Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives
This review appeared in Volume 3 (3) of The Semiotic Review of Books.
Occidental Poetics: Tradition and Progress. By Lubomir Dolezel. Lincoln and London: University of Nebraska Press, 1990. x + 261 pp. ISBN 0-80321685-8.
The aim of this book is to reconstruct the history of occidental poetics, which, according to Dolezel, is practically unknown (vii). In order to travel this course he distinguishes criticism from poetics, the former being a value-assigning activity spelling out arguable criteria for aesthetic evaluation (32) while the latter is cognitive in the sense of formulating the epistemology of a scientific study of poetic art (32) and gathering and providing increasing knowledge about literature (vii; 8). Dolezel presents his history according to what appears to him a natural divide: the first part of his book deals with pre-twentieth century stages during which the issues and problems of poetics were formulated. The second part of the book concentrates on twentieth-century structural trends where the above-mentioned issues receive their systematization. The cut off date for twentieth century poetics is 1945. Thus post-war poetics are not covered since they would have required a book of much greater length. The focus of his history is not on "unit ideas" or a search for predecessors, but on a history of a science of literature (2) which has evolving sets of assumptions, methods, and concepts which can be systematically reconstructed (4). The poetics Dolezel has in mind hinges on two interlinked assumptions: 1. Literature is a distinct, specific, creative productive activity. 2. Poetics is a systematic scientific study of literature. To trace the history of poetics involves the recognition that certain issues appear as successive reformulations and refinements of basic themata. The four which he singles out for his discussion are the idea of literature as structure; the doctrine of mimesis; the problems of poetic creativity; and the relation between literature and language.
I would map the discussion in the two parts of the book along three intersecting axes of development. 1. Chronological-geographical: from pre-twentieth-century Greece, Germany, and England to twentieth-century France, Switzerland, Russia and Prague. The basic division is the aforementioned formulation and systematization stages. 2. Thematic. In part one the four themata, the "invariants," as it were, of poetics, are foregrounded. In part two the focus is on different "schools" and their contribution to modern structural poetics. 3. Modelling, an axis that cuts across both parts of the book. The development of the mereological (concerned with the relationship of parts and wholes) model from its inception in Aristotle, through its morphological (organic) variant to its semiotic (twentieth-century) variant is traced. These intersecting axes make this book, despite its modest size, a very concentrated and substantial study.
Chapter one, dealing with Aristotle, lays the groundwork for all three axes of development and therefore deserves a detailed summary.
In discussing Aristotle Dolezel is not interested in dealing with often disputed topics such as mimesis or catharsis. Rather he focusses on the epistemological principles of the Poetics or "how Aristotle proceeded to gather knowledge about poetic art" (11) and the model Aristotle chose to represent this knowledge, which Dolezel carefully infers from Aristotle's implicit epistemology. Poetics is knowledge about the making of poetry and thus belongs within the productive sciences (13). This explains the dual aspect of poetics as both a theory of poetry (in the wide sense of verbal art) and as rational foundation for practice (14). As a specialized science, poetics is universalist in that it shares with other sciences the focus on general or universal rather than particular properties and in that it requires a method based on both induction and deduction and strives to organize knowledge on the principles of dimension. In the last requirement according ding to Dolezel, Aristotle seems to recognize different degrees of exactness (14-15). Dolezel reconstructs Aristotle's model of knowledge as a top-bottom model, i.e. the less abstract (lower level categories are derived from the more abstract or higher level ones. His model is thus stratificational comprising altogether four levels features of mimetic arts being the highest, through aspects of tragedy, parts of tragedy, (e.g. plot) and parts of parts (e.g. peripety) being the lowest. Dolezel also undertakes to clarify the different procedures applied in the unfolding of the model) and to show the dominance of one procedure (25h He uncovers three derivational procedures: the First is inferential (syllogistic) and covers levels one and two. On level three there is a shift to a mereological procedure- the focus now being on the relation of parts to whole- thus indicating tragedy is structurally modelled. Structure, in this sense, is based on the postulate of nonadditivity (the whole is more that the sum of its parts) and the postulate of nonadditivity (the whole is more than the sum of its parts) and the postulateof completeness, i.e. the enumeration of all possible constituent parts rendering the categorial sets exhaustive. The model is coherent but allows for no structural innovation (23; 179 n. 13), an idea that will be developed by Dolezel when he discusses normative poetics. The lowest level of the model is derived by a third, different, procedure, diaresis or division, where constituents are not decomposed into parts but specific properties of variants (e.g. of peripety) are given. By showing that the mereological procedure is central to Aristotle's Poetics Dolezel can claim that poetics at its inception is a scientific discipline defining its cognitive tasks, which will reappear in different and later poetics, as the study of structures, that is the study of emergent properties, of hierarchies finked by derivation and integration, of relationships between parts and parts and between parts and wholes (24). Aristotle's mereology draws on an analogy between poetic and Living structures and thus foreshadows the organic model of poetics which will dominate romantic poetics. Another aspect of Aristotle's poetics Dolezel highlights as important for later poetics is its exemplificatory "epistemic equipment" (25) which in a universalistic poetic works, though exemplification is used primarily with the lower levels of the model (e.g. plot constituents such as peripety exemplified by Oedipus Tyrannus).
Aristotle, as is commonly known, is also considered one of the founders of literary criticism. In the last part of his discussion, which I will not detail, Dolsel examines Aristotle in the light of his (Dolezel's) distinction between poetics and criticism, which cannot always be separated from each other in the Poetics. Ultimately, according to Dolezel, both poetics and criticism in Aristotle are founded on an intuitive axiology. Though, as Dolezel points out, Aristotle's theory is not an a priori construct, his poetics is filtered through a notion of an ideal structure (i.e., a few works selected into a privileged set, which are the basis and purpose of the theory thus necessarily including a restrictive, normative component (whose most evident mandestation will be neo-classicist poetics, which Dolezel mentions briefly (34)). Nevertheless Aristotle, in minor cryptic passages of the Poetics, opens the way for a descriptive poetics both in his fourfold typology of tragedy, which seems to include the notion of shimming dominants, and in the mention of the relationship between Fictional plots and historical events, two aspects which undermine the normative poetics of an ideal type.
While chapter one foregrounds the idea of literature as structure (the first of the themata), chapter two, which deals with Leibnitz and other poeticians of the German tradition, concentrates on the link between two themes, mimesis and creativity, both hinging on the relations of poetic art and the world. Dolezel briefly traces the shift in meaning of mimesis in normative poetics up to Leibnitz, showing that in the German tradition, but not exclusively there, the idea of mimesis as an imitation of nature (based on the norm of resemblance which began as a universal criterion, ended up as a view subject one since nature" is defined in vague terms and the poetician studies nature only through poetic representations and thus has no methodology for its cognition (35-36). Having established the problem, Dolezel concentrates on an emergent poetics, in the aesthetics of Baumgarten and Bëitinger, which did not dismantle mimetic poetics but liberalized its postulates (49). Both, while taking into account subjective factors, required a rationalistic theory of poetry and concentrated on logic and models of artistic creativity opening the way to a non-mimetic possible-worlds poetics. In this view poetic art is not subordinate to nature but parallel to it The artist is a maker who transforms possibles into fictional existents.
The important point in terms of the history of poetics is that the eighteenth-century German tradition offers a semiotics of imaginary worlds, whose most significant aspect is that poetic language is assigned a specific evocative function (47) distinguishing it from ordinary language. The study of the special evocative devotes specific to this language is considered the prime task of poetics(47-48).
In chapter three Dolezel focuses less on the thematic axis than on the model one, though the issue of creativity is discussed. He argues against the view that romantic poetics closes the development of poetics. Within romantic aesthetics Dolezel singles out Goethe's morphological model of nature as the basis for and an important stage from, as well as a reformulation of issues systematized in twentieth-century poetics. Specifically, the morphological model of poetics based on the theory of organic structures is a transition stage from, as well as a reformulation of, the Aristotelian mereological model Morphology is a 'theory of the formulation of complex structures from individual parts (56), which is why it is also a mereology. The contribution of romantic mereology (morphological model) is that poetics as a science of literature integrates the universal categories of Aristotelian poetics (genres) with an analysis and interpretation of specific works (56).
I will not detail Dolezel's discussion and developments of Goethe's two-level morphological model except to mention that Goethe's five postulates about structures, the first two of which are identical with aristotelian completeness and non-additivity, his Ur-type and metamorphosis make for a model that is dynamic and can account for invadance, in terms of which diversity of phenomena can be differentiated. These principles allow for creativity but also circumscribe it by definite laws. Goethe's biological model was extended, according to Dolezel, by Humboldt to poetics. Humboldt's main contribution is in emphasizing the creative role of the imagination and in his "zigzag method" which switches back and forth "from universal categories to concrete descriptions" (67) thus bridging the gap between theory of literature and text analysis.
Despite Humboldt's contribution to a creative view of language as a production of signifying systems (75), Humboldt did not develop a linguistic poetics, i.e., "the study of the verbal aspects of literature" (76). To fill in this lacuna in the development of poetics at its formulative stage Dolezel, in chapter four, turns to the fourth of his themata by examining the view of language and poetic language as expounded by Wordsworth, Coleridge and Frege, thus providing a gamut of conceptions of poetic language (94). Dolezel focuses on an aspect of Wordsworth and Coleridge not usually examined as an important event in the history of poetics the disagreements between the two. Both Wordsworth and Coleridge attack poetic diction and emphasize innovation and renewal as basic for poetic language. Woodsworth, though apparently advocating poetic language as ordinary language, contrains ordinariness by stressing expressivity, which is circumscribed by his assumption of a functional (teleological) view of language as pursuing an aesthetic aim (82) Coleridge differs from Wordsworth in seeing the relation of poetic and ordinary language as well as expressivity as solely a stylistic variable and not a linguistic constant. Also, unlike Wordsworth, Coleridge's poetic language is limited to metrical poetry. Since not all metrical composition is poetry nor is all poetry metrical, Coleridge passes the difficulty by postulating, according to Dolezel, a functional hierarchy distinguishing poetic from non-poetic (scientific) language: poetry has a primarily aesthetic function, science a cognitive one, though they are not mutually exclusive. The contrastive matrix Dolezel infers from Coleridge's theory is based on the presence or absence of 1. the aesthetic function, 2. metrical organization 3. and the absence or presence of communication of truth, all three of which distinguish poetic from nonpoetic language (88).
To highlight and supplement the semantic aruth-functional) status of poetic language for which Coleridge's theory has implications, Dolezel examines Frege's semantics. The contrast between poetic and non-poetic language can be explained in terms of Frege's distinction of two constituents of meaning: reference and sense. In the light of this distinction poetic language, as opposed to non-poetic language, is lacking in reference and thus has no truth value though it has sense.
The second part of the book, chapter Ew, takes up the systematization, in the twentieth century, by structural poetics of the themata and models discussed in part one. Structural poetics is enabled by an epistemological shift manifest in the supersession of the organic model by a semiotic one. The shift is evident in linguistics, which developed formal models, explicit theories and a technical terminology and thus made for a development of a linguistic poetics of a more exact character as well as a literary semantics focusing on what and how texts mean. Two distinct though overlapping semantic systems are discussed by Dolezel: The expressive semantics (stylistics) of Bally, Lanson and Grammont, who offered a "meticulous analysis of the expressive potential of language" (103), thus contributing to poetics of both verse and prose; and structural semantics. Breal and Saussure's linguistics, which constitute the dominant method of studying meaning in language.
Saussure's model is mereological in that it is selfcontained, interrelated and non-additive (113). He provides a technical dichotomous terminology (signifier/signified; difference/opposition; paradigmatic/syntagmatic) which is very influential Saussure's mereology posits the linguistic sign in terms of two different structural relationships: opposition and difference and paradigm Xsyntagm. This mereology far from being static can account for semantic change, according to DolezeN precisely because of the arbitrary fink of signifier and signified. To find an appropriate semantics for poetic language, where creative innovation and change are not based on analogy- the vehicle for change in ordinary language-- Dolezel turns to Saussure's study of anagrams, which purported to show that poetic meaning is a double structure of covert and overt elements. These studies, Dolezel says, indicate the major flaw in Saussure's attempt since his quest for '"organized' covert meaning ended as a purely subjective interpretive practice" (123).
Chapter six deals with Russian Formalism and its almost ignored antecedents in the German morphological poetics of narrative. Dolezel examines in detail DibeNus and others, and highlights their specific conceptual apparatus and also their conception of form as a "necessary skeleton" of poetic structure but an aesthetically empty one (127), their view of the cooperation between theoretical poetics and historical description (133) and their zigzag method as important contributions to poetics. In discussing the Formalists Dolezel concentrates on peripheral figures in whose work the notion of the dominant first appears He also shows how in some of the peripheral figures the basis for Propp's morphology was established, again unearthing historical background. Propp is briefly discussed mainly to point out that his narratology, in fine with the German morphological tradition, is a universalist semantics.
The last chapter discusses Prague school semiotics and concentrates on its mereological semiotic model and its systematized structural poetics. Prague school semiotics is based on the idea of literary communication and stands in contrast to 'the formulist, the expressive, the mimetic and the sociological' conceptions of art (148). It is interested in the specificity of literary communication. The mereological model it presents is stratified and hierarchical. It includes a supralinguistic stratum of thematics (necessarily expressed in language), and the aesthetic opposition of "material" and "form" is superimposed on all strata (155). Both themes and language are material which is transformed into poetic structure by form (i.e., the two procedures of deformation and organization). The mereological model is developed so that both material and formal constituents are assigned meaning-creating value and thus, from a semiotic perspective, a literary work " is a totally semanticized structure (156) at the same time as meaning is seen as a dynamic, bi-directional process of accumulation.
Dolezel also points out the lacuna in Prague poetics with regard to the question of reference and the problem of fictionality and highlights the pragmatic factors in semiotic poetics-- the concern with the communicating subjects (sender receiver) and with the social conditions surrounding the communication. The subjective factors are neutralized by "a balance between individual and supra-individual factors in literary communication" (161). The final aspect of the communicative semiotic poetics Dolezel discusses is that he calls transduction, a broad conception covering such phenomena as "literary tradition, intertextuality, influence and cross-cultural transference" are critical reception, that is the processing of metatexts, and literary adaptation, that is the literary transformations into other literary texts (171).
Dolezel's book seems to me to be an important contribution to the theoretical study of poetics and its history. Because of its wide scope (temporally and geographically) it discusses texts of lesser known European figures probably not familiar to many readers. Even when discussing well known figures, such as Aristotle, Coleridge and Wordsworth, Dolezel does not tread the hackneyed path. His comments, theoretical reconstruction of models, integration of different poetic traditions, notes, disagreements with accepted views (e.g. even the use of Else's translation of Aristotle's Poetics rather than Butcher's) are original and interesting.
The one flaw I found surprising in such a comprehensive discussion is that not so much as a mention is made of New Criticism as part of occidental poetics. Granted, it could be argued that New Criticism, as its name and its practice indicate, belongs to criticism rather than poetic, but I think that is misleading. Certainly it falls within the purview of the occidental tradition before 1945, at least in its formulative stage. Its influence on twentieth-century Anglo-American criticism and theory, even poststructuralism (in its total rejection of New Critical views) is not to be doubted. It is a "school" which views literature as a distinct object of study with its own unique language and also as a rival mode of cognitive knowledge to science. It is part of poetics in Dolezel's sense no less because of its "objectivist" attitude and methodology in imparting knowledge. It may not have presented a systematic model until after 1945 (Wellek and Warren, Wimsatt, Beardsley), but one can easily be inferred (as Dolezel has done for Coleridge, or for Aristotle's epistemology) and such a model is without doubt structural-mereological and morphological in its outlines. Perhaps New Criticism is so well known and has been studied so extensively that Dolezel felt he could easily omit it. But it deserves at least a reference in a note, even one stating why it is not discussed, or a brief mention such as the account of normative poetics between Aristotle and Leibnitz.
Despite this lacuna, the main contribution of Occidental Poetics is that, though it is a metapoetic text, Dolezel's book belongs in and exemplifies the tradition of poetics he discusses in that his own study conforms to the cognitive research tradition by his analytical classification of conceptual issues, his careful definition of his terms and his methodical, systematic development of the continuity and change he traces. It is an informative, erudite study, it advances knowledge about the invariants (themata) of literary theory and is convincingly and lucidly presented.
Willi Diengott is a Professor at The Open University of Israel in Tel Aviv.