The prestige of the classical tradition among medievalists is odd, and oddly persistent. It represents anxious deference to an antecedent culture, an anxiety we share, perhaps, with some (but by no means all) medieval writers and thinkers; but more remarkably, we all share it with each other across bounds of discipline and even ideology. A book need not be specially old-fashioned, nor specially new-fashioned, to be marked by the gestures, themes, and methods of the medievalist's reverence for the classical past.

That reverence is more remarkable because the medievalist's classical tradition is often something very different from the classicist's. Indeed, the classicist reading his medievalist colleagues' versions of antiquity often finds himself in a time warp, where the most acute and distinctively contemporary of critical gestures are made against a background of classical furniture that has grown noticeably musty around the edges.

Remarks such as these go directly to the heart of the book here under review. Rooted in a specific text, Chaucer's House of Fame, it identifies a salient theme so far inadequately explored, and the theme itself, not the text on the table, becomes the obsession of the author.

The book is bipartite: in the first part an exposition, drawing upon some important earlier work by others on labyrinths but adding fresh new material and insight as well, of the history of the labyrinth theme from Roman antiquity through the middle ages; in the second, extended readings proposed of the Aeneid, the Consolation of Philosophy, the Divine Comedy, and the House of Fame, each centered on the re-reading facilitated by the explorations in labyrinth themes of the first part.

The account of the history of the labyrinth theme is slightly weakened by the medievalist's choice of beginning: not Greek but Roman antiquity, and even there omitting Catullus 64 (which Doob remarks as `the most intriguing classical version of the myth' but omits as `irrelevant to the medieval tradition' [20n4]). The emphasis is on literary texts and discussions, but archaeology and artistic representations are far from neglected. The author has looked afresh at manuscript illuminations and added a few items to the standard catalogue of labyrinth representations. Two points of great interest emerge from the survey.

First, a marvelous ambiguity. It should be self-evident, but most will need to have it pointed out, that there are two ways in which a maze can be constructed or represented: first, as a multicursal, entangling structure, where even the cunning easily lose their way; second, as a unicursal, directed structure, where once you enter, your path is strictly and unerringly directed to the goal. But though verbal representations of labyrinths include both multicursal and unicursal models, visual representations of whatever kind are invariably unicursal: symmetrical, clean, and impossible to get lost in. And there is no sense of contradiction, even when a multicursal text is illustrated with a unicursal diagram. This state of affairs continues until the sixteenth and seventeenth centuries when the renaissance humanists (with their noted literal-minded adherence to antiquity) began introducing disorder and confusion to their diagrams and their actual mazes. The medieval authors and artists themselves were unconscious of the distinction themselves and could not have explained it if they were asked. The point is consonant with the deeper theme of the book: that the medieval student of labyrinths was poised between seeing disorder in the world and order in God, and finally could not produce a `real' maze without letting the divine order control its lineaments. (There is to be sure another paradox in representation: verbal representations often convey the experience of being lost within a labyrinth, but the visual representations always stand outside, from whence things look very different: one thinks of the lost tourists at Hampton Court in Three Men in a Boat who are finally rescued when a man on a ladder stands just outside the maze shouting directions to them -- it's always easy from there.)

Second, if there is one thing we all `know' about medieval mazes, it is that labyrinths portrayed on church floors were used by the stay-at-homes as substitute pilgrimage routes, to be tracked, perhaps on their knees, as a surrogate for a real voyage to Jerusalem. We are wrong: Doob is very convincing (at 119 and n7) in tracing that theory to origins in a French writer as recent as 1817, and very modest in presenting evidence and argument that must have been a delight to develop. In fact, the church-floor mazes are there on one level as a sign of the architect's genius (comparable to that of maze-making Daedalus), and on another for the way they evoked the myth of the inextricability of hell tamed by Theseus-like Christ.

All readers of this book will find links to their own studies and will see things that Doob's broad synoptic grasp has missed. Here I will suggest briefly some themes that seem to merit further exploration. (1) As Doob says (130), `it is presumably not by chance that virtually all French labyrinths have twelve concentric circles, possibly alluding the zodiac that occasionally accompanies the design and measures human time, within whose constraints the world only seems to be confusedly labyrinthine.' The resemblance is in fact powerful and convincing: the labyrinth is in some sense a counter-image of the whole cosmos, a man-made parody of the divine creation. Comparative study of the imagery of the two systems would repay itself. (2) Doob largely omits discussion of the Pythagorean hesitation at the bivium (mentioned briefly only at 46 and 239n30, not in the index), a want to be noticed in a book that sees mazes in all wanderings and perplexities of path. (3) In one passages (77) there is explicit link to the liberal arts and (shrewdly) to the claim of old `pagan' Symmachus that it is `non uno itinere' that men approach the divine secrets. The point is excellent if we realize and recall that for Boethius' innovation of the word quadrivium and the Carolingian back-formation of trivium, the image at hand was not that of paths leading out from a point of origin into a wide and wonderful world of learning, but rather a plurality of paths converging on a single, unitarian goal of divine wisdom. Put walls around that image, and it too is a labyrinth. This may be confirmed by Doob's quotation (88) of Augustine's contra Academicos 3.4.7, where she misses the point by failing to quote the Latin: `sed dum ad istarum disciplinarum quibus excoluntur animi circulum revocare vos cupio, metuo, ne vobis labyrinthus fiat.' In that place, `disciplinarum . . . circulum' is a nice a calque as you could ask for on the Greek enkuklios paideia: Augustine's text can thus be taken as recommending the liberal arts as a key that turns the multicursal labyrinth into the unicursal.

When the book passes beyond the broad picture of intellectual history (a very Warburgian exercise) to readings of texts that some qualms about the method must be brought to the fore. Briefly put, Doob is so emboldened by her labyrinthine adventures that she sees labyrinths if not everywhere, at least in central and crucial places. If the House of Fame explicitly plays off against Vergil and Dante, then readings of those two authors are given to show that the labyrinth is central to their masterworks, and Boethius is thrown in for good measure. The readings proposed are stimulating and interesting and challenging, they are also unconvincing. They are exercises in over-reading -- not necessarily a bad thing, as an exercise, but here they end by constructing a scaffolding that is too large and too rickety for the weight it is asked to bear.

Let Vergil stand for the others: (228) `the idea of the labyrinth constitutes a major if sometimes covert thread in the elaborate textus of the Aeneid, providing structural pattern and thematic leitmotif.' Alas, what she proves is that it is remarkable that Vergil does not use the image more explicitly. What she presents is in fact a fantasy counter-Aeneid, not the Aeneid itself, in willful disregard of the way books work. Thus she speaks (242) of Nisus and Euryalus `Imprisoned in their virtually impenetrable and inextricable camp and greedy for fame': the first half of that phrase is meant to tie us to the labyrinth tradition, the second to prepare us for the reading of the House of Fame that follows, but neither is necessary, and the first not at all well-grounded in the text. Most remarkable at 250: `Pietas, the paramount virtue in the Aeneid, is a quality most necessary in mazes, where success involves persistence in and respect for the paths and goals that are laid out, patient endurance of the unavoidable twistings and confusions that lie between bivia, and the exercise of moral [!!!] intelligence at points of choice.' A consistent and accurate sense of pietas will not be extracted from that sentence, and it must be remembered that moral intelligence is only necessary in metaphorical labyrinths.

The readings of Boethius and Dante similarly overpress their cases. Thus when we come to what is the center of the book's own labyrinth, the House of Fame, we are already out on a limb. On two pages opening the discussion (308-9), we get fourteen straining phrases attempting much too literally to link Chaucer to the over- read tradition so far presented: `probably knew', `had many opportunities to become familiar', `would have seized the chance', might well have learned', `would surely have made', `might have visited', `perhaps his readings . . . corresponded in some degree to the labyrinthine readings I have just offered', `it seems likely', `may also have alerted Chaucer', `a discrepancy that may inform The House of Fame', `did Chaucer also know . . . or . . . ?', `Did he know . . . Or . . . ?', ` Did he ponder . . . ?', `Chaucer could have gleaned'[1] The reading of the House of Fame that follows is probably the best thing in the book, and to be commended to all Chaucerians.

What in the end is a poem worth? Could not `Studies in the Backgrounds of the House of Fame' have been a worthy book to write and to read? This learned and lucid book yokes the study of the theme and the study of the poem and, by the excessive claims that link the two, does itself a disservice. Is there an implicit (and quite involuntary) undervaluation of the text -- too short for that seprate study? not canonical enough? Is it necessary to range beyond the text pursuing pervasive influences? What texts will this book make its readers read again? Few Vergilians or Boethians will even find the book, and the Chaucerian who very well should be driven back to the 2000 charming lines of the House of Fame has been taken a long and at times dangerously unconvincing detour, for all that it is conducted with grace and learning at every step. Surely the text could have been gotten at more directly: as it is, it lurks at the heart of a labyrinth that too few will penetrate.[2]

1. A much longer catalogue could be dressed of passages in which readings of earlier authors are qualified and `put in brackets', as it were: readers who wish to hear Doob's case for her style should study pages 192-5 carefully and decide for themselves. I would locate the focus of disagreement at 195, `I propose merely that the idea of the labyrinth is a useful model or analogue for, and codification of, the way much literature was thought to be designed, written, and understood.' That is a central statement of the book's theme, and the word `merely' in that sentence is the heart of the problem.

2. I will register briefly reservations about some of the learned apparatus. The bibliography of modern literature is excessively selective and excessively Anglophone. E.g., in the portmanteau general note on Dante seventeen items are quoted, all in English save one in Italian (and that by an Anglophone American). The handling of Latin texts is also discouraging. Translations (when not from the Loeb) are not quite reliable and regularly over-read mazes into the text (so `inremeabilis error' at 27 becomes `irretraceable maze'). Despite some fussing over exact senses of words, the TLL seems largely terra incognita, and I cannot tell whether she has read the TLL article on labyrinthus, for she seems to have all the references cited there, but may have gotten them through intermediaries.