This version is pre-editing; for the final version please see Comparative Critical Studies, Volume 10, No. 1, 2013, 68-91.
WHAT IS ‘COMPARATIVE’ LITERATURE?
This approaches the problems of self-definition surrounding ‘comparative literature’ by analysing the nature of comparison per se, a the place of this practice in literary criticism past and present. It argues that comparison in its broadest sense is involved in all thought, but that comparison in the strictest sense is involved in only a minority of criticism, whether described as comparative or otherwise. Certain works of literature call especially clearly for a comparative approach, through allusion to other works, or through establishing internally comparative structures (for example in parallel plots); such works might collectively be denoted by the noun phrase ‘comparative literature’. The nature of ‘comparability’ is analysed, and various factors affecting the results of comparison are noted in turn, the topic on which the comparanda are compared, their number, and the degree of detail of their description. It is argued that literary criticism would benefit from greater self-consciousness with regard to comparison, and that departments of comparative literature would be well-placed to lead the process of theorising comparison, which hitherto has been remarkably overlooked.
Comparison per se
This concerns a practice which is involved in all reading, yet has hardly ever been the explicit subject of literary theory. Comparison, in the broadest sense of term, is the mental process which enables us to perceive similarity and difference. Smells and ideas cannot be distinguished without perceiving their similarities and differences to others. Will cannot be exercised without comparing options; to choose comes from gusto, and involves, as Sainsbury’s would have us do, tasting the difference. A critic describes a literary work as mimetic only after comparing it with both life and other works; it is only comparatively mimetic. Matthew Arnold, who coined the term comparative literature as a translation of comparée, claimed in his inaugural lecture at Oxford University in 1857 that ‘No single event, no single literature is adequately comprehended except in relation to other events, to other literature’. In our own century, Richard Rorty wrote: ‘Good criticism is a matter of bouncing some of the books you have read off the rest of the books you have read’. He might have added that good reading of criticism involves bouncing the criticism you are reading off the rest of the criticism which you have read. No text reads the same twice. A rereader always compares the text as read with their memory of that text as read in the past; this comparison runs synchronously with a comparison of what they have just read with what they anticipate is coming next.
Yet these functions of cognition are so ubiquitous that their designation as comparison reduces the usefulness of the term. A narrower sense of the term is to denote the act or results of paying a similar quantity and quality of attention to a discrete number of in order to determine their similarities and differences with regard to possession, lack, or degree of possession of a particular quality. A minority of literary criticism practised today is of this kind – both of the inter-national, inter-linguistic, inter-artistic kind which presents itself as comparative literature, and of the criticism which includes none of these divisions (hence Earl Miner’s observation that ‘it is difficult to discover actual comparison in the world of the comparatists’). A comparison of George Eliot with George Sand on the subject of (for example) literary realism may have the interest, but also the complication, of involving linguistic and cultural variables which are not directly related to A comparison of George Eliot with Elizabeth Gaskell on literary realism, which involves fewer circumstantial variables, may be more cleanly comparative, and in this sense more comparative. Asymmetric comparison – paying a different quantity or quality of attention to two comparanda – has strong similarities with much criticism which is not usually considered comparative. For example, studying the influence of Miguel de Cervantes’s El ingenioso hidalgo don Quixote de la Mancha on Nikos Kazantsakis’s Βίος και Πολιτεία του Αλέξη Ζορμπά [Life and Adventures of Alexis Zorbas] has much in common with studying the representation of attitudes towards sex in rural 1930s Greece in the latter novel. In both cases one is looking for features of one complex object (a novel, and an aspect of a culture) in another; the discussion of any topic in literature involves a comparison of the form looking for X in Y.
In the sixth yearbook of the British Comparative Literature Association’s Comparative Criticism Elinor Shaffer commented: ‘Conducting a retrospective inquiry into specifically comparative modes of close analysis, we find that a very few comparative literary handbooks offered some direct discussion of comparative analysis of texts’. The position of comparison as a topic in philosophy is also undeservedly obscure (when trying to find materials on comparison, Earl Miner too found that ‘To my surprise, philosophers were equally dumb’). No English-language reference work of philosophy of which I am aware has an entry for the term, despite the fact that comparison is as important a method to philosophy as to literary criticism, and that it is fraught with philosophical implications. (Resemblance is discussed in philosophy, but is a simpler concept than comparison by virtue of lacking an epistemic agent). Therefore all literary criticism is comparative in a broad sense, whereas much criticism called comparative is not comparative in the strict sense of being conscious and methodical.
The academic subject comparative literature, it is commented with a frequency which has tamed it into a reassuring truism, is anxiogenic. This is partly because it is not easily defined either by method or matter. In the 1970s Robert Clements commented that ‘Comparative Literature sometimes figures in university curricula, but very few people know what they mean by the term’, and the last two decennil reports of the American Comparative Literature Association defined themselves as concerned with the state (and therefore also the nature) of the discipline, rather than, as previously, on the standard of what was performed within it. In 2006 Robert Weninger claimed that ‘nothing is written or published in comparative’, and pointed out that the Bernheimer report had dropped the proud initial capital letters from the discipline. (xii) Even this, it seemed, was too bold a move, and the Saussy report oscillated between comparative literature and Comparative Literature.
The problem of defining the subject by method is, as I have argued, that much of what is done under its remit is not methodically comparative. The bibliography of the British Comparative Literature Association’s first (1979) Yearbook of Comparative and General Literature contained sections for works concerning ‘Literary Genres, Types and Forms’, ‘Bible, Classical Antiquity’, ‘Individual Countries’, ‘Individual Authors’, and ‘Comparative, World and General Literature’. The three terms of the last category are often themselves imprecisely distinguished – just as общая литература, Allgemeiniteratur, littérature générale, and literatura universal are from the equivalents of comparative literature in their own languages.
The question then arises of whether comparative literature should simply be called, and become, the study of literature. Proponents of departments of literature include René Wellek and Austin Warren, who in 1949 argued against the idea of national literatures: ‘There’s just literature’. Fourteen years later Wellek wished that ‘we could simply speak of the study of literature […] and that there were, as Albert Thibaudet proposed, professors of literature just as there are professors of philosophy and of history’. And in 2006 Jonathan Culler argued against Charles Bernheimer that: ‘The turn to culture makes sense for national literature departments: the division of literature by national or linguistic boundaries was always rather dubious, but such divisions as these are a very reasonable way of organizing the study of culture’; this would leave comparative literature with the distinctive role of studying literature: ‘As the site of the study of literature in general, comparative literature would provide a home for poetics’. Objections to such a plan come from those who consider that literature should always be related to culture in its broadest sense, and to other art forms. In 1972 Levin, and in 1995 Bernheimer, argued that comparative literature did not and should not concern literature alone. It is also objected that the general study of literature in practice rarely fulfils that remit, consisting largely in the study of European literature and its nearest relatives. Clearly, European literature, not general, world, or comparative literature should be the title of courses if that is what is studied; mwriting in response to that report urge: ‘Study these interconnected European literatures, I say. They make sense together. They were made for each other’.
Both parts of the title comparative literature, then, imperfectly denote the subject’s de facto remit. works of criticism which explicitly compare literary works to each other – for example George Steiner’s Tolstoy or Dostoevsky: An Essay in Contrast (1959), or Colin Burrow’s Epic Romance: Homer to Milton (1993). The first is a cleanly bilateral comparison; the second traces the treatments of a common mode (epic romance) in a series of works across time, and in so doing compares those works as participants in a tradition. Criticism which, as Peter Brooks claimed of himself as a graduate student, is not ‘comparing literature, just working in more than one’be described (if any such description is necessary) as supra-national or supra-linguistic criticism – for example, Erich Auerbach’s Mimesis: Dargestellte Wirklichkeit in der abendländischen Literatur (1946). This work explores modes of mimesis in a large number of texts, but unlike Epic Romance: Homer to Milton its overwhelming concern is with its central topic, not with bringing its examples into mutual relation. In addition, criticism which brings literature into relationship with any other art form – for example Diane Gillespie’s The Sisters’ Arts: The Writing and Painting of Virginia Woolf and Vanessa Bell (1988) – should be classified as inter-artistic criticism rather than as comparative literature, even if (as is often the case with such projects) the institutional home of the work is a department of literature. , the anxiety surrounding the phrase comparative literature would be much diminished. Departments currently named comparative literature could be renamed as departments of inter-national literature, general literature, or literature.
I have already tried to define comparison as a , but a comparison is both an action and its outcome – making a comparison can refer both to the process of comparing and to the description of this process and its result (one cannot be described without the other). For example, were someone to say that a historian compares Hitler and Stalin, she might mean that the historian tries to discover the similarities and differences of these men, or that he draws attention to such similarities and differences as he has found them to possess. This is an important ambiguity – between the performance and results of comparison, between the discovery of results and their dissemination, and therefore between empiricism and rhetoric.
Language is not necessary to the performance of comparison, but it is to its description, in which it can prove limited. In English the language of comparison tends to imply one of three positions, which may be approximated to similarity, difference, and neutrality. One compares something and, with or to something else; and is neutral, with suggests the expectation of similarity, and to suggests the expectation of difference. Something is the same as something, but different to, from, or than it. Apart from the fact that to is more common in British English, and fromand than, in North American English, to implies orientation towards the differing other, from implies departure from it, and than implies an alternative to and possible displacement of it. The comparer should compare and choose her words with care. One likens to and contrasts with, but these collocative prepositions are perverse, and would in theory function better with the reverse verbs, liken with and contrast to. This would leave and as a neutral conjunction best befitting compare, as in ‘I intend to compare [contemporary Russian playwright] Vassily Sigarev and Martin McDonagh’.
In contrast to to contrast (contra-stare, to stand against), to compare also means to regard or represent as analogous or similar, and, intransitively, to be of the same or similar quality or value – as in ‘Ibsen compares with Lorca as a tragedian’ – hence examination questions beginning ‘compare and [by contrast] contrast’. H.M. Posnett, one of the first Anglophone theorists of comparative literature, in the 1880s implied identity by comparison when he asserted that ‘The most colourless proposition of the logician is either the assertion of a comparison, A is B, or the denial of a comparison, A is not B’., compare is an analogy, equal, or rival of something else. Many terms for comparison stress likeness over difference: to com–pare is to bring together parities, vergleichen makes gleich [the same], сравнить´ makes равный [equal], and a сравнение is a simile as well as a comparison. The ancient Greek παραβολή [from παρα plus βολή, a casting, throwing, or putting] is a placing side by side, or an analogy. In a parable, as in an allegory, something is made to stand for something else on the basis of similarity or translation; παραβολή was borrowed in the Latin parabola, or comparison, and in post-classical Latin it is an allegory, proverb, discourse, or speech – an expansion of meaning which acknowledges the importance of comparison to rhetoric. The Latin comparare also meant to place together, couple, unite, pit against, treat as equal. By contrast, the modern Greek term for comparison, συγκρίνω [to judge together] avoids the prejudgement of results which pertain to both compare and contrast. Whereas the Latin instruction cp. in practice often invites contrast, cf. invites open-minded comparison.
Of course, no two things (in sofar as absolutely discreet objects can even be conceived) are identical or absolutely different; they attract comparative investigation because they are felt to be a metaphor in Todorov’s sense – constituted by the tension of difference and resemblance, separateness and communication. That is, an initial comparison will have suggested either that the comparanda are different (an adjective used rhetorically to indicate that they are more different than might be expected) or, more often, that they are similar (that is, more similar than might be expected). A critic might decide to pursue a comparison of Pride and Prejudice and Mansfield Park in order to try to explain the striking differences of these nearly contemporaneous works by the same author; or might decide to compare Charles Dickens and Franz Kafka (as did Mark Spilka in 1963), having discerned that they share profound similarities as well as obvious differences. The idea of an initial comparison preceding further comparison indicates another ambiguity in the word, which can refer not just to a methodical process, but to the unexamined impression which prompts it.
When comparing literary works originating in different places, differences between them are the assumed basis, and one of the ends, of the investigation: the background of divergence against which the similarities which suggested the comparison appear, and the finer points which appear against those similarities. Earl Miner notes that ‘Comparison is feasible when presumptively or formally identical topics, conditions, or elements are identified. Of course what is presumptively but not actually identical soon betrays difference. With tact and luck, however, we may find the difference just great enough to provide interest, and the presumed identity strong enough to keep the comparison just.’ For example, Mark Spilka acknowledges that Dickens and Kafka might appear to be antitheses, as respectively English Gentile, Victorian, comic, extrovert, and moralistic – and German-Czech Jewish, twentiethcentury, pessimistic, introverted, and neurotic. Against these differences Spilka places what he discerns to be the two authors’ elective and accidental affinities – obsessions with familial exclusions, the bestiality in humanity, infantile perspectives, and the grotesque. In describing examples of these in each he continues to acknowledge their differences (for example Dickens’s greater habitual warmth), but is placing these differences in the new perspective of their similarities. Rather than being simply very different authors, Spilka makes them metaphors (in Todorov’s sense) for each other.
Description of difference in relation to an other is one aim of comparison, but description of difference in relation to the self is not. The very impulse to compare complex objects produces the attendant impulse to stabilize at least one of the comparanda rather than pay attention to the instabilities of all simultaneously; to limit the length of investigation of each literary work, for example, in order to maintain contact between them, and to limit the potentially-infinite discovery of difference in thick description – and, therefore, of non-comparability. Of course, a similar problem presents itself in relation to any literary analysis; the logical end point of concern for accuracy of description is simply to quote a text in its entirety. However, in comparative analysis the factor which limits accuracy of detail is not just the need to make one’s analysis finite, but to hold the comparanda in palpable mutual relationship. Clearly, Mark Spilka found that there were features of Kafka’s writings which could only or best be described as Dickensian, and of Dickens’s which could only or best be described as Kafkaesque – but this does not mean that he necessarily felt that he was giving an adequate description of either author or any one of their texts, taken alone.
In addition, the arts, unlike the sciences, are infrequently able to use quantitative units in comparison (although they might do so more often than they do), but rely on a crude vocabulary of identity, opposition, equilibrium, and comparatives, modified by intensifiers and qualifiers. Most comparative cadences in literary study assert either identity or difference: ‘Both Война и Мир [War and Peace] and Анна Каренина [Anna Karenina] treat the Russian peasantry as a repository of value’; ‘In contrast to War and Peace, Anna Karenina presents itself as a novel in the European mode’. The vague term relatively is used to indicate a relatively small degree of difference. The phrases just as and (more conscientiously) rather cover a range of degrees and types of similarity; whereas covers a range of differences; the present assertions are no more precise than what they describe. Or little more. Or hardly more. Yet in explicitly comparative work the degree of descriptive detail attained is crucial, since it determines what is described as a similarity and what as a difference. In practice the transition from the first to the second often involves a slight increase in detail. ‘Both Война и Мир [War and Peace] and Анна Каренина [Anna Karenina] treat the Russian peasantry as a repository of value’ – but ‘Whereas War and Peace actually presents admirable peasants, in Anna Karenina the idea of the Russian peasantry as a repository of value exists largely in Levin’s mind, and even that only once he has had the epiphany which concludes the novel’.
It is salutary to be reminded of the flexibility of such terms as similarity and difference, which are such heavily-used tools of thought, and are supposed antonyms; similarity is merely difference on a relatively small scale. A bespoke tailor will not make a pair of trousers exactly symmetrical; asymmetry is the product of attention to detail. Moreover, the choice of describing a relationship as a similarity or a difference can be determined in the interests of rhetoric. Indeed, comparisons have long been associated with not only rhetoric but odiousness: ‘Odyous of olde been comparisonis, And of comparisonis engendyrd is haterede’. (Lydgate, 1430) In Shakespeare’s works comparison is repeatedly characterised as quibbling, equivocation, jibing allusion and scoffing analogy. Rosaline in Love’s Labour’s Lost disparages Berowne as renowned as ‘a man replete with mocks;/ Full of comparisons and wounding flouts’; Beatrice predicts that Benedick will ‘but break a comparison or two on me; which,/ peradventure not marked or not laughed at, strikes him into melancholy’; Falstaff complains of Hal as ‘the most comparative, rascalliest, sweet young Prince’. (Love’s Labour’s Lost 5.2.30-31; Much Ado About Nothing 2.1.136-68; Henry IV Part 1 1.2.80-81).
But – if it is done conscientiously – one condition of a methodical comparison being considered worthy of pursuit is that the entities concerned are in fact comparable. Comparability involves a degree of similarity in the comparanda. Of course, any thing can be compared with anything else, and comparable resembles similar and differentin being a relative, not an absolute, term, the applicability of which rests on a comparison. The same is true of non- and in- comparable. Charges of non-comparability rhetorically assert that vacuity, tactlessness, unfairness, or some other wrong would be involved in pursuing a comparison. These are statements of value. The assertion that any one thing, rather than an assembly of things, is incomparable, or beyond compare, implies that the qualities which it has in common with the other things most like it are trivial in comparison to its distinguishing characteristic/s, and that the pursuance of comparison would involve paying insufficient attention to those characteristics, thus rendering the comparison either trivial or invalid. The assertion ‘you can’t compare J.K. Rowling to C.S. Lewis’ implicitly argues that their similarities of subject matter are unimportant compared to their differences of accomplishment, and that to describe either would be at best uninteresting and at worst insulting to Lewis. Similarly, Orsino tells Viola in Twelfth Night: ‘Make no compare/ Between that love a woman can bear me/ And that I owe Olivia’. (Twelfth Night: 2.4.100) Sometimes people describe a work of art as incomparable not only in order to express admiration for it, but also to imply that it is in the nature of the work’s excellence to determine a mode in which it alone should be explored, and which is always by far the most valuable mode in which to explore it. The phrase ‘that incomparable work of literature, Dante’s Divina Commedia’ asserts that the work is superlatively good, but also implies that it must be understood on its own terms. This is also what is meant by claims of uniqueness (since everything is unique as well as comparable, it too is a relative descriptor). The most extreme version of this argument is that the work’s own terms are the only ones on which it can be understood. The implication (rarely embraced) is that the work uses a private language in Wittgenstein’s sense, and is therefore incomprehensible. Comparatists assert both comparability and comprehensibility. Peter Szondi asserted that ‘Kein Kunstwerk behauptet daß es unvergleichbar ist (das behauptet allenfalls der Künstler oder de Kritiker), wohl aber verlangt es, daß es nicht verglichen werde’ [No work of art declares that it is incomparable (at most it is the artist or critic who claims that), but every work of art demands that it not be compared]. But this is clearly wrong- certain works do clearly ask to be compared with others: James Joyce’s Ulysses and Derek Walcott’s Omeros to the Ὀδύσσεια [Odyssey] of Homer for example. One entity of which incomprehensibility as well as incomparability is sometimes asserted is God – a claim made by several kinds of theism. The same compound assertion is made in order to express or advocate a sense of quasi-religious awe in relation to non-divine subjects – for example, since the later 1960s, to the Holocaust. A bill proposed to the Israeli Parliament in January 2012 would have outlawed comparisons to the Nazis.
Assertions of non-comparability as applied to combinations of objects are often based on a sense that the way in which they are most likely to be compared will not generate valid results. Neither apples nor oranges are proverbially asserted to be intrinsically incomparable, but they are asserted to be mutually non-comparable – presumably because their similarities of size and use generate the risk that they will be judged according to the same criteria, and an orange would be unfairly criticized as less crisp than an apple (in this sense the idiom may be contrasted with chalk and cheese – more of a contrast). Hazlitt stated that ‘Comparisons are […] impertinent, and lead only to the discovery of defects by making one thing the standard of another which has no relation to it’. Similarities should therefore not be allowed to obscure differences which affect comparability; this is perhaps the sense behind the perfect rhyme in the roughly-equivalent Serbian phrase поредити бабе и жабе [to compare grandmothers and toads]. A critic might for example assert that Madame Bovary and Effi Briest are not properly comparable as novels of adultery, because of the differences of time and place of their production.
The charge of incommensurability denies that a certain type of measure can be applied to all of the proposed comparanda. For example, Spanish has an idiom which disparages sumar peras con manzanas [adding pears and apples]. It is certainly possible to count pieces of fruit, but the specific category of pear-or-apple is, the idiom implies, of little interest. The Russian idiom сравнивать тёплое с мягким prohibits the comparison of the warm with the soft, since no single measure can be made of warmth and softness. Finally, certain qualities are differently perceived by different people. The Hungarian idiom ízlések és pofonok [tastes and smacks] suggests that the relative value of different objects cannot be absolutely decided if they are judged on qualities which are differently perceived by individuals, rather as two smacks in the face cannot be compared if they are received by different people. It is also the case that any comparison must be performed by one person. Certainly, non-subjective qualities can be determined as belonging to different objects by different people, as Franco Morretti argues.  The objects can be compared by a third person making use of their descriptions. Most comparative literary criticism – like any other kind – is done under the influence of others’ criticism of the works concerned. But for a comparison per se to take place, the comparandamust be apprehended by one mind. Relative unity of physical or conceptual place assists the equally-important unity of time. It helps to place the texts which a critic wishes to compare literally side by side. The end result of comparison is generated in an instant in which the qualities of the comparanda – a similar kind of grotesque in the works of Dickens and Kafka, for example – are simultaneously present to the comparer’s mind.
Comparison in Comparative Literature
I have suggested that today comparison is a minority pursuit. But i has played a central part in the development of criticism as a subject – notably, in a European context, in comparisons of the Greeks to the Romans, and the Ancients to the moderns (even today, as Earl Miner notes, ‘It is left to students of the Renaissance, the Middle Ages, and antiquity to do pretty much what the comparatists are thought to do, only under other names.’) Comparison and criticism were connected more systematically in the later nineteenth century, when literary studies were modelled on the evolving scientific disciplines: Literaturwissenschaft, on Naturwissenschaft. Specifically, comparative literature was modelled on other subjects with comparative in their titles, including philology, biology, and philosophy. Science, in part, proceeds inductively through comparison; experiments analyse a comparandum in relation to an isolated variable, and observe deviations from the secundum comparatum or control. Comparative philology operated by observing similarities in languages which had hitherto been assumed to be unconnected, and then by both using historical information to explain the connection, and by inducing historical hypotheses from the connection.
The results of such comparisons were sometimes explained, in the dimension of time, with the use of a tree metaphor. Indeed, tree-shaped comparativism has shown considerable durability in literary study, where it has tended either to point to similar social conditions generating similar literary phenomena (such as the rise of the novel in industrialising countries), or to posit direct influence between phenomena (such as naturalism as a development or offshoot of realism). Posnett tended to the first type of analysis in his comparisons of clan, town, national, and world literature. More recently, Franco Moretti described the history of British nineteenthcentury detective fiction in evolutionary terms, showing the results of his symmetric comparisons of novels in tree diagrams which showed the divergence and convergence over time of what he denoted as different genres.
The study of influence, which has proved a stronger and more enduring vein of criticism than social comparison, is necessarily asymmetric. Aleksei Veselovskii, brother of Aleksandr and co-founder of the Department of World Literature at Moscow University with Nikolai Storozhenko in 1873, stated at the beginning of his 1881 The Western Influence in New Russian Literature that ‘he exchange of ideas, images, fables, artistic forms between the tribes and peoples of the civilized world is one of the most important things studied by the still-young science of literary history’. In 1961 Henry Remak criticized French criticism for its emphasis on influence studies rather than comparison in the strictest sense, arguing that ‘Purely comparative subjects constitute an inexhaustible reservoir hardly tapped by contemporary scholars who seem to have forgotten that the name of our discipline is “comparative literature” not “influential literature”’. Similarly, in 1992 Peter V. Zima, who used the terms ‘der genetischer Vergleich’ for contact studies and ‘der typologischer Vergleich’ for the study of at ‘Ähnlichkeiten’ [similarities] which are produced, without contact, on the grounds of similar conditions of production or reception’, argued that ‘Die Erfahrungen der Sozialwissenschaften zeigen indessen, dass nicht die genetischen, sondern gerade die typologischen Beziehungen als Grundlage der Komparistik werden sollten’ [the experiences of the social sciences indicate that not genetic, but precisely typological relationships, should become the basis of comparativism]. In the following year Claudio Guillén distinguished three supranational bases for literary comparison: influence, socio-historic conditions, and critical methodology. Yet in practice, these bases are not easily separable. Zima had argued that ‘Kontakt und Analogiestudie einander ergänzen und in vielen Fällen nich unabhängig voneinander durchgeführt werden können’ [‘contact studies and analogy studies comple each other, and in many cases can’t be pursued in mutual independence’]. Furthermore,some critical methodology is necessary to the observation of those similarities and differences which may then be explained in terms of socio-historic conditions or influence; in this sense, Guillén’s third basis is the only ‘basis’, and the other two are contexts used to explain the results found thereon. For example, Moretti classified and compared detective novels according to the kinds of clues which they gave to the reader, and explained the survival and extinction of genres in terms of the literary marketplace. On the other hand, critical methodology can be viewed as culturally specific as the literature of which it purports to provide the basis of comparison. In 1992 Earl Miner argued that every culture bases its poetics on the basis of the most dominant genre at the time (in the West, drama; in the rest of the world, lyric), and that these theories, brought into inter-relation, should constitute the basis of intercultural poetics.
Most comparativism is not so rigorously theorised, however. In 2006 Jonathan Culler argued that ‘World literature courses that bring together the great books from around the world seem to base comparability on a notion of excellence, so that comparison – the principle of comparability – rather than opening new possibilities for cultural value, more often than not restricts and totalizes it’. However, courses of world and general literature do not necessarily assert the comparability of the works of literature they select, any more than they demand their comparison; indeed, they may purport to make their selection on the criterion of incomparability. When comparison is required, it will not be the comparison of excellence (the excellence of a work of literature being connected to its uniqueness), but will have reference to ‘specific intellectual norms or models – generic, thematic, historical’ which, Culler himself argues, ‘are subject to investigation and argument in ways that the vacuous bureaucratic norms are not’.
In the phrase comparative literature, comparative is the attribute of literature. Yet it is almost never understood in this way, the semantic meaning having drifted apart from the compositional meaning. The same is true of vergleichende Literatur and сравнительная литература, although not of comparée, in which the literature is the passive object of comparison. Clements noted that the equivalent
‘East Asian terms are a compound essentially of two substantives. The Chinese pi-chiao wên-hsüeh, the Japanese hikaku bungaku, and the Korean pigyo munhak consist of “comparison” plus “literature”. The terms thus denote […] the scientific comparison of two or more literatures without inclusion of adjectival modifiers. Perhaps if we followed suit and adopted the simple “literature comparison” we might eliminate a great deal of discussion’. On the other hand, Wellek considered that ‘There is little use in deploring the grammar of the term and to insist that it should be called “the comparative study of literature”, since everybody understands the elliptic usage’.
Yet the phrase comparative literature could be seen to point to a meaning which does correspond with its compositional sense: literature which invites the performance of internal comparison, or which, to put it another way, comparisons. This is to use the noun comparison in a sense distinct from the two discussed above, a process and its result. In this sense a comparison is a quality or set of qualities which may obviously or easily be interestingly compared with another quality or qualities in the same literary work. The work which contains them cannot be well understood without the performance of this comparison. Waiting for Godot is comparative between its first and second halves. William Thackeray’s Vanity Fair, Lev Tolstoi’s Anna Karenina, Hermann Hesse’s Narziss und Goldmund, and Patrick White’s Voss all constitute comparative literature – not by virtue of containing significant contrasts, which most literature does, but by virtue of presenting parallels and divisions between two protagonists which are more sustained and ostentatious than those presented by most literature. This might be the most useful, and grammatically cogent, application of the term comparative literature.
A comparison of novels as comparative works of literature is a second-order comparison similar to the comparison of ratios. This kind of comparison possesses the advantage of confessing the variable of context. To say that ‘Daniel’s relationship to Gwendolen is the equivalent in Daniel Deronda of Levin’s relationship to Anna in Anna Karenina’ is less problematic than claiming ‘Daniel is like Levin’, or ‘Gwendolen is like Anna’. According to Crookshank ‘he comparability of two facts is a function of the comparability of their contexts’, and ‘scrupulous criticism’ ‘forbids the possibly fortuitous resemblance between two several datadetached from their circumstances being taken as significant’. He goes on to assert that in the comparative methodology of science or philosophy, the ‘guiding principle will be analogy, reasoning in accordance with what in mathematics is called a proportion, that is to say, the equality between two ratios: A is to B as Y is to Z. Such an equivalence is compatible with no matter how great a heterogeneity between A and Y, B and Z. […] Confucius was in China that which Socrates was in Greece’. Of course, internal literary comparisons likewise involve differences of context, and all assertions of the similarities of Gwendolen and Alcharisi (in Daniel Deronda) should be contextualized by the two women’s very different family circumstances, native countries, and musical talents: a simile might run: ‘Gwendolen is to her circumstances as Alcharisi is to hers’. That is, any comparison of components of complex objects is, implicitly or otherwise, a comparison of ratios. One is reminded that ratio is the etymological ancestor of reason, pace René Etiemble’s claim that ‘comparaison n’est pas raison’. Comparing the comparisons of two stories in three novels of two times and two countries makes this fact particularly clear. Of course, the questions remain of the relationship of China to Greece, of Anna Karenina to Daniel Deronda, and of Russian to English. Gwendolen and Alcharisi are comparable in a way in which Gwendolen and Anna are not, simply because they are parts of the same work of art. In this sense the two levels of comparison – within, and between, works – are importantly distinguished.
Most comparisons have one or both of two motives: the desire to compare the comparanda (for example, Evgenii Zamiatin’s Мы [We] and Aldous Huxley’s Brave New World) and the desire to explore the topic or topics on which they are being compared (for example, dystopianism in early twentieth-century literature). In the latter case the comparanda will be chosen according to the topic, and will not necessarily be compared directly with each other (the two novels might be discussed in adjacent chapters rather than in the same). In the former case, topics of comparison will be suggested by the comparanda – but complex comparanda may suggest an infinitude of them. A full description of an act of literary comparison therefore contains an adverbial phrase: ‘I compare A and B with regard to C (and D and E and F)’. Earl Miner considered that ‘In literary study, the tertium comparationis is not only not immediately given. It is not given at all.’ But this fails to distinguish between different kinds of comparative topics and modes of comparison. Steiner posited an axis from literal translation of texts, through imitation, to what he called the interanimation of texts within a national, linguistic, or broader cultural region. This interanimation may be observed in relation to particular topics or qualities, on which they might be said to compare notes. Anna Karenina and Daniel Deronda, for example, could be compared with regard to love, lust, married life, double-plotting, tragedy, comedy, art, politics, intellectualism, cosmopolitanism, God, children, Schopenhauer, death, misanthropy, satire, horses, railways, symbolism, kitsch – amongst many others. Complex topics of comparison generate a field of comparison – a nexus of subject matter and methodologies within which the novels are compared on the possession of simpler qualities. For example, a consideration of the ways in which these novels are realist or otherwise would require their description with regard to a range of literary qualities. Erich Auerbach rightly observed that ‘Schon die Auffindung des Ansatzpunktes […] ist Intuition’ (even the discovery of the starting point is a matter of intuition). A topic for comparison may be intuited as the highest common factor of interest of all of the comparanda: a comparison of race-horses which have nothing else in common might well concern their ability to race against other horses. The category later-nineteenth earlytwentiethcentury European novel, however, is too to bring novels of which this is the highest common factor into direct comparative contact. The highest common factor is likely to shrink and loose interest, the more comparanda are involved.
The concept of a topic may be replaced by any one of several metaphors, each of which implies a slightly different comparative method. An axis of comparison implies a quality according to the degree of possession of which the comparanda are placed along a single axis. A fulcrum implies asymmetric comparison: the idea would be that performing comparison is like magnifying the force of a secundum comparatum through a lever resting on a tertium comparationis in order to lift the primum comparandum into clearer view or onto the same level as the secundum comparatum – for example, to lift Mary Shelley’s novel Frankenstein into clearer view by applying the force of John Milton’s Paradise Lost through a lever resting on the fulcrum of the Fall. Auerbach uses Ansatzpunkt to denote a point of vantage from which different cultural objects may be simultaneously viewed.
Comparisons of quantity (for example, amount of reference to God) can to some extent be distinguished from comparisons of quality (for example, conception of God). However, this distinction, which is apparently one of kind, could also be expressed as one of degree, just as distinctions of degree can also be expressed as distinctions of kind. Clearer is the distinction between comparisons which do, and do not, employ a standard external to the objects being compared. For example, Iasnaia Poliana – Tolstoy’s estate – and Saint Petersburg can be compared on their distance from the fixed third point which is Moscow, on the single axis of distance. In more complex comparisons the result is more ostensive. F.R. Leavis compares George Eliot and D.H. Lawrence on sex as follows: ‘the point may be made by saying that they are not only equally unlike Maupassant in their attitudes towards sex; they are unlike in the same way’ (which is like saying that Tula and Iasnaia Poliana both lie in roughly the same distance and direction from Moscow). Masaki Hirai in Sisters in Literature compares the relationships of the two sisters of Eliot’s Middlemarch, E.M. Forster’s Howard’s End, and Lawrence’s Women in Love to those of Antigone and Ismene in Sophocles’s Antigone, on the analogy of describing musical variations upon a theme. No conversion to a single axis is possible here, and nor was it when R.A. Jelliffe attempted to foster comparativism by teaching a course on tragedy: ‘constant reference was made […] both to the governing idea of the course, the idea of tragedy, and to the substance and the treatment of one of these plays compared with another’. The relation of Вишнëвый сад [The Cherry Orchard] and Beckett’s Happy Days to ‘the idea of tragedy’ cannot be placed on an axis, but both plays can be raised to view on the fulcrum of tragedy, or – to use an alternative metaphor – viewed from the tragic high ground.
On the other hand, comparanda may be compared in relation to qualities which are generated by their very comparison. This may be illustrated by Anna Karenina’s reaction to her husband Karenin on her return home from Moscow. Anna has frequent social contacts with many men, but after meeting Vronskii she does not compare him and Karenin in relation to a real or imagined standard of man, but judges each largely on possession of the quality attractiveness to Anna when the comparison is between Karenin and Vronskii. Of course, this quality has as much reference to Anna as to those men, and critics comparing literature – unlike women comparing men – should seek to exclude intrinsically personal reactions as far as possible. However, comparisons of complex objects inevitably generate qualities which are peculiar to that comparison. In this sense compared works of literature could be thought of as involved in a mutual process, as suggested by the reflexive Russian verb cоотноситься [to correspond with or compare oneself] which, unlike sravnivat’/sravnit’, exists only in the imperfective aspect, and so is a process rather than a finite action. Wayne C. Booth classified the questions which may be asked of a text into those which it invites, those to which it responds, and those by which it is violated. (1988: 89) Ideally, comparatists bring together works which are capable of conducting with each other a conversation, on one or more topics, which is worth overhearing for what the conversation reveals about themselves and/or the topic. The topic chosen is not necessarily that about which each individually has most to say – but it must be able to provide the focus of between thoses texts Ian McEwan’s novel Atonement could thus be compared to Beryl Bainbridge’s Master Georgie on the topic of the fictional recreation of historical war, to Thomas Hardy’s Jude the Obscure on limited mobility between English social classes, to Jane Austen’s Northanger Abbey on lurid fantaising by young girls, to E.M. Forster’s A Passage to India on false accusations of sexual malfeasance, to L.P. Hartley’s The Go-Between on childish misunderstanding of adult sexuality, to Vladimir Nabokov’s Lolita on sexuality across an age gap, and to Albert Camus’s La Peste on the late revelation of a narrator’s identity – but none of these topics would generate fruitful comparison with any of the other comparanda.When comparison is insensitively performed, it is apt to exaggerate either likeness or difference. Fluellen does the first in his attempt to demonstrate the likeness of ‘Macedon and Monmouth’ in Henry V:
if you look in the maps of the world, I warrant you shall find, in the comparisons between Macedon and Monmouth, that the situations, look you, is both alike. There is a river in Macedon, and there is also moreover a river at Monmouth. It is called Wye at Monmouth, but it is out of my prains what is the name of the other river; but tis all one, tis alike as my fingers is to my fingers, and there is salmons in both.
(Henry V: iv. 7. 21-31)
The significance of the results of Fluellen’s comparison is on his own terms clear: similarities of places imply the similarities of their rulers, and the similarities of their rulers support his King’s claim to France. Literary comparatists, by contrast, face the question about their efforts: cui bono? Atonement resembles Northanger Abbey in its concern with lurid fantaising by young girls, but is a tragedy whereas Northanger Abbey is a comedy. What is gained by these observations? The comparatist can respond in one or both of two ways. She can try to establish the reasons for the results, in terms of space, time, or influence; for example, one reason why Atonement resembles Northanger Abbey is that, as Ian McEwan acknowledged, Jane Austen’s novel was ‘on his mind’ during his planning of the novel. Or the critic can try to establish the results’ significance. The latter, which might be related to the reasons, could lie in an improved understanding of the texts and of their genres, of the authors’ lives, oeuvres, countries, and languages, and of cultural modes. Ian McEwan appears the more pitiless in his treatment of his character Briony Tallis, in relation to the fact that her mistaken fantasies, unlike those of Catherine Moreland, destroy her own and others’ lives, and that she is justified by the novel’s implied author in feeling lifelong guilt. On the other hand, the resemblances of Atonement to a large number of other novels adds another dimension to the novel’s metafictional aspect. Briony’s step in the direction of actuality when she reveals that Briony is not just a character in a novel but is that novel’s narrator is qualified by her own resemblance to another fictional character who made the same move before her in literary history – Camus’s Dr. Bernard Rieux (in La Peste). Literary comparisons worth the trouble of performing therefore contrast with the Hatter’s riddle to Alice at the tea-party in Wonderland, ‘Why is a raven like a writing desk?’ (95). Neither the Hatter nor the March Hare have the ‘slightest idea’. Solutions can be and have been found; Carroll himself suggested, when asked: ‘Because it can produce a few notes, though they are very flat’. But these solutions do not, individually or collectively, indicate that ravens and writing desks are in Todorov’s sense metaphors (constituted by the tension of difference and resemblance, separateness and communication). Nor does their discovery entail much interpretative risk; the degree of validity and profundity of the answers to the riddle found is immediately obvious.
One factor which greatly influences the outcome of comparison is the number of comparanda. The results of comparing two comparanda are more likely to be conceivable on a single axis, of which they may involuntarily be considered to mark out opposite ends. Modern English does not employ superlatives unless at least three comparandaare alluded to, but that does not prevent the illusion that the comparandum out of two which possesses more of a given quality, is most in possession of it, as in Shakespeare’s English: ‘Not to bestow my youngest daughter/ Before I have a husband for the elder’. (Taming of the Shrew) Leavis exemplifies and embraces the exaggeration to which such comparison can give rise in literary criticism: ‘Lawrence sees what the needs are, and understands their nature, so much better than George Eliot. In the comparison, in fact, we have to judge that George Eliot doesn’t understand them at all’. The addition of a third comparandum makes it more likely that the results will be conceived on a two-dimensional field, and can also dilute the generalizations which may be suggested by a comparison of two: Henrik Ibsen may seem strongly representative of Norway (as opposed to Britain) when compared to Oscar Wilde, but less so if August Strindberg is added as a third comparandum. With reference to the third language, which used to be required of many comparatists in the United States, Saussy wrote: ‘the third language, like an uninvited guest, points to the things that a two-language pattern leaves out’; ‘the apex of the triangle just determined is also a point from which a new angle opens up for measurement’. Fluellen might have had greater difficulty in demonstrating that King Henry was a second Alexander had a third point of comparison been involved. Bernheimer celebrates comparison for revealing external presences within works: ‘the voice of comparative literature is “unhomely” and this very quality of dispossession – a kind of haunting by otherness – is that voice’s great strength’. David Ferris goes further, in celebrating comparisons which do not generate coherent results in the bluntly paradoxical assertion: ‘We compare what cannot be compared’.
Like any assertion of incomparability, however, this is either a relative statement or untrue. It was noted at the beginning of this essay that comparison is intrinsic to thought and willed action. Given this, I would argue that it is worth sharpening one’s skills at comparison, and consciousness of comparison’s attractions and dangers, in the intellectually challenging but practically sheltered environment of literary criticism. Such criticism cultivates sensitivity, since comparison requires empirical openness to the precise location of the centre of gravity which permits a balance of separateness and communication between the comparanda. The term comparatively is related to relatively, with the latter understood not just in its connection to relativism (which is not a necessary concomitant of comparative thought), but to relationships – to the understanding of any phenomenon in its relevant contexts, and in the light of potential alternatives. At a political level, the willingness to compare one thing or oneself with an other or others undermines absolutism. And it is an ethically sound aim of human interaction for individuals to respect their own and each other’s quiddity, whilst reaching to find maximum common ground. Moreover, ethical analyses may be assisted by comparative reference to moral benchmarks: far from inducing ethical relativism, their use forbids it. For Earl Miner, too, comparativism and relativism are antithetical – partly because of the common ground introduced by a culturally-neutral formalism (‘The best solution to this problem of controlling relativism known to me is […] identification of formally identical features in the things being compared’), and partly because of the common ground constituted by bringing into mutual relationship culturally-specific poetics (‘The more diverse the literatures drawn on for evidence, the better founded will be any poetics we derive – as also the more difficult will be any issues of relativism’).
Since comparison is involved in all thought, thought about comparison is necessarily self-reflexive. This is one reason why analysis of the use of comparison in literary criticism should form an important part of literary theory, and why comparative literature courses as they currently exist can serve as a home for literary theory. The difference of degree rather than kind between similarity and difference, the mind’s tendency to seek out equivalents, and the limited attention paid to any individual object being compared, applies to comparison in its broadest sense, from which comparison in the narrower sense is distinguished as much by degree as by kind, and which is unconsciously performed in everything from understanding linguistic différance, to reading Atonement in relation to all of the novels which one remembers, to choosing one’s lover. Thinking about comparison gives a better sense of where art fits into life – how it relates to it – and how it compares to it.
 This is developed from material in the monograph The Art of Comparison: How Novels and Critics Compare(London: Legenda, 2011). Its three central chapters concern Daniel Deronda, Anna Karenina, and Women in Love as double-plotted novels, and compares how they compare their plots – hence the puns of the title and subtitle. The introduction and conclusion set up and reflect on this exercise, and reflect on the nature of comparison itself, inside and outside of literary study. This article draws on these outer, meta-critical, chapters; I am grateful to the editors for permission to do so.
 Quoted in Susan Bassnett-McGuire, Comparative Literature: A Critical Introduction (Oxford: Blackwell, 1993), p. 1.
 Richard Rorty, ‘Looking Back At “Literary Theory”, in Comparative Literature in the Age of Globalization, edited by Haun Saussy (Baltimore: Johns Hopkins University Press, 2006), pp. 63-67 (p. 64).
 Earl Miner, Comparative Poetics: An Intercultural Essay on Theories of Literature (Princeton: Princeton University Press, 1990), p. 20.
 Elinor Shaffer, ed., Comparative Criticism: A Yearbook (Cambridge: Cambridge University Press, 1979-), 1984: xiv.
 Miner, Comparative Poetics, p. 21.
 Charles Bernheimer opened the 1995 American Comparative Literature Association report with the statement: ‘Comparative literature is anxiogenic’. Charles Bernheimer, Comparative Literature in the Age of Multiculturalism (Baltimore: Johns Hopkins University Press, 1995), p. 1.
 Robert Weninger, ‘Comparative Literature at a Crossroads?’, Comparative Critical Studies 2006, xi-xix (xii).
 René Wellek and Austin Warren, Theory of Literature (London: Jonathan Cape, 1949), p. 49.
 René Wellek, ‘The Crisis of Comparative Literature’, in Concepts of Criticism, edited by Stephen G. Nichols (New Haven & London: Yale University Press, 1963), 282-295 (p. 290).
 Bernheimer, Comparative Literature in the Age of Multiculturalism, p. 240.
 Bernheimer, Comparative Literature in the Age of Multiculturalism, pp. 51-57.
 Bernheimer, Comparative Literature in the Age of Multiculturalism, p. 53.
 H.M. Posnett, Comparative Literature (London: Kegan Paul, 1886), p. 73.
 Tzvetan Todorov, Les Genres du discours (Paris: Seuil, 1978), p. 226.
 Mark Spilka, Dickens and Kafka: a Mutual Interpretation (London: Dobson, 1963).
 Miner, Comparative Poetics, p. 22.
 Peter Szondi, Hölderlin-Studien: mit einem Traktt über philologische Erkenntnis (Frankfurt am Main: Insel Verlag, 1967), p. 21.
 BBC News website 10.1.12 http://www.bbc.co.uk/news/world-middle-east-16488042 accessed 20.1.12.
 William Hazlitt, Table Talk (London: Pickering and Chatto, 1998), p. 92.
 Franco Moretti, Graphs, Maps, Trees: Abstract Models for a Literary History (London: Verso, 2007), pp. 4-5.
 Miner, Comparative Poetics, p. 22.
 Moretti, Graphs, Maps, Trees, pp. 73, 75.
 Quoted in Rachel Polonsky, English Literature and the Russian Aesthetic Renaissance (Cambridge: Cambridge University Press, 1998), p. 18.
 Henry Remak, ‘Comparative Literature, Its Definition and Function’, in Comparative Literature: Method and Perspective, edited by Newton Phelps
Stallknecht and Horst Frenz (Carbondale: Southern Illinois University Press, 1961), 3-37 (p. 5).
 Peter V. Zima, Komparistik (Tübingen: Francke, 1992), p. 94.
 Zima, Komparistik, p. 95.
 Moretti, Graphs, Maps, Trees, p. 75.
 Miner, Comparative Poetics, p. 24.
 Jonathan Culler, ‘Comparative Literature, at Last’, in Comparative Literature in the Age of Globalization , pp. 237-48 (p. 242).
 Culler, ‘Comparative Literature, at Last’, p. 244.
 Robert J. Clements, Comparative Literature as Academic Discipline: A Statement of Principles, Praxis, Standards (New York: Modern Language Association of America, 1978), p. 11.
 René Wellek, ‘The Crisis of Comparative Literature’, in Concepts of Criticism, edited by Stephen G. Nichols (New Haven & London: Yale University Press, 1963), 282-295 (p. 290).
 Paul Masson-Oursel, Comparative Philosophy (London: Kegan Paul, 1926), pp. 7, 50.
 Masson-Oursel, Comparative Philosophy, p. 44.
 René Étiemble, Comparaison n’est pas raison: la crise de la littérature comparée (Paris: Gallimard, 1963).
 Miner, Comparative Poetics, p. 232.
 Steiner, George, ‘What Is Comparative Literature?’, Comparative Criticism, 18 (1996), 157-71 (p. 436).
 Erich Auerbach, Mimesis: Dargestellte Wirklichkeit in der abendländischen Literatur (Bern: Francke, 1946), p. 306.
 Masako Hirai, Sisters in Literature: Female Sexuality in ‘Antigone’, ‘Middlemarch’, ‘Howards End’ and ‘Women in Love’ (London: Macmillan, 1998), 8, 25.
 R. A. Jelliffe, ‘An Experiment in Comparative Literature’, College English 9 (1947) 85-87 (p. 86).
 Wayne C. Booth, The Company We Keep: An Ethics of Fiction (Berkeley, Los Angeles, and London: University of California Press, 1988), p. 89.
 Lewis Carroll, The Annotated Alice (Harmondsworth: Penguin, 1970), p. 95.
 F.R. Leavis, D.H. Lawrence, Novelist (London: Chatto and Windus, 1955), pp. 115-16.
 Haun Saussy, ‘Comparative Literature?’, PMLA 1182 336-41 (336, 340).
 Bernheimer, Comparative Literature in the Age of Multiculturalism, p. 12.
 David Ferris, ‘Indiscipline’, in Comparative Literature in the Age of Globalization, pp. 78-99 (p. 91).
 Earl Miner, Comparative Poetics, p. 232, 238.