Allegory

Society for the Study of Early China
IMAGES OF ALLEGORY: A REVIEW ARTICLE

Author(s): CHARLES HARTMAN
Source: Early China, Vol. 14 (1989), pp. 183-200
Published by: Society for the Study of Early China
Stable URL: http://www.jstor.org/stable/23351416 .
Accessed: 07/09/2013 20:46
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
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 [email protected].
Society for the Study of Early China is collaborating with JSTOR to digitize, preserve and extend access to
Early China.
http://www.jstor.org
This content downloaded from 142.3.100.23 on Sat, 7 Sep 2013 20:46:14 PM

All use subject to JSTOR Terms and Conditions
IMAGES OF ALLEGORY: A REVIEW ARTICLE
CHARLES HARTMAN
Pauline Yu's new book, The Reading of Imagery in the Chinese Poetic
Tradition (Princeton: Princeton University Press, 1987), is the
culmination of a decade of research into the problems of imagery in
Chinese poetry. Her earlier well-received articles on metaphor and
allegory formed
intermediate stages in this research, and are
incorporated into this volume.1 Among her stated goals is a desire to
depart from the practice of discussing Chinese poetry using Western
rhetorical terms. "In particular, I have tried to delineate the cultural
presuppositions that I believe distinguish attitudes toward poetic

imagery in classical China from those commonly taken for granted in
the West. These dissimilarities argue against the all-too-frequent
unqualified adoption of Western rhetorical terms to apply to a body of
writing rooted in a very different set of assumptions..." (p. ix). What

follows this welcome announcement is a major book by a major scholar
on a major topic in the field. It is a book that every serious student of

Chinese poetry will read, and that every serious student of traditional
China should read, because its assumptions, its difficulties, and its
conclusions touch many of the central issues concerning the modern
study of intellectual life in ancient China.

In her earlier articles, Yu maintained that the Western terms
metaphor and allegory do not properly describe Chinese phenomena,

and this view is repeated and elaborated here. Yet, this book, the first
sentence of which reads "Imagery has been a central concern of Chinese
poetics from its very beginnings..." (p. 3), begins with a survey of the
term "imagery," in which she determines that the current meaning
1. Pauline Yu, "Metaphor and Chinese Poetry," CLEAR 3.2 (July 1981), 205-224, and
"Allegory, Allegoresis, and the Classic of Poetry," HJAS 43.2 (Dec. 1983), 377-412. I
would like to express my appreciation to Professor Edward L. Shaughnessy and to
two anonymous reviewers for their assistance in revising this article for publication in
Early China.
Early China 14(1989)

184 IMAGES OF ALLEGORY
dates from Thomas Hobbes (pp. 3-10). The reader might well ask if
and "allegory" are culture-bound Western terms that
"metaphor"
cannot be applied to Chinese experience, then how is "imagery," clearly
as Western a term as the other two, any more applicable?2
Yu prefers to speak of imagery rather than metaphor and allegory
because her review of the Western understanding of the latter two
terms reveals that both involve a spanning of "two ontologically distinct

realms, one concrete and the other abstract" 17) through the author's
(p.
Active, verbal creation. Metaphor and allegory are thus a reflection of
the "dualism at the heart of Western philosophy." Since Chinese
cosmology was monistic (there was no other realm beyond the tao
immanent in all things), it is impossible to speak of metaphor and

in a Chinese context. Rather, order was manifested and
allegory
maintained of correspondences that linked the forces
through systems
and entities present. This is the well-known philosopy of "organicism"
that has been described by Marcel Granet, Joseph Needham, and others.
Thus according to Yu, for the Chinese writer accustomed to "correlative
thinking," the connections between entities, including his own

emotions and physical realities outside the self, are seen as "pre
established." "Instead of the mimetic view that poetry is the imitation of
an action... it is seen here as a literal reaction of the poet to the world...."
(p. 35).
Yu attempts to work around these difficuties and to validate her
choice of terminology by identifying "image" with the hsiang of the

I-ching (pp. 37-43). Although her review of the Western scholarship
on the I-ching includes references to the difficulty of this facile
translation, the equation of hsiang with the "imagery" of Western
literary theory is accepted for the remainder of the book. Yet,
acceptance of the cultural implications inherent in /iswng/image is

certainly as dangerous as the facile translation of pi tb as "metaphor"
and hsing §5 as "allegory." However one may interpret hsiang, this
Chinese concept is certainly as far from Thomas Hobbes's conception of
2. Yu's initial efforts to distinguish between Chinese "imagery" and Western
"metaphor" are not maintained throughout the book. For instance, in the concluding
pages she writes that in China "imagery was not ultimately important for what it
presented directly but for what it concealed and evoked in the reader.... Such a
method then came be defined as in fact the method of bi xing and the quintessential^
'poetic' one, a position remarkably similar to ideas about metaphor in the West" (p. 217,
emphasis added). And the book closes with the sentence: "For all of these reasons,
then, Chinese poetic imagery should be distinguished from the metaphors and
allegories of Western literature, whose fundamental premises are so very different" (p.
218).

CHARLES HARTMAN 185
imagery as Yu asserts the Mao commentary is from allegory. If using
"metaphor" and "allegory" to speak of Chinese poetry distorts that

poetry because of the "cosmological gulf" between the two cultures,
then "imagery" is no less dangerous a term simply because, as Yu
contends, it is the more "inclusive" of the three (p. 11). One might point
out, for instance, that in the West the poet creates "images" just as he
creates "metaphors." But in China the hsiang, like natural
correspondences, are fixed. As Willard Peterson writes, they "are
independent of any human observer; they are 'out there/ whether or

not we look."3
It may also be useful to emphasize that Pauline Yu's observations on

the relationship of the Shih-ching to Western allegory are valid only to
the extent that her definition of Western allegory is sound.
— an
Unfortunately, the definition with which she has chosen to work
extended metaphor, fictionally developed by the author at two different
levels of meaning in an effort to span his dualistic cosmos— is
extremely conservative. It fails to do justice either to the richness and
diversity of the Western allegorical tradition or to the recent significant

advances in allegorical theory that have arisen in the wake of
deconstruction. Her survey of Western allegorical cites in
theory
passing Northrop Frye's well-known remark that "all commentary is
allegorical interpretation, an attaching of ideas to the structure of poetic
imagery" (p. 27). Although she does not pursue this line of thought,
Frye's view and the critical that have ensued from it call
developments
out for application in the Chinese cultural context, so overladen as this
latter is with the notions and of and, well,
practice commentary
polysemous discourse.
Rather than arguing that the traditional Chinese world was
cosmologically predisposed against allegory, one could perhaps make a
more convincing case that the system of correspondences inherent in
the Chinese Weltanschauung actually predisposed the Chinese writer to
allegory. Granet observed almost sixty years ago and with great insight
that
when Chinese writers of metaphors or allegories we should

speak
take heed. These terms denote not so much a process used men
by
of letters as a system used by moralists. Imagery is not employed
merely to simplify the idea or make it more attractive: in itself it has
3. Willard Peterson, "Making Connections: 'Commentary on the Attached

Verbalizations' of the Book of Change," HJAS 42.1 (June 1982), 80. One notes with
interest that Yu argues against Peterson's preferred translation of hsiang as "figure"
rather than the more accepted "image."

a moral value. This is evident in the case of certain themes. For
example, the picture of birds flying in couples is, in itself, an

exhortation to fidelity. If, then, metaphors borrowed from Nature
are used to give expression to the emotions, it is due not so much to
a consciousness of the beauty of Nature as to the fact that it is

moral to conform to Nature. Where, at first, one may be tempted to
see an artistic intention, there may be a moral intention.4
Likewise, Ch'ien Chung-shu §1 Iff in his T'an-i lu |*£|§ iU, a book

which Yu quotes often, after a discussion of Chinese concepts related to
Western allegory, emphasized how different Chinese and Western
cosmological assumptions are mirrored in the differing construction of
allegorical figures.
In Western literature, the mode called allegory is roughly similar to
these concepts. The Greek stoics had already begun the later trend
toward allegoresis. Dante, basing himself on contemporary

methods of biblical typological interpretation, expanded these to
art in a way that was similar to the devices of Tzu-hsia -pM. and
Wang I But there is a difference. In China, physical objects
were used as figures for human events: the relationship between
man and woman was used as a figure for the sentiments between
ruler and minister. Both figure and figured are real. But Dante used
human events as figures for religious truth: the relationship
between man and woman was used as a figure for the relations
between God and man. The figure was real, but the figured was
unreal. In the first case, poetry becomes history; in the second,
poetry becomes mysticism.5
4. Marcel Granet, Festivals and Songs of Ancient China, trans. E.D. Edwards (London:
Routledge and Sons, 1932), 50. Despite the cosmological gulf, a similar type of "pre
established" equivalence between natural imagery and moral value also occurred in
the traditional West. See, for instance, D.W. Robertson, A Preface to Chaucer (Princeton:
Princeton University Press, 1962), 388-390; Angus Fletcher, Allegory: The Theory of a
Symbolic Mode (Ithaca and London: Cornell University Press), 113ff, where reference is
made to the well-known book of E.M.W. Tillyard, The Elizabethan World Picture. For a
specific example, see R.E. Kaske, "The Summoner's Garleek, Onyons, and eek Lekes,"
Modern Language Notes 74 (1959), 481-484. See also the excellent summary "Correlative
Thought in Traditional Western Civilizations" in John Henderson, The Development

and Decline of Chinese Cosmology (New York: Columbia University Press, 1984), 54-58.
5. Ch'ien Chung-shu, T'an-i lu (Shanghai: K'ai-ming, 1948), 275 [rev. ed. Peking:
Chung-hua, 1984, p. 231]. Yu refers (p. 199, n.83) to the same page of Ch'ien's work but
does not mention this passage, even though its relevance to her argument seems
obvious. It is also striking that she makes no reference to Andrew Plaks's extensive
discussion of ontological dualism and allegory. Although his discussions center on

CHARLES HARTMAN 187
In other words, the fusion of natural image and moral meaning,

which in his dualistic universe the Western allegorist must labor to
establish, is given or "pre-established" for the Chinese author, whose

main difficulty is to divorce image from meaning, when he wishes to
use an established image with an other than established meaning. The
result, as Yu argues, may not be Western allegory, but it is certainly
polysemy.
This brings us to the core of Yu's argument, her contention that the
Chinese conception of a monistic universe precludes the possibility of
allegory and even of metaphor, as these terms have been understood in
the West. As we have seen above, her conservative definition of

Western allegory does much to vitiate this observation. Yet even
accepting her original formulation, I have serious doubts as to the
validity of the premise; for, although no one contests the ultimate

oneness of the traditional Chinese world, it is likewise well known that
from earliest times the Chinese have shown a clear preference for
thinking in dualistic terms. In short, although Chinese cosmology is

monistic, Chinese epistemology is decidedly dualistic. And Pauline Yu
has conflated the two.
This conflation and Yu's refusal to use the term allegory lead to
another of the central difficulties of the book: a pronounced

condescension toward the goals and concerns of traditional Chinese
exegesis. This is evident in her advocacy of the term "contextualization"
as a replacement for the usual term "allegorization" to characterize the

traditional exegesis of the Shih-ching. Contextualization seems to me
wrong on at least two counts. First, in relation to the Shih-ching itself, it

implies that the commentators have superimposed contexts on a group
of poems that were originally without them. Even though this

assumption may be partially true, it is certainly an oversimplification of
the complicated and drawn-out process by which an early version of the
Chinese which largely coincide with Yu's, are certainly also

fiction, his conclusions,
valid for Chinese
poetry. See Plaks, Archctype and Allegory in the Dream of the Red
Chamber (Princeton: Princeton University Press, 1976), 7, 108-110. There are also other
significant bibliographical omissions. Yu's discussions of the critical terms fu j@t, pi
tE, and hsing $11 make no reference to the researches of Chow Tse-tsung jc| £it/
who has traced the relationship of all "six principles of poetry" (liu i 7\S£) to early
shamanism. See "Ku-wu tui yueh-wu chi shih-ko fa-chan ti kung-hsien" ]*j
Ch'itw-hua hsiieh-pao NS. 13 (Dec. 1982), 1-25,
and Ku wu-i yii liu-shih k'ao ZE is |5| "a g$" ^ (Taipei: Lien-ching, 1986). Also
missing is reference to the important paper of Chou Shan, "Beginning with Images in
the Nature Poetry of Wang Wei," HJAS 42.1 (June 1982), 117-137, which contains much
revelant material.

text, the minor prefaces, and the Mao exegesis were fused into the
present Mao shih Cheng-chien In this sense, contextualization

does not accurately describe the commentators own perception of what
they were doing. Second, in relation to Chinese poetry in general, it

implies that there actually exist two groups of poems: those with
contexts and those without. I would contend that all texts have contexts,
even though these may have been lost over time and even though the
contexts as reconstructed by traditional commentators may seem
implausible or speculative to Western readers. The issue of meaning is
central here: no Chinese poem could exist in a sense that was
meaningful to a pre-modern Chinese reader without a context. If this

had been lost and the poem was considered culturally important
enough, a new context was supplied. And it is here that Frye's remark
that all commentary is allegory may begin to assume significance in the
Chinese setting. Whereas I could agree that the Shih-ching is not
allegorization, it certainly is polysemous in some sense— probably
closer to Western typology than to allegory— and contextualization,
although the term points toward the preference of Shih-ching

commentators for highly specific interpretations, avoids the central
issue of how they conceive of the texfs figurative structures as
contributing to its polysemy.
Pauline Yu's impatience with the implicit assumptions and formal

conventions of traditional Chinese commentary has seriously impaired
her study of the commentarial legacy of the Shih-ching and the Li sao
15- Her pages brim with sarcasm directed against commentators
individually and collectively. They "remain undaunted" (pp. 66, 89),
they "perform some exegetical contortions" (p. 75), they "continue along
avenues that should by now be quite predictable" (p. 125), and "they
would not agree with Karlgren..." (p. 68). This last remark is a telling
formulation. Although she has taken remarkable pains to place and
read Chinese poems in their cultural context, she is remarkably
disinclined to treat the commentaries in the same manner. And
although she has surmounted the methodological strictures that the
application of the New Criticism once
imposed on the modern
explication of Chinese
literary texts, she is still more than willing to
reject traditional Chinese criticism when this latter seems to her too
"far-fetched." Accordingly, the baseline for meaning is assumed to be
whatever Karlgren, Waley, Watson, or Hawkes have translated. She
repeatedly invokes the judgement of a hypothetical "modern reader" to

counter the perceived impossibility of traditional commentary (pp. 68,

CHARLES HARTMAN 189
75,92,127,132). The result is often a missed opportunity to illustrate the

real "commentarial legacy" of the texts under review.
Perhaps an example will illustrate this attitude. In her discussion of

the Shih-ching poem "Chiu yu" A IS (Mao #159), where the two
dominant images are the netting of rudd and bream and the flight of a
wild goose along an island, she recounts at length the differences
among the various commentators concerning the size of the net and the
nature of the fish. But the legacy of the commentaries to this poem is
not that they differ as to the details of the specific net, fish, and goose in
question, but that they helped to establish two powerful analogies that
became standard in later Chinese literature: the difficulty of netting
rudd and bream as a figure for the difficultyof recruiting the good man,
and the solitary flying habits of the wild goose as a figure for the
loftiness of the good man. Yu, of course, notes these analogies, but her
citation of Karlgren's translation, a practice standard throughout the
book, considerably muddles her discussion of the poem (pp. 65-67). She
shows no interest in relating these analogies to the poem's potential
meaning, which seems to remain for her the Karlgren translation. In this
case, however, Karlgren's translation is based on the researches of Wen
I-to |i} — i?, whose reading of this poem as a young woman's love
lament was part of his effort to establish the metaphorical relationship
between fish and sexual desire in the Shih-ching.6 Even though this
reading may reveal a valid aspect of the poem's early history, Wen I-to's
discovery does not negate the cultural power of traditional commentary

to mold the interpretation of this ode throughout the entire period of
pre-modern Chinese culture nor does it lessen the importance of this

commentary as the locus classicus for the two analogies mentioned
above.7
Despite its title, The Reading of Imagery in the Chinese Poetic Tradition is
not a book that attempts in a rigorous and objective way to examine the
assumptions and goals of traditional exegesis. Rather, Pauline Yu makes
use of what she perceives to be dissension within this tradition, or to

use her phrase, "disharmonious interpretations," to construct a number
of propositions that concern her views on how imagery works in

Chinese poetry. Much of her own exegesis of the Shih-ching and Li sao
6. Wen I-to, Shih-ching —■

t'ung-i1$ ® Us, Wen l-to ch'ilan-chiHIJ ^ ^ HI
(Shanghai: K'ai-ming, 1948), vol. II, 127.
7. For the consensus among traditional Shih-ching commentators that this poem
concerns the Duke of Chou's impending return to the West, see Ch'en Tzu-chan |?j!
"P M., Shih-ching chih-chich ill (Shanghai: Fu-tan ta-hsiieh ch'u-pan-she,
1983), vol. 1,503-505.

aims not to elucidate the texts but rather to elucidate disharmony
among the commentators. Much of this disharmony arises not from
anything in the commentaries themselves, but from Yu's impatience

with the tradition's own concerns and conventions. She is, for instance,
unconcerned to distinguish the various purposes or intended audiences
of different commentaries, often reading as disputes what are little more
than the same information directed to different audiences in different

periods. And, unconcerned as she is with questions of meaning, she is
usually derisive of the commentarial tradition's contributions in this
regard.
For example, distich six of the Li sao along with the translation of
Hawkes reads:
I dressed in selinea and shady angelica,

And twined autumn orchids to make a garland.8
Yu comments on these lines:
Thus Wang Yi 5^ (A- 110-120), the first major annotator of the

text, not surprisingly explains that this "garland" or girdle {pei
is "an image [xiang] for virtue," one of the earliest instances, to my
knowledge, where this term is taken from its context in the Classic
of Changes and applied to a literary text. Most commentators agree

with him, although Wang Yuan jEE (fl. 1600) argues that "the
ancients actually did use orchids to make garlands, so this is not
just a metaphor, and scholars should know this." This is only the
first of many disagreements between commentators on the extent
to which images should be taken literally (p. 89).
What Wang I says is this:
a m m,»tt. m,tfittmm it.

...Immm, nmmmz, mam t#
Jen [to string together] is to make a cord. Lan [orchid] is an aromatic

plant, fragrant in the autumn. Pei [girdle ornament] is to adorn; it is
that by which he gives form to his moral virtue. Therefore, the one
whose actions are pure and spotless wears fragrant plants as a
girdle ornament.... The meaning is that he has cultivated his [moral
8. David Hawkes, Ch'u Tz'u: The Songs of the South (London: Oxford University
Press, 1959), 22.

CHARLES HARTMAN 191
virtue to the point where it is] pure and spotless; and so he takes
selinea and angelica in order to wear them as clothing, and ties

autumn orchids in order to make a girdle ornament as adornment:
vastly he gathers all that is good in order to discipline and control

himself.9
What Wang Yuan says is this:
ay) b,Jtw,
Wang I says that "the meaning is that he has cultivated his [moral
virtue to the point where it is] pure and spotless; vastly he gathers
all that is good in order to discipline and control himself." This is
correct. Yet the "Internal Ordinances" [of the Li chi] state: "If
anyone should give a wife... an iris or an orchid, she should receive
and offer it to her parents-in-law." So the ancients indeed all once
used varieties of orchids and irises to make girdle ornaments; they
are not simply a metaphorical figure. This also is something
scholars should know.10
First, even if we were to accept the equation of hsiang and "image" as Yu

has suggested, hsiang is not a regular part of the commentarial technical
vocabulary relating to figurative expression. In brief, I do not believe

hsiang should be understood here as a technical literary term for
"imaging" or "figuring" (Wang I's usual term for this is yu [Ifi]);the
construction with so-i PJfJH,especially, argues against such a reading of
hsiang, which is here best understood in its normal, non-technical and
verbal meaning of "to give form or shape to." Second, it is obvious
when both commentators are translated in full that there is no
"disagreement... on the extent to which images should be taken
literally." Wang Yuan fully accepts Wang I's understanding of the
figurative value of the autumn orchids, and he supports, rather than as

Yu argues disagrees with, this interpretation by adding what might be
seen as an anthropological footnote: he cites a passage from the Li chi
9. Li sao tsuan i ed. Yu Kuo-en jS?®,® (Peking: Chung-hua, 1980), 31.

Cf. the similar interpretation and wording for distich 37 on p. 115.
10. Li sao tsuan i, 33. The U Chi reference can be found in Li chi cheng i ffflP.TFijf
(Shih-san ching chu-shu ed. [rpt. Peking: Chung-hua, 1980]), vol. II, 1463b; see also James
Legge, Li Chi: Book of Rites (rpt. New Hyde Park, N.V.: University Books, 1967), vol I,
458. Wang Yuan has abridged the original text, which reads in Legge's translation: "If
any one give the wife an article of food or dress, a piece of cloth or silk, a handkerchief
for her girdle, an iris or orchid, she should receive and offer it to her parents-in-law."

which he believes proves the ancients actually did wear orchids.

Scholars should also know this. For Wang Yuan, the existence of the Li
chi passsage confirming for him the literal fact of ancient orchid wearing
strengthens Wang I's understanding of this facfs figurative
significance: this is not only a metaphor. The point is crucial, and we
should recall Wang Yiian's comment together with Granefs
observation quoted above: cosmologically and epistemologically, the
traditional Chinese did not draw a dichotomy between a thing or action

and its moral significance. As Granet said, "the picture of birds flying in
couples is, in itself, an exhortation to fidelity." The fact that evidence
can be produced to prove the literal existence of a thing or action
confirms its moral significance.

Another example of this tendency to truncate or excerpt the
comments of individual commentators in order to exacerbate perceived

disputes among them occurs in the discussion of Li sao distiches 93-95.
These lines mark an important transition in the poem as Ch'ii Yuan |@
jl* begins his metaphorical journey into the sky in search of the fair one.
I yoked a team of jade dragons to a phoenix-figured car

And waited for the wind to come, to soar up on my journey.
In the morning I started on my way from Ts'ang-wu;

In the evening I came to the Garden of Paradise.
I wanted to stay a while in those fairy precincts,

But the swift-moving sun was dipping in the west.11
Pauline Yu writes of these lines:
Wang Yi at first seems to take the account literally, making sure

that we know that these dragons, for example, are of the hornless
variety (qiu £L) rather than horned (long jf|). Somewhat later,
however, it becomes clear that he reads this section as another
yet
veiled description of an attempt to make contact with the ruler, for
he notes that the word "fairy" (ling H) in distich 95 refers obliquely
to the ruler; Hong Xingzu further explains that "the dwelling place
of the spirits is a comparison to the ruler." Other commentators
whose remarks are collected in the Li sao zuan yi engage in a lively

debate, either supporting Wang Yi's contention or arguing that the
mythological elements should be taken at face value, rather than as
11. Hawkes, Ch'u Tz'u, 28.

CHARLES HARTMAN 193
political images (pp. 255-56). Zhu Xi, however, displays the least
literal cast of mind, for he advises us that the entire account is in
fact fictional: "these words mostly simulate— these objects and
events do not really exist" (p. 94).
First, Wang I's comment on the dragon horns is not gratuitous trivia.
Even less is it evidence that he "seems to take the account literally."
Wang I is doing here exactly what a good commentator should do. He is

indicating to his reader what the word ch'iu means: it is a young dragon
that has not yet grown his horns. It is important for the symbolism of
the poem that the reader understand that these ch'iu are young dragons
and not some other type of creature. Wang I does this in the language
and style of his time, and it is simply wrong to read more (or less) into
his comments.12
Second, Wang I does not shift his opinion on the figurative nature of
the passage. His comment on distich 93 closes with the remark that the
purpose of Ch'ii Yiian's journey is "to depart the vulgar of this world
and distance himself from the hoards of the small." Nor is there a "lively
debate" over whether the imagery is mythological or figurative, a
distinction I doubt would have made sense to the commentators in
question. There are a series of opinions on the meaning of the difficult

term ling so f§ and on the location of the "Garden of Paradise," but
these amount to little more than the routine philological give-and-take
so usual in classical, especially Li-sao, Most commentators
commentary.
favor some degree of figurative interpretation, and there is again the
usual give-and-take about how specific this interpretation should be.
The translation of Chu Hsi's comment ittJsAT,
1^1 Je ® and especially of the technical term chia-t'o flg It as
"simulate" fails to bring out the full meaning of the term. Although
chia-t'o can certainly mean "to simulate" (see Mathews), the word is part
of the technical vocabulary of figurative expression and refers to the
"borrowing" of meaning that takes place in metaphor, with emphasis
12. The text of the Shuo-wen chieli-tzu a work contemporaneous with
Wang I, defines
the ch'iu as "a young dragon with horns" (lung tzu yu chiao-che jt|-?
iei), but there is considerable evidence from later citations of this text that the
correct reading should be "a dragon without horns" (lung wu chiao<he ^),
and this latter reading was accepted by Tuan Yii-ts'ai IgtHi Wt, Shuo-wen chieh-tzu chu
ft (rpt. Taipei: Shih-chieh, 1961), 13A.14b. This was the tradition for later
commentators. See, for instance, Ssu-ma Hsiang-ju WJ "Shang-lin fu" h
IS in Wen hsiian (rpt. Shanghai: Ku-chi, 1986), vol. 1,371 where Li Shan's
gloss on ch'iu reads "lung yell, wu chiao yiieh ch'iu yeh" fitfc/ H 4L tfc/ which
succinctly combines the two points of Wang I: ch'iu are dragons; they have not yet
grown horns.

on the conjured and/or invented nature of the ultimate target meaning

of the passage. This literary meaning is related to the modern, non
literary meaning of the word as "pretext." The author describes one
thing as a pretext for something else. Chu Hsi is not emphasizing the
"fictional" aspect of the text, but rather its figurative meaning: do not
think that what occurs in the text from this point on (i.e. mounting to
the sky on dragons, etc.) actually happened, these are only figures for
something else. Or, as August Pfizmaier correctly remarked in a note,
probably based on Chu Hsi, to his 1852 translation of the text: "Was nun
folgt bis zu dem Schlusse des Gedichtes, ist durchaus lauter Pi-yii, eine
Vereinigung von Allegorie und Fabel, und die Handlungen diirften nur
als Ideen betrachtet werden."13
Yu's observations on literary texts are interspersed with a history of
the understanding and use of the terms pi and hsing (pp. 57-65,159-167,
181-187, 211-216). Two technical problems mar these otherwise useful
discussions: 1) a lack of attention to detail in the translation of what are

admittedly some very difficult texts,14 and 2) a tendency to remove these
passages from their surrounding contexts (either in the texts themselves
13. August Pfizmaier, "Das Li-sao und die neun Gesange," Denkschriften der
kaiserlichen Akademie der Wissenschaften. Philosophisch-historische Classe (Vienna), 3
(1852), 167, n.5.
14. For instance, Yu's treatment (pp. 182-187) of the material on Chiao-jan
and Wang Ch'ang-ling ~F rI from the Bunkyo hifuron JxC Iw has suffered
from her failure to consult the 1983 edition of Wang Li-ch'i 3E^[J§S, Wen-clung mi-fix
lun chiao-chu fff Iw (Peking: Chung-kuo she-hui k'o-hsiieh ch'u-pan
she) or the important study by Wang Meng-ou ^ IS, "Wang Ch'ang-ling sheng
p'ing chi ch'i shih-lun" -r r jj? S -B- StF Bffl(rpt. in Wang Meng-ou, Ku-tien
wen-hstieh lun t'an-so t*j [Taipei: Cheng-chung, 1984], 259-294). The
latter in particular demonstrates through a comparison of parallel passages that the
definitions of the "Six Principles of Poetry" Tn fg attributed to Chiao-jan and Wang
Ch'ang-ling in the Bunkyo hifuron (Wang ed., pp. 158-162) are not textually secure. As a
result, evidence drawn from these texts to support Yu's conclusion concerning a
"fundamental confusion" between pi and hsing is highly tenuous. Wang Meng-ou has
also shown (pp. 273,292n.29) that the texts of two passages Yu has translated from the
Shih ko of Wang Ch'ang-ling are probably Five Dynasties reworkings of earlier
material, and that the Bunkyo hifuron parallel passages are to be preferred. Strangely,
she has not used the secure and very interesting
textually material in Wang's
"Seventeen (Wang ed., pp. 114ff.), for which there is a preliminary study
Dispositions"
in English by Joseph J. Lee, Wang Ch'ang-ling (Boston: Twayne, 1982), 67-82. The
seventh of these "dispositions" entitled "enigmatic comparison" (mi-pi ^ tb)> which
is illustrated by one of Wang Ch'ang-ling's own poems and his own commentary,
shows him in the process of doing something very close to making his own
metaphors, using as always the stock images, but investing them with new, personal
values (Wang ed., p. 124-125; Lee, pp. 81-82).

CHARLES HARTMAN 195
or in the general opinions of their authors) in a way that sometimes

distorts their meaning.15
A major contention is that the history of the terms pi and hsing is both
confused and confusing; and most scholars who have worked on these
texts would probably concur. She consistently translates hsing as
"stimulus" and pi as "comparison," following established usage and

the controversial terms allegory and metaphor. But during her
avoiding
introduction of the earliest and most critical definitions of these terms
(pp. 58-59) she herself introduces considerable confusion by translating

three other terms all as "comparison." Thus she translates Cheng
Chung's definition of pi (pi-che, pi fang yti wu tb#, tbTij as

"a comparison makes a comparison to an object," ignoring the meaning
of fang as "likeness, similitude." One might suggest something like "pi is
to compare (lit. set side-by-side) likenesses with respect to things," to
bring out Cheng Chung7s idea that a pi is a comparison based on some
similarity between two objects. On the same page she also translates
both p'i § and yti (Mr)

as "comparison." The latter is especially confusing,
since term for figuration,
yti is perhaps the most widespread technical
and since this translation is maintained throughout the book (pp. 66,74,
78, 183). It leads her at once into a major error when she declares that
15. For example, Yu's translation and interpretation of Ssu-ma Kuang's WJ ,^§ 7^
comment on Tu Fu's fi iff "Spring Gaze" (pp. 197-198) exactly reverses the intent of
the passage, which is not talking about imagery but about meaning. The passage
opens with a quotation from the Shih-ching poem "T'iao chih hua" g i|I (Mao
#233) combined with a fragment of the Mao-Cheng commentary to that poem, which
Yu has omitted from her translation:
hmss. #a&h, n k ma
sto mzm&mt. jgntSAtitt^s
mmxzM.
The ewes have
big heads; the Three Stars are seen in the fish trap. [Mao
commentary:] 'It cannot last for long.' When the ancients composed poetry they
put value on meaning that was beyond the words, so that people could obtain
that meaning only after considerable thought. Therefore, those who spoke were
without fault; and those who heard were sufficiently admonished. In recent
times, only Tu Fu has perfected this style of the Shih-ching poets....
Ssu-ma Kuang is not simply making a perfunctory bow toward the classics; he is
making a direct connection between the didactic function of the Shih-ching poets and
Tu Fu. "It cannot last for long" is Mao's reading of the opening couplet to mean that
the Chou state will soon fall (see Bernhard Karglren, "Glosses on the Siao Ya Odes,"
BMFEA 16 [1944], 165-166). Ssu-ma Kuang is reading Tu Fu the same way he read the
Shih-ching. It may well be that, as Yu argues, use of imagery such as one encounters in
"Spring Gaze" is not present in Chinese poetry until the T'ang; but that is clearly not
Ssu-ma Kuang's view.

1% IMAGES OF ALLEGORY
the T'ang commentator K'ung Ying-ta ?L H M "confuse[s] the two

functions [of pi and hsing] even more thoroughly." But K'ung Ying-ta is
not writing about pi and hsing, he is writing about yti and hsing. By
equating yti and pi, she has herself confused the functions even more
thoroughly. Thus, Cheng Hsuan's routine formula hsing-che yti
is invariably translated "this stimulus is a comparison to the fact that...,"
which must surely leave non-sinological readers wondering why the
Chinese were unable to distinguish two of their most basic critical
terms.
As an of the second tendency, we may turn to Yu's

example
discussion of a well-known passage in which the Ming scholar Li
Tung-yang (1447-1516) advocates the superiority of pi and hsing
over fu !®, because the former two "consist of relying on objects in order
to lodge one's When words come to an end but meaning is

feelings....
endless, then the spirit will be refreshed and move as if in flight, hands
will dance and feet will stomp without one's being aware of it. This is
the reason why poetry values emotional thought and disesteems real
events" p. 213). Yu comments:
Although Li Dongyang's mention of dancing hands and stomping

feet recalls the orthodox Confucian Great Preface to the Classic of
Poetry, the rest of this passage explicitly denies the main premise of
the Preface as well as the shorter prefaces to individual poems in
the anthology: that the songs allude to concrete historical events or
situations during the Zhou dynasty.... Li Dongyang rejects both the

direct mode of presentation of exposition and the possibility of
topical allusion as limiting the emotionally evocative powers of
poetry.
Li Tung-yang was presented to the emperor at age three, passed the
chin-shih at age sixteen, held continuous official position for fiftyyears,

and was the preeminent literatus of the capital for much of that time.16It
seems hardly likely that he would allude to the Preface in one line and
deny its major premise in the next. "Real events" does not refer to the
specific topical references of Shih-ching exegesis; rather, the contrast

with "emotional thought" simply restates the commonplace that poetry
best concerns itself not with the specific, concrete details of an affair but
with the emotions behind it. Far from denying the premises of the
Preface, Li Tung-yang is reaffirming them by incorporating the Ch'an
16. See Daniel Bryant's entry on Li Tung-yang in The Indiana Companion to

Traditional Chinese Literature, ed. William J. Nienhauser, Jr., et al. (Bloomington:
Indiana University Press, 1986), 554-555.

CHARLES HARTMAN 197
based "poetics of transcendence," which he inherited from Yen Yii's H

33 Ts'ang-lang $hih-hua M t$ IS and which he promoted, into the
orthodox understanding of the Preface: the poems of the Shih-ching do
not narrate the details of specific Chou events (shih-shih H Jt); they
speak of them in terms of the "lodged" (yu "0H)"feelings and thoughts"
(ch'ing-ssu fjf Jgjt)of the authors. In short, Li Tung-yang projects the
"poetics of transcendence" back to the authors of the Shih-ching and
thus reasserts the collection's value for his own time.
To illustrate the implications of these two basic problems that I
perceive in Yu's work— the insistence on elevating "imagery" over
metaphor and allegory and the reluctance to accept the Chinese
commentarial tradition on its own terms— it will be helpful in

conclusion to turn briefly to her discussion of later poetry.
There are many fine observations on individual poems and poets in
the fourth and fifthchapters of the book, which form a distinct unit as
opposed to the earlier sections on the Shih-ching and Li sao. Specifically,
her insistence that the legacy of Mao-Cheng and Wang I left Chinese
writers and readers with a tendency to think of poems as responses to
real situations and her corresponding reservations toward general and
universal readings of the "Nineteen Ancient Poems" and Juan Chi PicIf
are well-conceived and well-argued. Yet, a primary feature of this book,
its propensity to formulate literary issues as bifurcated into opposing
polarities (thus the imagery of the Shih-ching is juxtapositional, that of
the Ch'u-tz'u is substitutive; in the Six Dynasties, imagery is literal or
emblematic), although perhaps helping to bring clarity to a sometimes
confusing mass of material, also tends to mask the often considerable
degree of interaction between what are presented as opposing

tendencies. Although Yu usually admits the possibility of such
interaction in her theoretical sections, she seldom explores these
possibilities in her exegesis of actual poems.

For example, her dichotomy of oblique vs. direct referentiality in
T'ang poetry obscures the degree to which the old emblematic values of
the images retained their power in the hands of skillful poets working
in the new tradition. In short, where Yu perceives a new development
toward the "poetics of transcendence" that eclipsed the older poetic
tradition, I would rather conceive a dramatic expansion of the old
tradition as it assimilated new developments from within (the landscape
tradition of T'ao Ch'ien |Sj and Hsieh Ling-yiin ||} 1§ jig) and
absorbed new elements from without (Buddhist epistemology). This
latter view is probably closer to the Chinese poetic tradition's view of
itself. Yu's evidence for any meaningful renunciation of the emblematic

legacy of Mao-Cheng/Wang I within the Chinese tradition is open to

questions of interpretation of the sort I have indicated above with
reference to her citation of Li Tung-yang.
These difficulties are nowhere more apparent than in her discussion
on pp. 191-198 of several famous Li Po and Tu Fu poems, all of which

are presented as examples of "a plausibly observed scene," in which the
poet "preserves the primacy of the literal object as observed." This
vacillation as to whether we are dealing with a plausible scene or a
literal scene is one source of confusion throughout this section. The
discussion begins with Li Po's "Seeing Off a Friend" ("Sung yu-jen" 3^

M A)/ the middle couplets of which are, as Yu translates them:
tkM-mimmmmfc.
From this place once parting has ended,

The lone tumbleweed flies a myriad miles.
Floating clouds: a traveler's thoughts.
Setting sun: an old friend's feelings.

She insists that all of these — and sun —
images tumbleweed, clouds,
form part of a plausibly observed scene, and this is certainly so. Yet,
each brings to that scene connotations from established prior usage that
Li Po uses in different ways. The tumbleweed as a figure for the lonely

traveler is standard, and Li Po fully accepts that equation here without
modification. The floating clouds, however, are more complicated. They
certainly do have connotations from the "Nineteen Ancient Poems" as
images for slanderers, but this image is hardly appropriate here. But
there are other associations of "floating clouds," including an example
from the Wen hstian that is remarkably similar to Li Po's use here. The
first of Li Ling's ^ three poems for Su Wu has the following
middle couplets:
\mnmt
Looking up, we see floating clouds rush by,

quickly passing each over the other.
In waves of wind, they soon lose their place,
and each comes to rest in a far corner of the sky.17
17. Wen hsiiatt (rpt. Shanghai: Ku-chi, 1986), vol. Ill, 1352. Cf. trans, in Erwin von
Zach, Die chinesische Anthologie (Cambridge: Harvard Univ. Press, 1958), vol. 1,520, and
Arthur Waley, Chinese Poems (London: Allen and Unwin, 1946), 44.

CHARLES HARTMAN 199
Li Shan notes that these couplets "figure human travel" (i yii jen chih
k'e-yu iiiifii So Yu's contention that "the clouds and sun are
not effaced from the scene in favor of another situation" is not entirely
accurate. The clouds at least, like the tumbleweed, clearly come into the
poem with something more than their value as plausible elements in a
visual scene. In short, the whole concept of "image" in Chinese poetry is
more complicated than what first meets the eye. These "floating clouds"
are actually many "images" depending on the context from which they
come and on the text into which they go.
The attempt to explicate this set of poems as literal scenic descriptions
fails completely in the discussion of Tu Fu's "Yangtse and Han"
("Chiang Han" Yu sees in this poem "knowledge conveyed not
through abstract statement... but in the coherently progressing images
of a natural scene." Yet the old horse of the last couplet, "Since ancient
times they've kept a place for the old horses/ no need for them to take
the long road" is certainly not part of any
natural scene. He is rather one of the most standard of images with
"pre-established meaning," according to most commentators deriving

from Han Fei tzu And he is the culminating figure in the poem.
Likewise, in the third couplet, "In setting sun, heart still hale;/ in
autumn wind, my illness getting better" ($$ 0 <L>30 {fir, ® 8ft.!£),
although it is true that Tu Fu reverses the conventional associations of
setting sun and autumn wind, the tension and beauty of the line arises
from the very fact that these images already have such associations.18 Tu
Fu is simply making use of them in a different way. And certainly not all
critics agreed that both images were to be understood as part of an
hypothetical "scenic tableau." In fact, the Ch'ing critic Huang Sheng ft

felt compelled to answer unnamed earlier critics who castigated Tu
Fu for mentioning both sun and moon in the same poem (thus creating
what was obviously felt to be an impossible scenic tableau) by pointing
out that "the setting sun is simply a figure for one's closing years." In
other words, as far as Huang Sheng is concerned, the sun is not even
literally present in the poem. The line simply means "in my closing
years I still feel hale."
I have dwelt on these distinctions because I believe Pauline Yu has
overemphasized the degree to which the T'ang poets rejected the
principles upon which the old poetics of the classics were based. First,
there is no reason an image that is part of a legitimate "plausible scene"
cannot also have fixed associations. Li Po's "floating clouds" is a good
18. See Kao Yu-kung and Mei Tsu-lin, "Meaning, Metaphor, and Allusion in T'ang
Poetry," HJAS 38.2 (Dec. 1978), 290-291.

example. This is especially true in the Chinese world, where the very
landscape itself was felt to have moral power and significance. One
should also recall that Chinese poets were often disposed to construct
real-life artificial "scenic tableaux" that were nevertheless also symbolic,

so that the finished poems "describing" these scenes would be both
literal and figurative at the same time. There are many examples from
the miniature gardens of the T'ang to the grape arbor in the Chin P'ing
Mei Moreover, drawing too strong a distinction between these
two kinds of images ignores the cultural context in which Chinese
poetry was learned and practiced. Scholars began to learn poetry by

memorizing the pre-established associations of standard images; this is
clear from the existence of T'ang pedagogical texts like the Erh-nan mi
chih Hf.19But more accomplished poets quickly learned to go
beyond these associations or to adapt them to their own artistic uses, as
Tu Fu does in "Yangtse and Han." Thus the identical "image" with
"pre-established meaning" could have different meanings depending
on the skill of the poet.
I hope the above remarks will be taken as an indication of the
importance I perceive in The Reading of Imagery in the Chinese Poetic

Tradition. It is to Pauline Yu's immense credit that she has undertaken to
plot some basic correlations over so long a time period between the
classics of Chinese poetics and the poems themselves. Her ambitious
formulations of many issues central to our understanding of Chinese
literary history will long remain focal points for other scholars working
in the field.
19. Yu's identification of the Erh-nan mi-chih with

exclusively emblematic poetiy
that inserts into the narrative "images that are not plausible elements of an empirically
observed scene" (p. 185) is vitiated by the three poems that conclude the text and
whose explication demonstrates the author's own view of how the analogies he has
listed apply to contemporary The poems, by Huang-fu Jan Je 11} -p}-, Li Chia
poetry.
yu and Li Tuan are all typical late High Tang style, all veiy much in
the "transcendence" school.
Once should also note that the Erh-nan mi-chih is by no
means an isolated of its genre. Wang Meng-ou
example ("Wang Ch'ang-ling sheng
p'ing," 322-323) has pointed out that this style of poetry criticism emphasizing specific
one-to-one correspondences between image and signification was popular in the
Ch'an monasteries of the late Tang and Five Dynasties periods, exactly concurrent
with the full flowering of Yu's "poetics of transcendence." Clearly, these issues await
further research.


Allegory

Uploaded by

Copyright:

Available Formats

Allegory

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Allegory

Uploaded by

Copyright:

Available Formats

Society for the Study of Early China

IMAGES OF ALLEGORY: A REVIEW ARTICLE

This content downloaded from 142.3.100.23 on Sat, 7 Sep 2013 20:46:14 PM

Chinese poetry. Her earlier well-received articles on metaphor and

presuppositions that I believe distinguish attitudes toward poetic

writing rooted in a very different set of assumptions..." (p. ix). What

on a major topic in the field. It is a book that every serious student of

study of intellectual life in ancient China.

metaphor and allegory do not properly describe Chinese phenomena,

Early China 14(1989)

This content downloaded from 142.3.100.23 on Sat, 7 Sep 2013 20:46:14 PM

terms reveals that both involve a spanning of "two ontologically distinct

immanent in all things), it is impossible to speak of metaphor and

Thus according to Yu, for the Chinese writer accustomed to "correlative

thinking," the connections between entities, including his own

choice of terminology by identifying "image" with the hsiang of the

literary theory is accepted for the remainder of the book. Yet,

acceptance of the cultural implications inherent in /iswng/image is

2. Yu's initial efforts to distinguish between Chinese "imagery" and Western

This content downloaded from 142.3.100.23 on Sat, 7 Sep 2013 20:46:14 PM

imagery as Yu asserts the Mao commentary is from allegory. If using

"metaphor" and "allegory" to speak of Chinese poetry distorts that

independent of any human observer; they are 'out there/ whether or

It may also be useful to emphasize that Pauline Yu's observations on

diversity of the Western allegorical tradition or to the recent significant

when Chinese writers of metaphors or allegories we should

3. Willard Peterson, "Making Connections: 'Commentary on the Attached

This content downloaded from 142.3.100.23 on Sat, 7 Sep 2013 20:46:14 PM

a moral value. This is evident in the case of certain themes. For

example, the picture of birds flying in couples is, in itself, an

a consciousness of the beauty of Nature as to the fact that it is

see an artistic intention, there may be a moral intention.4

Likewise, Ch'ien Chung-shu §1 Iff in his T'an-i lu |*£|§ iU, a book

cosmological assumptions are mirrored in the differing construction of

toward allegoresis. Dante, basing himself on contemporary

unreal. In the first case, poetry becomes history; in the second,

poetry becomes mysticism.5

Thought in Traditional Western Civilizations" in John Henderson, The Development

This content downloaded from 142.3.100.23 on Sat, 7 Sep 2013 20:46:14 PM

In other words, the fusion of natural image and moral meaning,

establish, is given or "pre-established" for the Chinese author, whose

the West. As we have seen above, her conservative definition of

accepting her original formulation, I have serious doubts as to the

validity of the premise; for, although no one contests the ultimate

thinking in dualistic terms. In short, although Chinese cosmology is

another of the central difficulties of the book: a pronounced

exegesis. This is evident in her advocacy of the term "contextualization"

as a replacement for the usual term "allegorization" to characterize the

wrong on at least two counts. First, in relation to the Shih-ching itself, it

of poems that were originally without them. Even though this

Chinese which largely coincide with Yu's, are certainly also

This content downloaded from 142.3.100.23 on Sat, 7 Sep 2013 20:46:14 PM

present Mao shih Cheng-chien In this sense, contextualization

they were doing. Second, in relation to Chinese poetry in general, it

contexts as reconstructed by traditional commentators may seem

implausible or speculative to Western readers. The issue of meaning is

central here: no Chinese poem could exist in a sense that was

meaningful to a pre-modern Chinese reader without a context. If this

although the term points toward the preference of Shih-ching