of emphasis. How fully are Chaucer's 'themes"' (440).
pilgrims representative of fourteenth-century En- At present, pluralization appears to have real,
gland? How reductive, how much confined to a if unstated, limits. For instance, there has been no
single class, is the "world picture" Tillyard found rush to defend the sentimental description and in-
in the Elizabethans? While one of the accepted spirational storytelling that delighted our grand-
values of literature, one reason for selecting older parents. The generation educated early in this
works, has been to convey a sense of how the world century still happily quoted "Little Orphant
was then, recent "historicizing" has shifted this Annie," "Excelsior," "Curfew Must Not Ring
emphasis to an analysis of the unconscious as- Tonight," "Casabianca," and "The Good Time
sumptions of earlier writers (as revealed by the Coming," but the antielitist impulse has yet to re-
conscious psychological or political assumptions habilitate Mrs. Hemans or Charles Mackay.
of the critic)12 or-as Annette Kolodny writes, cit-
ing Jane Tomkins-to an analysis of "'how and The Selection of Texts as the Selection
why specific texts have power in the world' (or do of Readings
not attain power, as the case may be) at any given
moment" (304). Whatever the functions governing selections, it is
Pluralizing. Though the attention given to lit- important to recognize that although a canon is
erature written by, or representing the experience nominally made up of texts, it is actually made up
of, women and ethnic minorities may seem espe- not of texts in themselves but of texts as read.
cially characteristic of the 1970s and 1980s, there When the church found ways of accommodating
was a better balance at the turn of the century than pagan authors to Christian belief, what it admit-
now. The 1890s in England were a strongly ted to its selection of Greek and Roman philoso-
pluralizing time in which writers expressing the phers and poets were particular readings of the
points of view of the Irish, the Scots, women, and texts. Augustine thus compares the acceptable
the poor were selected precisely because they teachings of the pagans to the "vases and orna-
represented perspectives outside the dominant ments of gold and silver and clothing which the
one. Stedman's 1895 anthology includes, for in- Israelites took with them secretly when they fled,
stance, sections for poets of Australasia and as if to put them to a better use" (75).
Canada and, among those who would now be Leaping to the present, one can cite The Catcher
called poets of the people, Thomas Cooper, in the Rye, which owes its continuing place in the
Ebenezer Elliott, and Ebenezer Jones. A full third nonce canon to its mimetic function as a portrait
of the poets included are women. By contrast, in of adolescence but which can be selected,
Quiller-Couch's 1912 Oxford Book of Victorian according to Richard Ohmann's reading, as a neo-
Verse only a sixth of the poets are women (a frac- Marxist text that intends to reveal the omni-
tion maintained in Christopher Ricks's 1987 New presence of capitalist ideology. Again, as Annette
Oxford Book of Victorian Verse), Irish poets are Kolodny argues more clearly and explicitly than
less frequent, and Indian and Australian poets most other critics seeking to expand the canon, the
have largely disappeared. In "Race and Gender in reading of unfamiliar literary texts and unfamiliar
the Shaping of the American Literary Canon" types of criticism should not merely make one
Paul Lauter notes a similar phenomenon in an- comfortable with new texts but defamiliarize the
thologies of American literature, with the percen- texts in the current critical and pedagogical
tages of black and women writers declining canons. While her already influential 1985 essay
markedly between 1919 and 1950. Though Lauter argues for expansion and pluralization of the offi-
elsewhere sees the invisible hand of monopoly cial canon, it implies as well that future selections
capitalism in canon selection, he here lists the should include only those works from the present
causes as "the eastern male elite's professionaliza- critical and pedagogical canons whose texts as
tion of the teaching of literature," the formalist read after defamiliarization meet whatever selec-
aesthetic, and the "historiographic organization tion criteria are applied.

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
118 Canonicity

The Ultimate Function of Canons Is history only as a tawdr

to Compete pursued and ideologie
Though ideologically or
If we have not one canon of literature but many, cite the relativity of "tru
no canon formation but, rather, constant pro- assume that their own
are, if not absolute, a g
cesses of text selection, no selection based on a sin-
gle criterion, and no escape from the necessity others.
of In contrast, thos
selection, to attack The Canon is to misconceive sistent indeterminizing
the problem. Similarly, to attribute all selectionthere are no absolute tr
processes to the influence of power is radically discover them if they do
ferred for its presumed
simplistic, unless power and influence are defined
tion is perennially ope
so broadly that they include all social motivation.
The dominant conventions in a particular societymust accept as absolute
and values or fall into a
at a particular time obviously derive their power
from some source. But the possible sources are dox (all beliefs are relativ
many-political, economic, moral, aesthetic, all beliefs are relative),
metaphysical, religious, and psychological-and beginning no later than
since they appear to be closely intertwined, thetion by Socrates, who
question of which, if any, underlies all the others
promulgating, truth.14
remains moot. One can argue that all human the "philosophical" (as
choices are at root political, or economic, orcal") tradition of liber
moral, or aesthetic, or metaphysical, or psycho-cisely on this argument,
logical, but little is illuminated by this tactic. 3standard selection of
Whatever the motivation of canon selection, it ground that there can b
remains important to contest the constrictive ten- can be no final truth. N
dencies of the critical and pedagogical lists. Givenever relativist, must ch
the forces generating the professional recirculationto talk about, just as r
ones to read.
of texts, we risk intellectual stagnation if we do not
champion new selections based on new criteria. ItAt the practical level, there will always be com-
is something else, however, to read as an ideologi-
peting canons: it is impossible to avoid the ques-
cal censor. Much contemporary commentarytion on of which texts one wishes to share or discuss
the diachronic canon seems intended to discredit in one's anthology, or critical article, or syllabus,
any text that arguably supports authority, elitism, or polemic. Recent textbook anthologies have fat-
or capitalism. But simply to emphasize the "elitist" tened noticeably in their editors' attempts to rep-
and "capitalist" assumptions of older texts seems resent greater cultural diversity, but the length of
designed more to invoke magical exorcism than to semesters has unfortunately remained the same.
do anything else. As Murray Krieger writes: As teachers of literature we have thus again had
to become more consciously selective. This is all
to the good to the extent that it makes us recog-
To reject our revered masterpieces, then, is really to nize the clarification of a literary text's functions
reject the political institutions at work in the cultures
as the necessary prolegomenon to the process of
that produced them. It is as if, by turning against an
aesthetic monument . . . the anti-elitist critics some-
selection. Critics continue to agree with Bacon, if
one can judge from what they write, that two of
how can wish out of existence the reactionary politi-
cal context that may have been thriving when the work the primary activities of criticism are providing
was created. (155) "brief censure and judgment of authors; that men
may make some election unto themselves what
books to read" and setting an order of studies so
Charles Altieri similarly questions
"that men may know in what order "the
or pursuitherm
to
tics of suspicion": "It
read"is
(86, a
182);mistake to isread
the requisite qualification that cu

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
 Wendell V Harris 119

"judgment and censure" and the preferable order of "the books publicly read in Christian assemblies," or one
of reading depend on criteria that depend on that was canonical in the sense of "excellent as a model," or
one that embodied a regulative principle.
purposes.
2Kimball's Orators and Philosophers is a masterly survey
None of the functions I have outlined is either
of the tensions and later confusions between the "oratorical"
nefarious or trivial. It is well to have some knowl-
(artes liberales) tradition that sought to imbue leaders of so-
edge of major literary-historical influences on ciety with the proper virtues and the "philosophical" tradi-
texts and some familiarity with the sources of the tion underlain by a skepticism that required an unending
search for ever-elusive truths. Isocrates, Cicero, and Quintilian
literary and philosophical allusions authors have
are key figures in the first tradition; Plato, Socrates, and Aqui-
expected their educated readers to share. It is well nas in the second.
to bring some historical perspective to contem- 3By way of comparison, the sixth of the thirty-nine Arti-
porary debates: to see the scandals of television cles of Religion of the Anglican Church reads, "Holy Scrip-
evangelism against the background of the ture containeth all things necessary to salvation: so that
Pardoner's Tale, contemporary theological dis- whatsoever is not read therein, nor may be proved thereby,
is not to be required of any man, that it should be believed
putes against Pope's Essay on Man, questions of
as an article of the Faith, or be thought requisite or neces-
the limits of individual freedom against Mill's On sary to salvation. In the name of Holy Scripture we do un-
Liberty, and arguments over corporate responsi- derstand those canonical Books of the Old and New
bility for the environment against Ruskin's essays Testament, of whose authority was never any doubt in t
Church."
on political economy. It is well to recognize that
4I am using significance to contrast with intentional mean-
much of our literature assumes male, Anglo-
ing, following the distinction developed by E. D. Hirsch in
Saxon, competitively individualistic biases. It is
Validity in Interpretation.
well to encounter models of prose that persua- 5Older works that had not become part of, or had dropped
sively use traditionally effective rhetorical devices out of, the diachronic canon may of course enter it belatedly
and well to encounter individualistic, counter- if they are fortunate in their sponsorship-Eliot's sponsor-
cultural prose. But no selection of texts that can ship of Donne is the standard example-and sufficiently
malleable to be linked to current cultural and critical interests.
be fitted into the one literature course, or perhaps
6There seems at least a short-term popular canon in the
the two or three, taken by the average under- area of fiction and nonfiction: Harriet Doerr's Stones for
graduate-or into the dozen or so literature Ibarra, Tony Hillerman's Navajo mysteries, Gerald Durrell's
courses taken by the baccalaureate major-can autobiographical stories of animal collecting, and Cleveland
adequately provide all that background. We need Amory's tale The Cat That Came for Christmas are perhaps
more than ever, then, to be honest with ourselves examples of more or less recent books that, though appropri-
ately advertised, came to be known primarily through the
and with our students about the limited purposes recommendations of one reader to another.
both of individual courses and of the requirements 7As stated in 1252 by "the masters of the English nation,
for our degrees-to be honest about what our teaching in Paris," the requirements for the bachelor-of-arts
selection of texts and our approach to them does degree were attendance in "lectures in arts for five years or
not accomplish. If The Canon no longer lives, the four at least at Paris continually or elsewhere in a university
of arts," at which the student should have heard lectures on
reason is that it never did; there have been and are
a large portion of Aristotle-the Praedicamenta, the
only selections with purposes. If anything has been Periarmeniae, the Topics, the Divisions, the Elenci, the Prior
clarified by the last twenty years of critical alarms Analytics, the Posterior Analytics, and De anima-together
and excursions, it is the multiplicity of possible with the Sex Principia, the eighteen books of Priscian's gram-
purposes. mar, and the Barbarismus of Donatus (Thorndike 53-54).
8It is worth noting that E. C. Stedman similarly drew on
ephemeral publications in compiling his Victorian Anthol-
ogy a hundred years later.
9Gerald Graff's chapter "What Was New Criticism" in his
Literature against Itself succinctly sums up the contradictions
within the New Critical program. Examples of the "all-
Notes
embracing dichotomies" Crane cites are "good and evil, love
and hate, harmony and strife, order and disorder, eternity and
'Under the heading "canon" in the eleventh edition of thetime, reality and appearance." Crane comments, "Of such
Encyclopaedia Britannica, Samuel Davidson summarizes threeuniversal contraries, not restricted in their applicability to any
opinions: in ecclesiastical use canon originally meant a list
kind of work, whether lyric, narrative, or dramatic, it will be

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
120 Canonicity

easy enough for us to acquire an adequate supply, and once Works Cited
we have them . . . it will seldom be hard to discover their
Altieri, Charles. "An Idea and Ideal of a Literary Canon."
presence in poems as organizing principles of symbolic con-
tent" (123-24). Critical Inquiry 10 (1983): 37-60.
Augustine, Bishop of Hippo. On Christian Doctrine. Trans.
10For a presentation of the second set of views of Arnold,
see Harris. D. W. Robertson, Jr. Indianapolis: Bobbs, 1958.
Bacon, Francis. The Advancement of Learning. Ed. W. A.
1 lIt is instructive to note how frequently arguments against
the possibility of a shared culture nevertheless rely on one. Wright. Oxford: Clarendon-Oxford UP, 1900.
Bloom, Harold, and Lionel Trilling. Victorian Prose and
Clifford Geertz, for instance, argues that the "enormous mul-
tiplicity" and the "radical variousness of the way we think Poetry. New York: Oxford UP, 1973.
now" make the transmission of a core of shared cultural Bowyer, John Wilson, and John Lee Brooks. The Victorian
knowledge impossible (161). What gives his article depth,
Age: Prose, Poetry, and Drama. New York: Appleton,
1954.
however, is the multiplicity of his own allusions to the pre-
sumed knowledge of the educated reader. He appears toBurke,
ex- Kenneth. The Philosophy of Literary Form. 1941. New
pect appropriate responses, for instance, when he mentions
York: Vintage, 1957.
Copernicus, Freud, Bach, Ptolemy, Einstein, Malinowski, Crane, R. S. The Language of Criticism and the Structure
Boas, Whorf, the Tewas language, Bertrand Russell, Kant, of Poetry. Toronto: U of Toronto P, 1953.
Berkeley's Esse est percipi, the Pyrenees, William James, Curtius, Ernst Robert. European Literature and the Latin
Henry James, Yeats, C. P. Snow, mandarins, and Eden. Middle Ages. Trans. Willard R. Trask. Bollingen Series
12For instance, Jonathan Dollimore writes in the foreword 36. Princeton: Princeton UP, 1953.
to Political Shakespeare: "Finally, cultural materialism does
Dollimore, Jonathan, and Alan Sinfield, eds. Political
not pretend to political neutrality. . . . Cultural material- Shakespeare: New Essays in Cultural Materialism. Man-
ism does not, like much established literary criticism, attempt chester, Eng.: Manchester UP, 1985.
to mystify its perspective as the natural, obvious, or right in-
Erasmus, Desiderius. De Ratione Studii. Concerning the Aim
terpretation of an allegedly given textual fact. On the con- and Method of Education. Ed. and trans. William Har-
trary, it registers its commitment to the transformation of arison Woodward. New York: Teachers Coll. P, 1964.
Felperin, Howard. Beyond Deconstruction: The Uses and
social order which exploits people on grounds of race, gender
and class" (viii). Abuses of Literary Theory. Oxford: Clarendon-Oxford
13In one sense, whatever exists, including all cultural be- UP, 1985.
liefs, conventions, artifacts, and conditions, can be described
Fowler, Alastair. "Genre and the Literary Canon." New Liter-
as demonstrating the triumph of one force or power over ary History 11 (1979): 97-119.
another: the power of tradition, education, religion, politi-
Geertz, Clifford. "The Way We Think Now: Toward an Eth-
cal structures, science, logic, capitalism, socialism, selfishness,nography of Modern Thought." Local Knowledge: Fur-
hatred, ignorance, benevolence, self-interest, advertising,ther Essays in Interpretive Anthropology. New York:
propaganda, personal experience, the press, the constitution
Basic, 1983. 147-63.
of the human mind and body, the knowledge of the brevity
Golding, Alan C. "A History of American Poetry Antholo-
of life, the need for love and approval-the list is endless.
gies." Canons. Ed. Robert von Hallberg. Chicago: U of
All the powers or forces influencing human decisions inter- Chicago P, 1984. 279-307.
act to produce the total social structure of any given moment.
Graff, Gerald. Literature against Itself Chicago: U of Chicago
The canonical status of a literary text-like the economic sta-
P, 1979.
tus of a rock musician, the reputation of a painter, the pu- Wendell V. "The Continuously Creative Function of
Harris,
rity of the air and water, the desirability of consumer goods,Arnoldian Criticism." Victorian Poetry 26 (Spring-
or the majority positions on taxes, abortion, and nuclearSummer 1988): 117-33.
power-can only be understood as the result of multiple
Hirsch, E. D., Jr. Validity in Interpretation. New Haven: Yale
causes. To attribute any cultural phenomenon to a single UP, 1967.
"power"-that of capitalism, or male prejudice, or political
Kenner, Hugh. "The Making of the Modernist Canon."
corruption, or economic greed, or moral idealism-is as naiveCanons. Ed. Robert von Hallberg. Chicago: U of
as to think such powers can be ignored. Chicago P, 1984. 363-75.
14See Bruce Kimball's tracing of this "philosophical" orien-
Kermode, Frank. "Institutional Control of Interpretation."
tation from Plato to the present. The second part of the Salmagundi 43 (1979): 72-86.
seventh chapter of Walter Pater's Plato and PlatonismKimball,
em- Bruce. Orators and Philosophers: A History of the
phasizes (and perhaps magnifies) this aspect of Plato's Idea of Liberal Education. New York: Teachers Coll. P,
thought, carefully outlining the "dialectic method, this con-
1986.
tinuous discourse with one's self" (185), and the spirit "which
Kolodny, Annette. "The Integrity of Memory: Creating a New
to the last will have its diffidence and reserve, its scruples and
Literary History of the United States." American Liter-
second thoughts" (196). ature 57 (1985): 291-307.

This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
 Wendell V Harms
Krieger, Murray. Words about Words about Words: Theory,
Criticism, and the Literary Text. Baltimore: Johns Hop-
kins UP, 1988.
Lauter, Paul. "Caste, Class, and Canon." A Gift of Tongues.
Ed. Marie Harris and Kathleen Aguerro. Athens: U of
Georgia P, 1987. 51-82.
. "Race and Gender in the Shaping of the American
Literary Canon: A Case Study from the Twenties." Fem-
inist Studies 9 (1983): 435-65.
Levin, Harry. "Core, Canon, Curriculum." College English 43
(1981): 352-62.
Ohmann, Richard, with Carol Ohmann. "A Case Study in
Canon Formation: Reviewers, Critics, and The Catcher
in the Rye." Politics or Letters. By Richard Ohmann.
Middletown: Wesleyan UP, 1987. 45-67.
Pater, Walter. Plato and Platonism. London: Macmillan, 1910.
Pfeiffer, Rudolf. History of Classical Scholarship from the
This content downloaded from 12.156.20.2 on Fri, 13 Dec 2019 20:18:52 UTC
All use subject to https://about.jstor.org/terms
121
Beginning to the End of the Hellenistic Age. Oxford:
Clarendon-Oxford UP, 1968.
Quiller-Couch, Arthur, ed. The Oxford Book of Victorian
Verse. Oxford: Clarendon-Oxford UP, 1912.
Ricks, Christopher, ed. The New Oxford Book of Victorian
Verse. Oxford: Oxford UP, 1987.
Smith, Barbara Herrnstein. "Contingencies of Value." Criti-
cal Inquiry 10 (1983): 1-35.
Stange, G. Robert. "1887 and the Making of the Victorian
Canon." Victorian Studies 25 (1987): 151-68.
Stedman, E. C. A Victorian Anthology, 1837-1895. Boston:
Houghton, 1895.
Thorndike, Lynn. University Records and Life in the Middle
Ages. New York: Columbia UP, 1944.
Woods, George Benjamin, ed. Poetry of the Victorian Period.
Chicago: Scott, 1930. 2nd and 3rd eds. with Jerome
Hamilton Buckley. 1955, 1965.