[In the following essay, Lenaghan traces the textual history of William Caxton's 1484 English translation of Aesop's fables, discusses the popularity of the fable format during the Middle Ages. Lenaghan suggests that Caxton's treatment of the fable anticipates aspects of Renaissance humanism.]

Aesop's fables have been popular from papyrus to television. While our children read them in school, we read Marianne Moore's translations of La Fontaine or Thurber's fables in The New Yorker and thus repeat once more an old pattern, for Roman and medieval schoolboys did their primer fables while their parents read about the Horatian mice or Chauntecleer and Pertelote. The practice of attributing such fables to Aesop is centuries-old, but since for almost as many centuries there has been no firm knowledge about him, the identification has been more legendary than biographical.

Under the impulse of their unquestionable vitality fables proliferated, particularly during the Middle Ages, in a variety of collections which were usually credited to Aesop as the ultimate source but which were quite different from one another in significant ways. Since no fables can be surely identified as Aesop's, his name is best taken as a generic label, and it is more accurate to speak of the Aesopic fables in a given collection than of Aesop's fables. The chief ancient collections are those of Babrius, a hellenized Roman whose Greek fables were probably composed in the latter half of the first century A.D., and of Phaedrus, a Roman writer whose Latin verse fables were composed toward the middle of the same century. Although both these collections were unknown in the West during the Middle Ages, the fables of Phaedrus did circulate in a prose reworking that went under the name Romulus. The fables of the Romulus collection were accepted as Aesop's by a German physician and man of letters, Heinrich Steinhöwel, when he attempted to assemble Aesop's fables in the fifteenth century.

The fifteenth century was an age of transition in northern Europe, and in England William Caxton conveniently embodied some of this transitional division. He was a printer, hence modern, of traditional books, hence medieval. He was born about 1432 and spent most of his adult life as a mercer in Bruges when that city was the seat of the Dukes of Burgundy and an important cultural center. There Caxton developed the literary interests which animated his work as a printer when he returned to England and set up his press at Westminister in 1476, Since Aesopic fables were both popular and traditional, they were a natural choice for early printers, and Caxton was no exception. Steinhöwel's collection was the largest available in the fifteenth century, and it had been translated into French by an Augustinian monk in Lyons and printed there in 1480.¹

Caxton's Aesop

In 1483 William Caxton began to print his translation of the French translation of Steinhöwel's fables:

Here begynneth the book of the subtyl historyes and Fables of Esope whiche were translated out of Frensshe in to Englysshe by wylliam Caxton at westmynstre In the yere of oure Lorde M.CCCC.lxxxiij.

Inasmuch as the colophon dates the completion of the book on March 26, 1484, the second day of Caxton's new year, he must have done most of the printing in 1483.

And here with I fynysshe this book/ translated & emprynted by me William Caxton at westmynstre in thabbey/ And fynysshed thexxvj daye of Marche the yere of oure lord M CCCC lxxxiiij. And the fyrst yere of the regne of kyng Rychard the thyrdde

Three complete or nearly complete copies of this edition are known to survive: a perfect copy in the Royal Library at Windsor; a copy in the British Museum, which lacks the first leaf but has the complete text; and a copy at the Bodleian, which lacks eight leaves. Caxton printed only this one edition, but the London printer, Richard Pynson, printed two more editions, one about 1497 and one about 1500. All these editions are profusely illustrated with woodcuts modeled on those in the French Esope, which were traced copies of the original German woodcuts.

Caxton's Aesop contains 167 fables and tales and a Life of Aesop which is itself a composite of various types of folktales. Although Caxton's collection is basically an English version of the one assembled by Steinhöwel,² its contents do differ slightly from those of its basic source. It does not include two tales which Steinhöwel took from Petrus Alphonsus and one which he took from Poggio; the same tales were omitted by the intermediate French translator and editor, Julien Macho. On the other hand, at the end of the selection from Poggio it does include six tales neither in Steinhöwel nor in the French.

Caxton's edition is divided into seven main parts: (1) the Life of Aesop translated into Latin from the Greek by the Italian humanist, Rinuccio da Castiglione of Arezzo; (2) the four books of the Romulus collection of Aesop's fables; (3) a fifth book of seventeen fables of Aesop which had not been included in the registers of the Romulus collection; (4) seventeen fables of Aesop newly translated by Rinuccio and so also not included in the Romulus collection; (5) a selection of twenty-seven fables of Avianus, a Roman fabulist of about A.D. 400; (6) a selection of thirteen tales from the Disciplina Clericalis of Petrus Alphonsus, a Spanish Jew converted to Christianity in the twelfth century; and (7) a group of thirteen tales, attributed by Caxton to the Italian humanist, Poggio Bracciolini (of these, seven came from Poggio's Facetiae through Steinhöwel and the French text; to them Caxton added four other tales from the Facetiae and two more which are apparently original with him). These components range in date from the late fourth century to the mid-fifteenth; the older ones, having been popular for centuries, were no doubt easily accessible to Steinhöwel, and both Poggio's Facetiae and Rinuccio's fables were in print in the 1470's.

1. The Life of Aesop.³ The Life, as taken from Rinuccio, is a crude, episodic 'biography,' presenting in a series of tales Aesop's rise from slave to royal adviser and concluding with his death at the hands of the jealous Delphians. The earliest version comes from Egypt and dates from the first century A.D.,⁴ but most of the versions known in modern times have been associated with the name of Maximus Planudes, a thirteenth-century Byzantine monk whowas an emissary to Italy. Steinhöwel's version of the Life was formerly thought to derive from the Greek of Planudes, but Professor Perry has shown that the versions of Rinuccio and Planudes descend from different branches of a complex of manuscripts stemming ultimately from an eleventh-century Byzantine reworking of an ancient version of the Life.⁵
2. The Romulus Collection.⁶ These fables make up the largest unit in the collection, accounting for almost half of its total number of fables and tales. The Romulus collection is dated between A.D. 350 and 600 by its editor. The name, Romulus, comes from a prefatory letter in which a writer of that name recommends to his son the collection of Aesop's fables he has translated from Greek into Latin, though in fact the fables are a prose reworking of the verse fables of Phaedrus. Since the fables of Phaedrus survived as the Romulus collection, that collection became Aesop's fables in the Latin West, and medieval references to Aesop's fables usually meant the Romulus fables. The Steinhöwel-Caxton version of these fables, like most other versions of the collection, is divided into four books, but it is the only version that is evenly divided into four books of twenty fables each. This unique symmetry produces some peculiarities of omission (the boy and the scorpion, Aesopica, 199; and the thirsty crow, later included as the twentieth fable in the selection from Avianus) and addition (the pine tree and the reed, Romulus IV, 20).
3. The Fifth Book.⁷ This book is made up of seventeen fables ascribed to Aesop. Caxton, following the French translator, says they are Aesop's despite the fact that they were not included in any of the registers for the books of the Romulus collection. Steinhöwel calls them extravagantes antique because they were old (antique) and because they were not fixed in the books of the standard Romulus collection of Aesop's fables (extravagantes). Eleven of the seventeen do, however, appear in a collection contained in a fifteenth-century manuscript in Munich along with twenty-eight fables from the standard Romulus collection. Inclusion of these fables in another Romulus collection would presumably have been a satisfactory warrant for Steinhöwel to attribute them to Aesop as extravagantes antique. No doubt these eleven fables had been 'extravagant' for some time: five of them appear, wholly or in part, in the twelfth-century collection of Marie de France, and, of course, separate episodes and motifs can be found in versions antedating those of the Munich manuscript (e.g., no. 16 and no. 86 of the Dialogus Creaturarum). Although Steinhöwel designated all seventeen fables as extravagantes, six of them appear for the first time in his collection and therefore present rather more of a problem than the other eleven. There seems no way of deciding whether Steinhöwel was following a collection now lost, which contained all seventeen extravagantes, or whether he simply added to those of the Munich manuscript six more fables and tales which he thought were also extravagantes. He does say at the end of thebook that he is not certain whether these fables are Aesop's, but because the French translator omits this passage, no such doubt appears in Caxton's text, and he attaches the extravagantes as an additional fifth book to the traditional four of the Romulus collection.
4. The Selection from Rinuccio's Fables.⁸ This grouping of seventeen fables comes from the collection which Rinuccio translated from the Greek in 1448 and which was accompanied by his Life of Aesop. Rinuccio asserts that he has translated only those fables he chanced to find and that he makes no claims to completeness; nevertheless, his collection is large, numbering one hundred fables. Some of them are also in the Romulus collection, and since Steinhöwel had already included these fables of Aesop in his collection, there was no need to duplicate them from Rinuccio. However, there are also fables in Rinuccio's collection which do not appear elsewhere in Steinhöwel and which he did not include in his selection from Rinuccio. This is the first suggestion that Steinhöwel would be satisfied with something less than completeness. Of course, aside from reasons of practicality and convenience, it is quite possible that Steinhöwel may have rated Rinuccio's authority less than that of tradition (he speaks of the new translation) and so have felt freer to exercise a choice about what was to go into his collection. In this case the French translator and Caxton repeat Steinhöwel's editorial remarks and inform their readers that these fables are Aesop's, but from a new translation instead of the old one made by Romulus; therefore, the selection is not given a book number like the fables of Romulus or designated by the name of the author like the non-Aesopic sections of the collection.
5. The Selection from the Fables of Avianus.⁹ This selection of twenty-seven fables is taken from the collection of Avianus, which is a reworking in distichs of forty-two fables from the Greek of Babrius and dates from about 400. Of the fifteen fables which do not appear in this section of the Steinhöwel-Caxton collection, only one appears elsewhere.
6. The Selection from the Disciplina Clericalis of Petrus Alphonsus.'¹⁰ Caxton's selection numbers thirteen fables (really, twelve tales and one fable, Al. 9), twelve of which are taken from the thirty-four tales of the Disciplina Clericalis. Steinhöwel may have added the additional tale (Al. 12) to his selection, or he may have used a variant text which included this tale. The Disciplina Clericalis is the oldest medieval collection of 'Eastern' tales and was written in the early twelfth century by Petrus Alphonsus.
7. The Selection from the Facetiae of Poggio Bracciolini.¹¹ The Facetiae are a collection of frivolous stories assembled sometime after 1450 for the amusement of Poggio's colleagues of the papal Curia. The selection which Caxton prints as 'fables of Poge the Florentyn' is made up of thirteen fables and tales. Seven of these tales come straight from the intermediate French edition where Julien Macho reduced Steinhöwel's selection of eight by one. Four additional tales come from the Facetiae independently of Steinhöwel and Macho, and the last two tales are apparently Caxton's own. These last six tales form a textual unit in that they follow one another without separate numbering, the only pieces in the entire collection which are so presented. Following the pattern set for the first seven tales of the selection, Caxton credits Poggio with the first four tales of the additional unit, the ones taken independently from the Facetiae. He then introduces his own tale of the widow's reply (Poggio no. 12) impersonally and he introduces himself at the beginning of his last tale—of the good, simple priest (Poggio no. 13), 'Now thenne I wylle fynysshe alle these fables wyth this tale that foloweth whiche a worshipful preest and a parsone told me late.'

It is something of a question how Caxton, who followed the French translation of Steinhöwel's collection so faithfully, obtained his four additional tales from the Facetiae independently of the French translator and of Steinhöwel. Robert H. Wilson, on the strength of the appearance of one of the four in a French Esope of 1532, suggests a common source for the French collection and Caxton.¹² Brunet does mention an early French translation of some of the Facetiae and cites an attribution to Julien Macho,¹³ but I have been unable to learn anything more about the book. If this book existed in the 1480's, it could have been Mr. Wilson's hypothesized source. It is also quite possible, however, that Caxton (or someone else) could have taken the tales directly from the Facetiae and appended them just as Steinhöwel might have appended his extravagantes to the fifth book of fables. While the tale of the widow's reply (Poggio no. 12) could have been associated with Poggio's collection before Caxton made his additions to the selection from it, it could also have been his own rather more sober variation of the usually frivolous 'reply' as represenited elsewhere in the Facetiae. As for the last tale, I see no reason to doubt Caxton's attribution to a 'worshipful preest.' Whatever the source of these last two tales or the details of their transmission, their overt morality has the effect of returning the reader to a serious atmosphere of instruction, and by virtue of their strategic final position they do much to shape an impression of the book.

Caxton's Aesop is shaped, then, in the pattern Steinhöwel gave his original collection. At the center is the Romulus collection—Aesop's fables—to which the Life is an obviously suitable preface. The extravagantes antique and the fables from the new translation of Rinuccio, though not perhaps of equal authority, supplement the basic collection. The selection from the fables of Avianus, though they were never supposed to be Aesop's, was an appropriate addition to the collection because these fables had long been paired with the fables of Romulus as school texts and thetraditional link between them must have been very strong. The selections from the Disciplina Clericalis of Petrus Alphonsus and Poggio's Facetiae are less clearly appropriate, and some of Steinhöwel's editorial remarks show that he felt this diminished suitability. Though Caxton does not make explicit any such editorial doubts, his reshaping of the last part of the final selection implies that though he wanted to entertain his reader, the traditional literary function of instruction was still an active principle in his work. It is the special quality of the fable that it is the most elementary combination of the two functions.

The Aesopic Fable in the Middle Ages

Perhaps just because the fable is so rudimentary a combination of the basic literary functions, it was an extremely popular literary form during the Middle Ages, and perhaps for the same reason fables were freely altered and recast throughout the period. Consequently the history of the medieval fable is a complicated one and any survey must be either very short, a general designation of the main sources, or very long, an analysis of the intricacies of relationship among the various collections. A brief sketch should be sufficient to locate Caxton's Aesop against the general medieval background.

There are two main source collections for the medieval fable, both of late Latin origin: the prose fables of Romulus and the verse fables of Avianus. Those of Avianus were selected and translated from the Greek fables of Babrius. Those of Romulus are prose reworkings of the 'standard' ancient Latin collection, the fables of Phaedrus, which were unknown in their original form through most of the Middle Ages. The Romulus fables became the most important Aesopic collection during the Middle Ages, in part perhaps because they were derived from the 'standard' antique collection, in part because they were more numerous than the fables of Avianus, but chiefly because of the prefatory letter which claimed that they were Aesop's fables translated directly from the Greek. Whatever the cause, the Romulus fables were taken as Aesop's fables during the Middle Ages. They were translated, versified, recast with new or additional morals, and selected for smaller collections. The fables of Avianus underwent the same process. The resulting tangle of versions and collections can be most easily appreciated by scanning the lists and tables of Leopold Hervieux's Les Fabulistes latius. Accretion to so varying a corpus was almost inevitable. The process of this accretion and, indeed, the whole tangle of the medieval fable are conveniently ordered in Ben Edwin Perry's Aesopica where the development of the Aesopic corpus is clearly laid out.

Caxton's Aesop can be seen as a historical anthology of the medieval Aesopic fable. The Life establishes an identity for the legendary author of the fables. Then come Aesop's fablesthemselves, in the Romulus collection, some medieval additions to that corpus, and the variant, parallel collection of Avianus. The tales of the Disciplina Clericalis serve as a reminder that the fable was often combined with other sorts of didactic literature and was often thought of more as one kind of moral narrative than as an independent literary genre. The Aesop also implies a rounding off of this medieval history by inclusion of the work of the fifteenth-century Italian humanists, Rinuccio and Poggio Bracciolini. Rinuccio's translation is newly made from Greek sources and is an early sign of the re-established contact with the Greek world. Although Poggio's tales are usually unlike the typical Romulus fable in several ways, the novel difference is implicit in his title, and explicit in his preface where he says that entertainment and style are his only concerns; the tales are not concerned with morality. That the new mixes so easily with the old testifies both to the vitality and the elasticity of the fable as a genre.

The vitality of the genre is an obvious incentive to the collector but its elasticity is just as obviously a problem for him. He has to decide what a fable is before he can assemble a collection. Ancient and medieval authorities defined the fable as a fiction made up to represent some aspect of human activity.¹⁴ This amounts to saying that the fable is a fundamental form of literature. The definition sets out the combination of the traditional literary elements—story and moral—and the concomitant aims—amusement and instruction. Caxton's Aesop is true to this tradition; it is an unsophisticated collection of literary elements unabashedly confident in the worth of those elements. Because, however, their combination in the fable is so rudimentary, it is difficult to say which of the elements is more important, and so the genre carries about it an ambiguity of intention that makes it difficult to declare the actual primacy of one or the other of the two aims. Caxton is not especially sensitive to this ambiguity, but the original collector, Heinrich Steinhöwel, was much more conscious of the problem. He apologizes for the unseemly tales of Poggio, and his preface insists that fables are valuable for their morals. There can be no misunderstanding Steinhöwel's conception of the fable's purpose: it is didactic. He provides an index to help to realize that aim by arranging appropriate fables under topical headings like Aigensinnig, Ayd, Alter.¹⁵

When Steinhöwel turns from the fable's function to its nature, or make-up, his discussion becomes more confusing. He borrows a definition from the medieval encyclopedist, Isidore of Seville, which is really a combination of definitions. He starts with etymology: 'The poets have taken the name fable from the Latin word fando, that is to say in German that fables are not about things which actually happened but are only verbal inventions, and they have been composed so that through the invented words of unreasoning animals lower than himself a man may recognize an image of the ways and habits of human virtue.¹⁶ He then defines by description and division of fables according to their characters. He finally defines by function: some fables are composed to give pleasure, some to represent human nature for didactic ends, and some to give allegorical explanations of natural phenomena (Steinhöwels Asop, p. 5). Steinhöwel, Isidore, and most twentieth-century readers would agree that a fable is a moralized animal tale. But in picking up Isidore's etymology, Steinhöwel emphasizes what medieval authorities considered the essential feature of the fable: it is 'made up'; it narrates events that did not happen. Steinhöwel's borrowed definition is really a combination of at least four different, though not mutually exclusive, definitions: (1) it is fiction in the sense that it did not really happen; (2) it is literary entertainment; (3) it is poetic fiction with double or allegorical significance; and (4) it is a moral tale, usually with animal characters. The first seems to be the essential definition but it is very general. The fourth is most familiar to modern readers, and if Isidore's examples are indicative, it seems to have been uppermost in his mind. Since the fourth is easily contained in the first, it is possible that fable may often refer specifically to a moralized animal tale even when the context stresses its more general definition as fiction. These two principal definitions are ancient and medieval commonplaces, one from rhetoric and one from grammar. The rhetorician defined fabula as fiction, that type of narration which describes events that could not have happened. The grammar teacher stressed the utility of the fable as a simple, instructive tale. Although Steinhöwel's definition of the fable only confuses his discussion, tacitly admitting the ambiguity of the genre, he does come down squarely on the side of the grammar teacher in his insistence on the morality of the fable.

The grammar teacher, probably more than anyone else, was responsible for the medieval currency of the Aesopic fable. He was actively aware of the reasons for the durability of the genre. The suitability of the fable as primer material is obvious. The lists of school authors are seldom without an Aesopic collection and the same is true of certain earlier manuscripts which seem to have served as primer texts in the ninth and tenth centuries. The fable was also the vehicle of the first of a series of paraphrase exercises in composition which are described in Priscian's Praeexercitamina.¹⁷ Though the series probably did not survive intact through the Middle Ages, there is ample evidence that fables did serve as paraphrase vehicles for teaching composition. This pedagogical utility is due to the simplicity that so strongly characterizes the genre; yet that very simplicity, enforced by association with the schoolroom, makes the fable especially suitable for a special kind of sophisticated irony that subordinates instruction to amusement. By virtue of the naïveté associated with the genre, a sophisticated fabulist can take the ironic stance of Socratic mock innocence. Chaucer's Nun's Priest's Tale is a fine single example of this sort of fable, and the Reynard tales are often expanded developments of the same kind of irony.

The conclusion to be drawn, then, is that in the fifteenth century, as for many centuries before, the fable was a literary genre, most obviously a simple one, but also possessed of a basic ambiguity sufficient to invite sophisticated, ironic manipulation.

Because of its generic ambiguity, then, the fable was many things in the fifteenth century: folktale, pedagogical device, sermon exemplum, literary genre; it could reach a wide range of intelligences; it could be bluntly assertive or cleverly ironic; didactic or skeptical. It is no wonder that the definition Steinhöwel appropriated from Isidore turns out to be a rather confusing combination of definitions. A man may emphasize any feature or function he chooses. Steinhöwel chose with the grammarians to emphasize the didactic function, but by quoting Isidore, he noticed almost all the others. Texts of such various appeal, sanctioned by so long a tradition, would naturally suggest themselves for printing. Practical financial questions aside, the problem would have been to make a selection from the welter of texts circulating in the fifteenth century. Heinrich Steinhöwel, Italian-educated and prominent both because of his social position and his literary activities, had the authority to make that choice. His decision was quite simple: he made a collection of collections, and Johann Zainer printed the lavish German and Latin edition of 1476/77.

The Esopus was widely reprinted; it was translated into French by Julien Macho of Lyons and printed there in 1480. Caxton translated the French text, and so within six years of its first printing Steinhöwel's collection passed through an intermediate French edition into Caxton's English.

The Editions of Steinhöwel and Julien Macho

In one sense, description and discussion of Caxton's collection make superfluous any discussion of its originals, Steinhöwel's collection and its French translation, because much of what should be said of them has already been said. Still, Caxton's collection does differ in certain respects from its originals, and a full view of his work is impossible without a look at them and the process of their transference into his collection.

Steinhöwel's Esopus was an unmistakably lavish production, offering a profusion of woodcuts, more fables than any earlier edition, a bilingual text, and, for the first three books of fables, still another Latin version in distichs. The pattern of the collection indicates that Steinhöwel was a self-conscious, and, to a degree, a discriminating collector ('completed are the extravagantes antique, ascribed to Aesop, I do not know whether truly or falsely¹⁸). Steinhöwel included, either complete or in part, almost all the major collections of fables popular in the Middle Ages. His printer undertook to produce a book worthy of such a collection, and perhaps the best measure of this is the presentation of the first three books of Romulus in three parallel versions: the Latinprose of the Romulus collection itself, the German translation of it, and a set of Latin verse fables.

Of these three versions, the Latin prose of the Romulus collection is basic, and Steinhöwel's German version tends to follow it closely with some expansion.

Verum ut vitam hominum et mores ostenderet, inducit aves, arbores, bestias et pecora loquentes, pro vana cuiuslibet fabula, ut noverint homines, fabularum cur sit inventum genus, aperte et breviter narravit. Apposuitque vera malis. Composuit integra bonis. Scripsit calumnias malorum, argumenta improborum.

Aber darumb daz er das leben der menschen und iere sitten erzögen möchte, hat er in syn fabeln redend fogel, böm, wilde und zäme tier, hirs, wolf, füchs, löwen, rinder, schauff, gaiss und andre gezogen, nach gebürlikait ainer ieden fabel, daruss man lycht und verstentlich kennen mag, warunb die gewonhait in fabeln ze schryben sye erfunden. Er hat die warhait zuo den bösen geseczet, das guot zuo guoten. Er beschrybt die böslist der untrüwen ankläger der gericht, und erdichte fürzüg der unfrummen.¹⁹

The translation is close, 'translated from Latin into German by Doctor Heinrich Steinhöwel not well but clearly, taken not word for word, but meaning for meaning, to the greater clarification of the text often with a few added or deleted words.'²⁰ The most obvious change is the expansion of the dramatis bestiae, but also noteworthy are the change of vana fabula to syn fabel and the reminder in nach geburlikait ainer ieden fabel that there are traditional expectations for certain animals (the lion is a king; the fox is clever; the wolf is violent).

The third of the parallel versions—the Latin verse fables—derives from the basic Romulus collection as a poetic version of it, but it has more the status of an independent collection. These distichs date from the twelfth century and are attributed to an unknown Walterius. They were already extremely popular; Hervieux lists 129 Latin and vernacular manuscripts of it. The fifteenth-century Italian translations by Zuccho and Tuppo seem to have been more popular than the Latin translations made from the Greek by Lorenzo Valla and Rinuccio. To judge from the number of editions, however, the most popular fifteenth-century verse collection was the Esopus Moralisatus, which was nothing more than Walter's collection with the superaddition (by the thirteenth century) of a set of moralizations appended to the epimythia of the fables. Thus, by including the fables of Walter, Steinhöwel in effect appropriated the major rival collection of the fifteenth century.

Though Steinhöwel's sources are all identified, the specific texts he used have not been established. In any event, Steinhöwel would have felt free to exercise his own judgment in preparing his texts. For example, though Rinuccio says that fables are important fortheir wisdom and although his Argumentum fabularum describes an explicitly moral wisdom, in his prologue it is a rather worldly wisdom. Steinhöwel resisted the lead suggested by Rinuccio and poured the 'new' translation into the old moral bottles so that it became formally indistinguishable from the old. He added Latin and German promythia, the initial, generalized statements of the morals, to all except three of his selections from Rinuccio's fables (and he managed a German promythium for one of the three). The effect, though he may have intended nothing more than to produce a formal similarity, is certainly to underscore the morals. Similar literary and didactic standards can also be seen in his treatment of his selection from Poggio. After a tale of unquestionable impropriety (Poggio no. 4), he writes a lengthy apology: he included such stories not because he liked them but because of his respect for the famous Poggio, and when he thinks of how he has fallen from his high purposes to such stories, he sees that he must have done with them. To prove his sincerity he devotes a page to describing, occasionally in detail, the stories he will not tell. He does, however, conclude the book with four harmless tales, the last being a fable of a fox and a cock (Poggio no. 7). Needless to say, Steinhöwel is hard put to find edifying morals for improper material, but it is testimony to his determination that he supplies promythia for three of the eight pieces. These promythia are more interesting for what they reveal about Steinhöwel's sense of genre than for any applicability of moral to tale. The concluding fable he introduces, 'Concerning this, hear a fable,' but each of the two tales (Poggio no. 6 and one which does not appear in Caxton) he introduces, 'Concerning this, hear an amusing tale.' This kind of apology and this kind of discrimination are editorial manifestations of Steinhöwel's interest in the questions of the nature and function of the fable, and they are precisely what, for lack of interest or for lack of space, are omitted from Julien Macho's translation and therefore from Caxton's.

Steinhöwel's collection was reprinted several times immediately after the first edition. The German text alone was reprinted by Gunther Zainer in Augsburg in 1478 and twice more in Augsburg (1479 and 1480) by Anton Sorg, who also reprinted the Latin text alone in 1480. The French translation was made by an Augustinian monk of Lyons, Julien Macho, an obscure figure about whom little is known aside from the fact that he translated and edited a number of books for Lyonnais printers.²¹ The more or less automatic expectation, particularly in view of his other translations, is that he would have made the French translation from a Latin text. This would seem to be borne out by a comparison of some sample passages.

Quidam ex scolaribus percipiens Xanthum vino paulisper gravatum ait: Dic mihi, preceptor, homo unus mare totum bibere posset? Quid ni, Xanthus ait, nam egoipse totum ebiberem. Et scolaris inquit: Et si non biberis, quid deponis? Domum, Xanthus air (Steinhöwels Äsop, p. 23).

Et adoncques vng des escolliers voyant que xantus estoyt bien devin charge. Mon maistre ie te demade si vng homme pourroit boire tout la mer. Et pourquoy non dist xantus moy mesme le la boyray toute.

Et lescollier luy dist. et si tu ne la boys que veulx tu perdre.

Et xantus dist ma maison.²²

Furthermore, there are occasional mistakes which would suggest that the translator had his eye on the Latin text: 'vng prestre qui auoit nom Isidis,' 'ecce Ysidis sacerdos,' 'ain prester der gottin Ysidis'; or 'lesquelz seruiteurs auoient nom Grammaticus Saltes & esope,' 'hi fuerunt grammaticus, psaltes atque Esopus,' 'die warent ain grammaticus, ein harpfer und Esopus.'²³

On the other hand, there are passages which suggest that the translator had his eye on the German text:

de grece au pres de troye la grande dune ville appelle amonneo, is … natione Phrygius, ex Ammonio Phrygie pago fuit, der gegent Phrigia, dar inn Troya gelegen ist, von Ammonio dem wyler geboren.²⁴

Steinhöwel offers a full Latin version of the fable of Venus and the hen (Rom. III, 8) but stops his German version short, saying he will not put what the hen said into German. The French translator follows the German version, 'We shall leave it in Latin,' even though the Latin he leaves it in is not in his book (see note 239; p. 110). The promythia of the fables of Avianus also indicate that the translator sometimes followed the German text. There are no promythia for some of the fables in this collection, but Steinhöwel, perhaps because of an understandable reluctance to imitate the poetry of Avianus, did not supply the missing Latin promythia, as he did for the fables of Rinuccio. Instead he contented himself with supplying German promythia, which some of the French promythia very much resemble. Although Steinhöwel no doubt modeled some of his German promythia on the corresponding Latin epimythia, the final, generalized statement of a fable's moral, and the French translator could easily have done the same (e.g. Avianus no. 9), there are other cases where it is unlikely that the French translator followed Steinhöwel's Latin (Avianus no. 3), or impossible (Avianus no. 8). It therefore seems clear that Julien Macho made some use of the German text.

If a comparison of the texts suggests that the French translation follows both the German and Latin texts, the obvious original for the translation. would seem to be Steinhöwel's first bilingual edition printed at Ulm. Unfortunately for this conclusion, there are woodcuts in the French edition of 1480 which are not in Steinhöwel's first edition but do appear in the subsequent editions printed at Augsburg.²⁵ Since none of these editions was bilingual, Julien Macho probably worked from the bilingual first edition, and the designer of the woodcuts had one of the Augsburg editions.

The French translation is a much reduced version of the lavish Steinhöwel first edition. It is not bilingual, and gone with the Latin texts are three of Steinhöwel's tales (Alphonsus nos. 13, 14 and Poggio no. 1). Also dropped is Steinhöwel's editorial apparatus: his introduction about the nature and function of the fable, his topical index of morals, and his apology for Poggio's tales. Apparently, Julien Macho was less concerned than Steinhöwel to define the fable as a literary genre and saw less need to stress the didactic function of that genre. He does, however, make his position clear by adding to the preface to the second book of Romulus some remarks on the function and value of fables (which reappear in Caxton's edition, p. 89):

Every fable is invented to show men what they ought to follow and what they ought to flee. For fables mean as much in poetry as words in theology. And so I shall write fables to show the ways of good men.²⁶

Still, the omission, for whatever reason, of so much of Steinhöwel's express emphasis on the morality of the fable does have the effect of leaving the reader free to read with as little attention to morality as the fables themselves permit. The result is that the French translation concentrates somewhat less on instruction than does Steinhöwel's collection, and so permits relatively more attention to entertainment.

Caxton's Prose

Caxton's dependence on Steinhöwel is counterbalanced less by his own independence of judgment than by his dependence on the French translator. This is not to minimize the independence implicit in his additions to the collection but simply to state the obvious—Caxton could only see Steinhöwel's collection as the French translator chose to show it to him. The pattern of Caxton's Aesop reveals the larger shape of this dependence, but the actual processes of translation best reveal its particulars, as can be seen by comparing passages from Caxton's edition with passages from the French edition of 1480.

Des mocqueurs esope a fait vne telle fable dung asne qui recontra vng lyon. Et lasne luy dist. Monfrere dieu te gart. Et le lyon commence a [c 8] branler la teste par grant ayr et a grant paine peust il refraindre son yre que de ses dens ne le deuourast. Et Adoncques dist en soy mesmes il nappertient pas que vne dens si noble touche a vne beste si ville. Car celluy qui est saige ne doit blesser le fol auoir cure de ses parolles mais le fault laisser aller pour tel quil est.

Of them whiche mocken other esope reherceth such a fable Ther was an asse which met with a lyon to whom he said my broder god saue the & the lyon shaked his hede [e 4^v] and with grete payne he myght hold his courage/to haue forth with deuoured hym/But the lyon sayd to hym self/It behoueth not that teethe soo noble and sofayre as myn be touchen not/ne byten suche a fowle beest/For he that is wyse must not hurte the foole ne take hede to his wordes/but lete hym go for suche as he is (Rom. I, II)

The most important point is again the obvious one: Caxton follows the French closely, word for word at times. His changes are usually minor: for example, the addition of 'ne byten,' the omission of par grant ayr, the alteration of dung asne to 'Ther was an asse,' the expansion of Des Mocqueurs to 'Of them whiche mocken other.' There are occasional difficulties: his infinitive is an awkward rendering of que de ses dens ne le deuourast. As might be expected in so faithful a translation, Caxton anglicizes and directly adopts many of the original French words. But Caxton's syntax is clearly English. Occasionally difficulties arise from a lapse into French patterns (e.g. 'took of the most best metes,' 39:24), from the reproduction of difficulties in the French original, and from requirements made by our English but not by Caxton's (e.g. pronouns, 102:21-24; omitted subject, 172:1, and levels of discourse, 91:12-14). In short, Caxton wanted to render the fables directly into English and on the whole he did his work efficiently.

He made some changes, and the most important of these—the additions to the selection of tales from Poggio—have already been discussed. There are also two fairly long omissions: a passage in the Life about urinating (see note 43; p. 35), and another of heavy-handed quibbling (see note 50; p. 39). Caxton might seem to be acting from a moral or aesthetic sensitivity, especially since there is a comparable graphic omission in the deletion from his frontispiece of the striking French example of hindsight (see Caxton's frontispiece in appendix); but there are too many indelicate episodes which are not slighted to admit of generalizations about Caxton's delicacy. Similarly, in a tale of cuckoldry (Poggio no. 1) he changed French dieu to the Holy Ghost, but whether this tempers or refines the irreligion is hard to say. It is probably safest to regard these changes as fortuitous.

It is possible, however, to make some general inferences about Caxton's handling of his text. To start negatively, Caxton is often neglectful of narrative precision or nicety. He sometimes abridges the French narrative and drops out words or phrases on which a later statement depends. For example, the country mouse gave his urbane cousin 'of such mete as he had,' instead of 'some grains of wheat and some water to drink' which better anticipates the conclusion, 'I had leuer ete some come in the feldes' (see note 160; p. 81). There are other such places (e.g. note 78; p. 53), places where his narrative is less tidy (e.g. note 349; p. 152), and places where he changes the identity of an animal in spite of a clearly corrective woodcut (e.g. Aesop before Croesus in the Life, or Romulus IV, 14). That such blemishes are more matters of carelessness than ineptitude is clear from the direct and effective narration of Caxton's own concluding tale (Poggio no. 13).

If such careless omission and alteration show that narrative precision was not a major interest, a positive interest in style may be inferred from one kind of addition Caxton made to his text. A casual reading shows that he is not always as direct as he is in his final tale; indeed he often swells up his prose to a remarkable prolixity (e.g. 91:6-8). This tendency suggests an interest in elevation of style, but the focused augmentation of doubled phrases is a safer basis for inference than the more general notion of prolixity. Caxton obviously wants to use two words where one would suffice. This doubling was common enough in his day and, on the sample of the first half of his book, I found that thirty per cent of Caxton's doublets are simply translations of doublets in the French text. Of the remainder, slightly more than half are formed by adding an English word to the French ('doubte ne drede' from 'doubter,' 'tryst and sorowful' from 'tryste'). A glance at almost any page will provide further examples of these counterpoised doublets. They are best understood in the light of the dilemma Caxton describes in the prologue to his Eneydos; he was attracted to the book by 'the fayr and honest termes and wordes in frenshe' and he had been urged by great clerks 'to wryte the moste curyous termes,' but others wanted him to stick to 'olde and homely termes.'²⁷ Caxton's response is often to do both, or at least to link an English word with a French one. As a result, some doublets did have the effect of 'augmenting' the language, and in some cases this effect seems to have been intentional (see notes 239; p. 109 and 190; p. 93). In other doublets, however, there are no French words, or else the French word had long been naturalized and could not be considered 'curyous.' Caxton's enrichment of his native language was occasional and unsystematic; it came naturally in the effort to elevate his style. Professor N. F. Blake has remarked that doublets in Caxton's Reynard the Fox come most frequently in didactic or descriptive passages and in beginnings and endings rather than in the course of the narrative.²⁸ This is less clearly true of the fables, where there are a good many doublets added in the narrative sections; but in some fables the frequency of doublets goes up quite markedly in the promythia and epimythia. These are statements of sentence, and the doublets are the straightforward means of lending them dignity.

Caxton's prose, especially the doublet, is a guide to the nature of his book. His method of elevating his style was consistent with contemporary Burgundian literary values and was thus old-fashioned or medieval. His famous prologue to his Eneydos was similarly old-fashioned, but its deference showed a receptivity, a willingness to be taught, and further suggested that this willingness was proper for gentlemen concerned with humane studies. In other words Caxton's general concern with style, though it was old-fashioned and Burgundian in its particulars, put him in the humanistic channel through which the new instruction was eventually to come. In his Aesop Caxton actually had harbingers of the new age, the humanists, Rinuccio and Poggio, but there is no sign that he had any practical recognition of this fact. Steinhöwel, of course, did have some idea of the values of the Italian humanistsand his Esopus, with its pedagogical associations, is a partial and elementary reflection of those values. By eliminating the Latin texts of the Esopus, Julien Macho eliminated the explicit signs of that reflection and so Caxton stands almost as far from Steinhöwel as his woodcuts are from their German originals. Yet Caxton's treatment of his text betrays some apprehension of the general values which animated Steinhöwel's collection and which, because of the traditional link between the Aesopic fable and elementary grammatical study, were still implicit even in the reduced version that Caxton translated and printed.

Notes

¹ Ben Edwin Perry's Aesopica (Urbana, 1952) is the authoritative text but his volume in the Loeb Classical Library, Babrius and Phaedrus (Cambridge, Mass., 1964), is a more convenient introduction to the Aesopic fable. For Caxton, Nellie Slayton Aurner's Caxton: Mirror of Fifteenth Century Letters (London, 1926) is a good general introduction, as is, more briefly, the relevant part of the Oxford History of English Literature: H. S. Bennett, Chaucer and the Fifteenth Century (Oxford, 1947), pp. 203-213.

²Steinhöwels Asop, ed. Hermann Osterley (Tubingen, 1873).

³ There is no modern edition of Rinuccio's Life and Fables. The incunabula are listed in Der Gesamtkatalog der Wiegendrucke, nos. 335-344.

⁴ Ben Edwin Perry, Aesopica, p. 5; Babrius and Phaedrus, p. xlvi.

⁵Aesopica, pp. 22, 28.

⁶Der Lateinische Äsop des Romulus und die Prosafassungen des Phadrus, ed. Georg Thiele (Heidelberg, 1910).

⁷ The closest thing to a source text for these fables is Monachii Romulleae et Extravagantes Fabulae, ed. Leopold Hervieux, Les Fabulistes Latins, 2nd ed. (Paris, 1894), II, 262-290, and Aesopica, pp. 696-704.

⁸Der Gesamtkatalog der Wiegendrucke, nos. 335-344.

⁹ Avianus, Fables, ed. Robinson Ellis (Oxford, 1887); the fables of Avianus are also included in a volume of the Loeb Classical Library, Minor Latin Poets, ed. J. Wright Duff and Arnold M. Duff (Cambridge, Mass., 1935).

¹⁰ Petrus Alfonsus, Disciplina Clericalis, ed. Alfons Hilka and Werner Soderhjelm (Helsinki, 1911).

¹¹ Poggio Bracciolini, Facetiae, ed. anon. (Paris, 1879).

¹² 'The Poggiana in Caxton's Esope,' The Philological Quarterly, 30:350 (1951).

¹³ Gustave Brunet, La France Litteraire au XVe Siecle (Paris, 1865), p. 165.

¹⁴ Ben Edwin Perry's 'Fable,' Studium Generale, 12:19-23 (1959), is an authoritative modern definition.

¹⁵ 'Obstinate, Oath, Old Age.'

¹⁶ 'Die poeten den namen fabel von dem latinischen wort fando habent genommen, daz ist ze tiitsch reden, wann fabel synt nit geschechene ding, sonder allain mit worten erdichte ding, und sint darumb erdacht worden, daz man durch erdichte wort der unverniinftigen tier under in selber ain ynbildung des wesens und sitten der menschlichen wiirde erkennet' (Steinhöwels Äsop, p. 5).

¹⁷ Priscianus, Praeexercitamina, ed. Heinrich Keil in Grammatici Latini (Leipzig, 1858), II, 430.

¹⁸ 'Finite sunt extravagantes antique, ascripte Esope, nescio si vere vel ficte' (Steinhöwels Asop, p. 241).

¹⁹'But because he wanted to instruct the life of men and their customs he has drawn into his fables speaking birds, trees, wild and tame beasts, stags, wolves, foxes, lions, oxen, sheep, goats and others, according to the requirements of each fable, from this one may easily and clearly recognize why the custom of writing fables was invented. He has told the truth to the wicked, good to the good. He describes the wickedness of the false accusers of the just and sets down the progress of the ungodly' (Steinhöwels Äsop, p. 78).

²⁰ 'uss latin von doctore Hainrico Stainhowel schlecht und verstentlich getiutschet, nit wort uss wort, sunder sin uss sin, um merer lutrung wegen des textes oft mit wenig zugelegten oder abgebrochnen worten gezogen' (Steinhöwels Äsop, p. 4).

²¹ J. B. Wadsworth, Lyons, 1473-1503 (Cambridge, Mass., 1962), pp. 22-25.

²² And then one of the scholars, seeing that Xantus was full of wine, said, 'Master, I ask you if a man can drink all the sea?' 'And why not?' said Xantus, 'I shall drink it all myself.' And the scholar said to him, 'And if you do not drink it all, what will you forfeit?' And Xantus said, 'My house.'

²³ 'a priest who was named Isidis'; the French translator apparentlymissed the genitive, Ysidis, and read it in apposition with sacerdos, a reading impossible from the German; 'the which servants were named Grammaticus, Saltes and Aesop'; the translator read grammaticus and psaltes as proper names like Esopus, again a reading impossible from the German.

²⁴ 'from Greece in the country of Troy the great of a town called Amonneo.' The Latin locates Ammonio in Phrygia; the German joins it with Troy and the French drops Phrygia.

²⁵ Claude Dalbanne and E. Droz, 'Etude sur l'Illustration des Fables,' Les Subtiles Fables d'Esope (Lyons, 1926), p. 160.

²⁶ Toute fable est trouure pour demonstrer aux hommes quelle chose ilz doyuent ensuyure et quelle chose ilz doyuent fouyr. Car autant veult dire fable en poesie comme parolles en theologie. Et pource iescripray fables pour monstrer les meurs des bons hommes.

²⁷Caxton's Eneydos, ed. W. T. Culley and F. J. Furnivall, Early English Text Society, Extra Series 57 (London, 1890), pp. 1-3.

²⁸ 'William Caxton's Reynard the Fox and his Dutch Original,' Bulletin of the John Rylands Library, 46:320-321 (1964).