Medieval invention is considered first in relation to the powers of the rational and the irrational: the outcome of an artefact may have particular persuasive (more rarely supernatural) force which casts light on the makers of the artefact themselves. Was especially remarkable art necessarily the product of virtue? The answer in Antiquity and the medieval world was clearly “no”. The rational and rhetorical idea of persuasion touched and blurred with the idea of seduction, and the agency in the making of an artefact might appear supernatural, free of human agency. But exceptional effects are not norms, and the concern of this paper is with the everyday, the social, rather than the exceptional and unique. This is explored in the second half by thinking about rules and conventions that are agreed socially, and in regard to which the aesthetic and the ethical once more touch and blur. Stress is laid on the medieval pleasures of taking conventions and playing with them in the process of art realization, a social process which could ultimately reaffirm the authority of traditional starting points.

DOI “O for a Muse of fire, that would ascend/ The brightest heaven of invention”, says the prologue to Shakespeare’s Henry V. Artistic invention in the Middle Ages was not so major a theme as it was to become in the Renaissance, when its potencies were an important way of signalling the competitive machismo of great artists. Yet the rhetorical idea of invention, that finding out of ideas and material set out in Cicero’s De inventione, was perfectly familiar to the Middle Ages, as readers of Geoffrey of Vinsauf’s Poetria nova will know.¹ It was first a literary idea, and though I do not find it greatly emphasized in the artistic annals of the period, it was an idea subject to the formulations of Greco-Roman rationality. It belongs fully within the Latin and romance tradition. Consider that famous sixteenth-century expression “Se non è vero è ben trovato”—“if it be not true it is at least well feigned”, or, in more modern speech “It may not be true but, goddamit, I like it.” Trovato means to “find” and probably comes from the later Latin tropare, “to compose” or “invent”, a word which in turn produces trover, trope, troubador and so on.² This troping occurs on folio 15 of Villard de Honnecourt’s portfolio where a church plan is labelled “Kevilars de honecourt trova & pierres de corbie” (fig. 1).³ Under the chord of the apse of the plan was added an inscription in brownish ink “istud bresbiterium invenerunt ulardus de hunecort et petrus de corbeia inter se disputando” (Villard de Honnecourt and Pierre de Corbeil made up this presbytery, debating between themselves). Invention is a subject for argument, disputation, even “wrestling” intellectually, as the nearby drawing of two wrestlers suggests. It entailed argy-bargy, force, but also reasoning.

Many aspects of rhetoric and the history of ethics teach us that art is in part a matter of rational judgement, of self-conscious calculation. For Aristotle the purpose of making and hence skill (téchnē, Latin ars) is to produce a thing (Aristotle takes the example of building), while action has as its end a judgement, ideally practical wisdom or phronēsis. Téchnē is thus a ”reasoned productive state” whose action is poietike—an idea which led Horace, among others, to think of poetry as something “crafted”, hence his work Ars poetica, literally the “poetic craft”. In his Metaphysics (Iine 1) Aristotle considers téchnē to be a product of remembered experience from which general method and system is derived. Hence some have written of the “Craft of Thought”: a fertile idea partly because it indicates the crafty, educable nature of thinking but also, per contra, the thoughtful, educable nature of craftiness.⁴ The idea that highly skilled artists, architects especially, possessed (and were defined professionally by the possession of) knowledge of certainties and causes—what scholastics called scientia—suggests that medieval invention was primarily rational. This idea was developed from Aristotle’s teaching about causes, and that it was applied to architects is vouched for by Christine de Pizan.⁵

So the “brightest heaven” might suggest that everything was reason, sweetness, and light. But the classical and medieval worlds tell us that this was not exactly so. Much of what is interesting about medieval ideas of creativity lies at the junction of rationality and that which is less obviously rational, that which works in ways that are not immediately obvious, but which are immediately effective. In fact, for an important tradition in medieval thinking that was not Aristotelian in nature, it was exactly in the mysterious powers of art that danger lay. Suspicion of such powers originated in Platonic thought, in the belief made apparent in the Symposium and Phaedrus that there is something erotic and hence corruptible in the arts of communication.⁶ Quintilian (Institutio Oratoria, 2.15) thought along similar lines in attempting to rescue rhetoric from the charge that it is merely persuasive, for, as he says in the same place, money, the “mere look of a man”, courtesans, flatterers and corrupters are all persuasive. This made persuasion a larger thing than speech alone, a thing of “looks”. It is an easy matter to trace suspicion of such sensuous things as colour all the way from rhetorical thought through to monastic critiques of the eleventh and twelfth centuries and later Lollard and antifraternal literature such as Pierce the Ploughman’s Crede. The word persuasion is related to the word suavitas, sweetness, a feature of the middle of the three levels of rhetoric. Sweetness, like most interesting aesthetic ideas, is notoriously tricky and ambivalent, at once winning and pleasing, but also tinged with the potential for human error. The fruits and serpent of Eden were seductive: Honorius Augustodunensis indeed refers to the “serpens persuadens” of Genesis.⁷

Persuasive and talented artists might themselves be morally ambivalent or worse. Daedalus and some of the talented human monsters in Pliny’s Historia naturalis (35.71) such as Parrhasius, “fecundus artifex sed quo nemo insolentius usus sit Gloria artis”, a prolific artist but one who enjoyed the glory of his art with unparalleled arrogance, form part of the same tradition as Matthew Paris’s condemnation of the mason Hugh of Goldclif at St Albans as “vir quidem fallax et falsidicus sed artifex praeelectus”, a deceitful man and a liar, but a pre-eminent craftsman.⁸ William of Sens at Canterbury had fallen from the scaffold, according to Gervase’s elliptical words, either because of the “vengeance of God or spite of the Devil”. Dante, of course, associates great painters with Pride in his Purgatorio (Canto XI).

So there was a dark side to invention. Extraordinary invention is both miraculous and divine, yet art, craft, and cunning all have an aspect that, as commentators during the Renaissance were to emphasize, was melancholy, or actually morally suspect. Aristotle’s writing on oratory made it possible to think that the ethos or character of an artist lay within his work, not within himself: a bad man could give a good speech (Quintilian disagreed). Cunning, deriving from the Old German kennen, points to knowledge, and to the word Kunst, “art” or “know-how”. It also flags up the possibility of deceit. The word craft was developed from strength, power, and force (as in the German kraft). But while medieval things might have force or virtus, as Xénia Muratova notes, there is no evidence that medieval craftsmen themselves were regarded or regarded themselves as magicians.⁹ They possessed strength—the strength needed to “honour” material. They were not always morally good, for persuaders are not always nice. But they did not weave spells. The powers of art and the powers of invention were not exactly the same.

There is a fine line between the power of seduction—a power which, as Eve found out, does not operate upon rational judgement but which seeks rather to sidestep it with appalling consequences—and persuasion, in the working of which thought and judgement are integral. The two are distinct yet related. A seducer cleverly and even unconsciously overcomes well thought-through and rationally held core values, is a ruinous corrupter, from the Latin corruptio and corrumpere, to seduce, corrupt, or spoil. A persuader, however, produces conviction by appealing consciously both to heart and mind. Persuasion does not typically entail coercion or duress. And there is little evidence that in the Middle Ages the powers of rhetorical persuasion were deemed to be actually magical. Some present-day writers suggest that rhetoric and magic are in fact the same. This seems to me to stretch the definition of magic a little far.¹⁰ We might speak of “magic” metaphorically, as in the “magical” effect of a successful work of art, without insisting on the literal operation of anything supernatural. But this is a Romantic turn of phrase, not a medieval one. Those (unusual) words of Master Gregorius, the supposed English clerical visitor to Rome who wrote of the “magical persuasion” (magicam persuasionem), of the statue of Venus on the Roman forum to which he was repeatedly compelled to return, say something principally about the dark power of ancient idols: this remained a trope of fetishism into the modern era. I think his words are periphrastic and allude to erotic charge—all this pointing us back to that Platonic idea of rhetorical persuasion as seduction.¹¹ His stress on compulsion precisely reminds us that incantatio, fascination, and enchantment had strong negative associations in the medieval mind.¹² Wicked pagan things were an artful class apart.

Here it is useful to make two additional points. The first is that there certainly was an account of medieval making (if not exactly invention) which stressed supernatural or superhuman origin. There were classes of medieval object which came into existence without apparent human agency and which could be said to possess their own intentionality. Seals are an instance. The emergence of an image on soft wax produced by a die is, metaphorically speaking, a sort of miracle. For Bedos-Rezak the seal impression supplies analogies for theological thinking about prototype, image, and incarnation, and about intention.¹³ The image emerges from nowhere as if it possessed intentionality; and indeed there is a case for saying that, whatever we think about authorship, the intentio auctoris can be the property not of a person but of a text or art object. The issue here is the way the outward configuration of an object or text steers us, leads us, towards an effect or affect. Intention is change, indicating the apparent operation of will or voluntas. On its own an object cannot do this simply—the context or “occasion” must be right.

This idea of an origin and intention was connected to the idea of authority. Medievalists regularly encounter the idea of the miraculous or God-given or angel-induced image, be it an image, illumination, image-relic and so on, or indeed a building. Giraldus Cambrensis writes in this way about an Irish Gospel book and its miracles of miniaturization.¹⁴ Until 1200 or so there was also a class of buildings inspired in their basic arrangement—scale especially—by divine mandate: an instance is the miracle of Gunzo at Cluny in which the dimensions of the great new church were dictated apostolically in a vision. God and the saints set out the measurements of churches as if they were Noah’s Ark. These measurements were, inventively speaking, a cognitive model or starting point, the sort of basic formula discussed in Richard Krautheimer’s still fundamental article of 1942 on architectural iconography.¹⁵ We are familiar with these transferences in pre-Gothic England as at Lincoln or Norwich or Canterbury. In such cases we are dealing with a handing down, tradition: obedience and human agency are necessary to record and implement the numbers.

After origins, effects: the wonder-response to great skill, great scale, or great effect is very well documented in the Middle Ages, as Caroline Bynum noted.¹⁶ In such extraordinary cases the agency at work may be perceived as in some way supernatural (as in the case of the actually miraculous), but it is usually rooted in the human sensory faculties, unlike the most rare transcendent, visionary, or ecstatic experiences, events actually beyond experience and discourse. For the most part even experiences of wonder are really experiences not of enchantment but engagement. Enchantment is a form of force majeure working one way upon a possibly unwilling or unconscious “recipient”. In contrast, engagement (like persuasion) is a two-way pledge requiring consenting adults. It is true that occasionally the effects of art are astonishing and that we react to them with wonder, that initial disorientation and questioning which precedes reasoned reflection on “how it was done”: “Hoc opus eximium vario celamine mirum”—“this extraordinary work, a marvel of varied relief”—as it says on the twelfth-century apse reliefs from Königslutter (fig. 2).¹⁷ Wondering might entail an emotion, or a kind of hesitation, even the attribution of the source of the wonder to supernatural authority, angelic hands: but the effects, no matter how remarkable, are still within the realm of nature whatever their causes. In the appreciation of the elegance and complexity of the marble used in the new choir of Lincoln Cathedral begun in the 1190s, the Metrical Life of St Hugh of Lincoln (lines 879–80) states that on inspection the stone can “suspend minds” in trying to decide what kind of stone was used, jasper or marble. But the reflection which follows, like the sculpted surfaces, is palpable, consisting of a sensory and intellectual weighing up of such things as how the effects were reached, or what their purpose might be. Wondering is not the end, but the start of thought and enquiry. Wonder gets us going.

In breathing human life back into medieval artefacts—the great concern of much present-day art-historical and cultural criticism in its reaction to an over-intellectualized twentieth-century Kunstwissenschaft—we do well to remember those darker powers of moral uncertainty and ambivalence which combine in one of western Europe’s great traditions of scepticism about human attainment, that art possesses dangers, that it has a side not exactly captured by reasoning. However, having issued that reminder, it is important not to lose sight of another truth about medieval crafting: that it was not always engaged in the operation of, or search for, the exceptional. The medieval idea of craft connected to the “good life”: the daily business of the craftsman, that harmonizing of means and ends had to do with getting things “right”.  The modern idea of art, formed under the pressures of Romantic thought as a special, even magical, zone of experience, is quite alien to the ordinary daily preoccupation with “rightness”. This is why the formal arrangement, the crafting and calculation of what are now called works of art, matters, and why the demotion of the notion of craft in relation to the special zone of fine art—a nineteenth-century “achievement”—has not been helpful.

The artisan judges what is ordinarily right, not exceptionally right. Artisans do this because durable works of art must fit, be useful, and be tolerant of the vagaries of personal subjective mood and circumstances. The sense of something being ordinarily right is embedded in the conventions of téchnē or ars. Getting things right entails the values of proportion and measure. These underwrote the idea of the worthiness of something, whether or not it was “honest”. The word “measure” has its root in the Latin mensurare, to establish quantity; but it also has a second, as it were qualitative, sense of regulate, moderate, restrain. “Proportion”, like “honesty” was an important value in art commissioning as in social interaction more generally. In John Lydgate’s Song of Just Mesure, he states that everything contrived by man “standith in proporcioun”.¹⁸ “Proportion” here is not simply a matter of geometric quantity or number but, following the Book of Wisdom 11:21 and the authority of Augustine, also weight: in Mesure is tresour, Lydgate also asks

What may avaylle al your ymagynynges
Withoute proporciouns of weyghte and just mesour?¹⁹

The idea of weight as well as balance is important for any three-dimensional art such as sculpture, as it is for architecture. Balance—if not exact symmetry, at least a calculated negotiation of the forces in a form to produce harmony—mattered in art production as much as in the maintenance of the humours of the body, and therefore health.²⁰ The balance might be easily upset, indeed its force in aesthetic matters came precisely from its inherent precariousness, the artist striking a tensile balance without de-energizing the final result. And as with proportion, so with balance: such terms were linked in a chain of association, like an affinity, with ethical ideas such as equity or justice: “measure” connected to the sense of “rightness”.²¹ The verbal symmetry between the Latin verb artare, to bind, contract, or limit, and ars, usually taken at this time as “craft”, prompted a line of thought on just this problem. In his Etymologies (1.1.2), Isidore uses and develops this ethically when he states of the words ars-artus that “ars has got its name because it consists of artis, instructions and rules. Others say this word is derived from the Greek apotesaretes, i.e. from virtue, the name they use for knowledge.”

It is helpful to separate rules from conventions. Conventions are essentially guidelines indicating that if x is done y will probably follow. Inventiveness (a term perhaps preferable to the ontologically more problematical “creativity”) is thus more conventional than bound by rules. We might think of it as rule-referred, but not rule-bound. Rule-referral creates the scope for those winning and even unexpected deployments of rules which go further and seem even more “right”—“rightness” always being relational, occasional.²² In this sense of rightness lies the idea of nifty selection from the options, or elegance. Elegance (the term comes from the Latin to “select”) is, however, based on skilled choice from limited, not unlimited options. In radical cases—much rarer—the artisan may end up changing the rules altogether. In society too there are those who are outlaws beyond the rules, those who are within but who create new rules or change the old ones, and those who just follow.

But all artists, like all writers, know that rules, like plans, are only a starting point: that they may be referred to, but also pulled around, in the service of art. This exposes a difficulty with the transfer of modern deterministic or rule-bound notions of plans to medieval invention. As no less an authority than Le Corbusier said: “The plan is what determines everything; it is the decisive moment.”²³ In understanding medieval art this is not entirely helpful, though its consequences have been thoroughly worked out by those subscribing to the notion of “Platonic geometry” as an archetypical basis for medieval architecture.²⁴ The whole point about an archetype is that it cannot be exactly replicated. There is no support in Geoffrey of Vinsauf (or in rhetorical teaching) for assuming that the “plan”, or consilium, is something worked out in full in all its ideas: Geoffrey (lines 55–59) likens it to an architectural plan but also to a map in which the general scope of the voyage is known (its “Cadiz”) but not the detailed itinerary. Even when working within limitations there is room for manoeuvre. Rules may trap but they may also direct: that is what you and I sometimes call “style”. The Romantic idea that pre-planned work stifles all opportunity for “creative” spontaneity during execution evinces a curious modern failure to understand just what a plan is—a rational plan was in essence a guideline, a framework for inventive development, like the notation of a jazz riff providing a basic structure of harmony and bars upon which musical invention is founded. To imagine that it was purely a constraint is unconsciously to pursue a notion of individual spontaneity or creative fantasy as against a notion of rule-assisted (as opposed to rule-directed) activity. Jazz riffs are, after all, founded on pretty well invariant formal progressions such as twelve-bar blues, but the freedom of the riff is not a subversion of the given matrix of harmonic narrative, but a sort of affirmation of the basic soundness of the underlying formal structure.

Activity which is rule-referred rather than rule-changing always permits the unexpected or aleatoric, but steers it: this is a basis of style. Ideas are “found out” by artists as they are realized, and are not preordained. This is why the idea of ludus is important: the gaming element is fun, and also social, because it relies on surprise, anticipation, improvisation, and more deeply on an implicit common ground of understanding and constraint.²⁵ Take those fabulous jazz riffs, manuscript marginalia.²⁶ Marginalia are a specialized, but by no means isolated, instance of the larger phenomenon of the ludic in medieval invention—play that can be serious and high-minded and ritualized, but which fundamentally energizes the whole process of creation. The area of creative play is necessarily specialized, in the sense that it implies the priority of conventions or rules—the order of the text page—which must first be understood and assimilated in order to be cast aside: these rules or conventions might belong equally to the relatively enclosed order of a craft with its specific training and procedure, or to the literary or intellectual practices of certain professions, or to playing chess or making music. The recent dominance of modern literary and social theory in discussing the marginal has I think led us to overlook the serio-ludic aspect of craft itself, in which persuasion, the power to bait and lure, lies in the witty manipulation of things based on artisanal experience and method, as much as in the signification of things themselves. These are not just production-line jokes and nor are they necessarily about meaning or symbolism or power. The claim is much larger: it is that in the witty manipulation of things lies their actual power to charm, move, and convince, to create an experience.²⁷ Marginalia are spread across all Gothic artworks because this is what the intended experience required, not least as a matter of style.

Theories of “freedom” run the risk of overlooking the key dynamic of order and disorder in any inventive process: that order and disorder are symbiotic and confer significance upon one another in a way that renders theories of mutual subversion hazardous. The first thing any book-maker did to an illuminated page was rule it with a ruler. Practical experience alone shows that drawing a babewyn against such a framework becomes a pleasurable unruling, not least when the look of the page has stiffened into highly regularized column formats with running heads, framing ornamental bars, and a clear hierarchy of text sizes, and when the people executing such pages are making them by the thousand. Serial production for large markets, and formal standardization were arguably a fundamental part of the dynamic of the “marginal”, the wit of the book lying, as it were, in the self-conscious exploitation of page layout as an object of artistic comment in, and on, itself. The order of the book has conferred significance on the self-conscious (but artful) pushing of the boundaries of orderliness. So the page’s internal relations are necessarily artificial, and in the creation of a zone of unruliness lay something ritualized.

Toying with the relation of the permanent and impermanent also provided one of the pleasures recognizable in the extempore or aleatoric. The history of cultural production in the pre-modern world tells us that art does not always consist of fixed things, permanencies, data. Much production of the highest order has always been transient, evanescent, ephemeral: modernists and postmodernists too rightly make much of ephemera, of the “happening”, trying to win back the life and vitality of art. I think there are at least two aspects to ludus and ephemerality. One is that Gothic art placed them at the centre of what it valued. Hence it disturbed the Renaissance purveyors of virtù, manly reliability, stability. At the end of the fifteenth century the Italian architectural theorist Filarete discussed the barbarity of Gothic architecture by association with metalwork in his Trattato: “The goldsmiths fashioned buildings like tabernacles and thuribles, and they made real buildings in the same manner, though these had nothing to do with architecture . . . These modes and customs they have received, as I said, from across the mountains [oltramontani], from the Germans and the French.” Vasari hammered in the final nail in the introduction to architecture in the 1550 edition of his Lives. Not only were “German” works very different from classical architecture, they had portals with columns so “subtle and twisted” as to be incapable of support; these architects “made such a malediction of little tabernacles [maledizzione di tabernacolini], put one above the other, that they cannot stand.” Note the diminutive, tabernacolini: this practice is an assault on proper hierarchy, magnificentia, it confuses what is small with what is great. Finally, says Vasari, in their lack of Vitruvian firmitas, such buildings “seem made more of paper than of stone or marble”.

Contrast the situation in Gothic literature which precisely, I think, bears Filarete out: this was not the operation of a stupid prejudice. What to the Italians seemed like an unstable, and destabilizing, aesthetic vice was to their northern counterparts a fundamental aspect of invention, not least literary invention. A passage in the late fourteenth-century alliterative poem Sir Gawain and the Green Knight describes Bertilak’s castle as a vision of glinting white towers, thick with battlements and pinnacles:

So mony pynakle payntet watz poudred ayquere
Among þe castel carnelez, clambred so þik,
Þat pared out of papure purely hit semed.²⁸

“Seemed” captures both a sense of uncertainty and a lack of solidity: things “pared out of papure” and also “poynted of golde” are found as table ornaments at Belshazzar’s feast in the contemporary alliterative poem Cleanness in the same manuscript source as Sir Gawain and the Green Knight. Cleanness (lines 1458–59) anticipates another theme of the Italian critics, the exchangeability of stone and metalwork, magnificence and munificence (the magnificence of the miniature), in describing Solomonic liturgical vessels shaped “as casteles arayed/ Enbaned under batelment with bantelles quoynt”. But—it has to be said—these surface effects of tinsel and instability were built on solid literary foundations which positively celebrated the ludic, the mixed and impure, the transient and the outlandish, the deliberately confused, as features of poetic invention which architecture, or the poet as master builder, seemed especially suited to expressing. Literary considerations aside, it seems possible that this was so because of the peculiar freedoms of English architectural invention itself: it served poets well, because poetry, or a sort of poetic imagination and playfulness, had already served it.

The second aspect of ephemerality is the extremely high standing of extempore performance—of spur of the moment, tightrope-precarious invention. Before the eighteenth century, and certainly in the Middle Ages, the majority of musical performance of any kind was of this type. Composers tended not to write music down: they memorized and improvised. What was valued was the power of extempore performance, far more so than memory-feat or sight reading. Extemporization, typically based upon the amplification of some small theme, was the real test of invention. This I find helpful, because my own experience suggests that we respond especially strongly to brilliant extemporization and that music reveals the truth of this most directly. An astonishing feat of skill, especially rational or disciplined skill in performance, is where I think a sense of the operation of something beyond the bounds of normality, something that some might, metaphorically, call “magical”, first arises. “Where is it coming from? How on earth is this possible?” These are the wonder responses: for wonder provokes the search for understanding—whence, how? To witness a double fugue in strict counterpoint composed flawlessly on the spot is to witness, I think, one of the hardest things anyone can do. There is a mixture of awe but also puzzlement, rather than simply respectful approval. In fact a common response is laughter, delight, and then a sense of “coming home” or restoration.

The Greeks saw something in this effect which was more than persuasion. One example is Longinus’s Peri Hypsous, “concerning the lofty”, written in the first centuries AD as a deliberate critique of persuasive rhetoric Roman-style, but cleverly taking the form of a rhetorical treatise. Longinus thinks of the effect of terrific speeches. At 1.4 he describes how wonderful effects are sudden, a sort of flash or corruscation: “sublimity, brought out at just the right moment makes everything different, like lightning, and directly shows that ‘all-at-once’ capacity of the speaker.” And importantly, at 7.2, by true sublimity “our soul somehow is both lifted up and—taking on a kind of proud possession—filled with delight and great glory, as if our soul itself had created what it just heard.”²⁹ Longinus demonstrates the revenge of Greek religion on Roman eloquence, and precisely indicates the difficulty of conflating the persuasive with the supernatural: as Curtius observed, quoting Longinus, ‘‘the extraordinary cuts the tie between rhetoric and literature . . . what inspires wonder in us is in every way superior to what is only convincing and pleasing.’’³⁰ Art at its highest does not persuade, according to this view; it just operates without argument upon us till we are outside of ourselves, as if in a religious state, or like magic. The lightning strike of sublimity written of by Longinus finds later echoes in Byzantine writing (e.g. Michael Psellos). But he was unread in the Latin Middle Ages, being revived only in the Early Modern period in writing about the Sublime.

For all the potencies of invention my concern, at heart, is with the everyday and not the exceptional experience. I want to close with two brief reflections on freedom and constraint: on tradition, and the creative powers of discourse that allow tradition to flourish rather than just repeat.³¹ That the idea of tradition, literally handing on, was powerful in the Middle Ages goes without saying, bearing in mind that the idea of tradition should be understood dynamically. Following an authority was a central way of gaining authority: medieval invention was mimetic, imitative, in the deep sense that it was self-referential (and not “naturalistic”). This concerned the relationship of invention to the existing stock of older material, as it were the “inventory”. There is much truth in the idea that the medieval relation of past and present was one of typological recurrence, as in Krautheimer’s notion of iconography. When we think of an image not being made by human hands but passed down to us, we are conceding something about its authority as well as its facture. The practical, thoughtful recourse to older models underlay, for the present writer, some of the striking aspects of English architecture at its most innovative in the period of the Decorated Style. Fourteenth-century English architects were sometimes just as stimulated by the inventory of twelfth-century architecture as they were by the work of their contemporaries in France. The discussion of innovation in architecture has tended to be assimilated more readily to an idea of modernity than to an idea of intelligent critique of the existing “stock” of architectural ideas. Another powerful instance of this is also provided by manuscript illumination at those moments when art innovation occurs. An instance, to cite Otto Pächt, would be the so-called Giottesque episode in fourteenth-century English illumination when it fell for a moment under the influence of trecento Italy.³² For Pächt, as for Panofsky, this moment was a moment of modernization, of aesthetic liberation.³³ The art of the Mediterranean lands had come to the rescue of tradition-bound northern artists, not least by the introduction of small-scale oblique spatial settings for figures in illuminated manuscripts. In the old Panofskian model (derived ultimately from Jacob Burckhardt), the triumphs of “illusionism” are basically teleological: they witness a stage on a pre-ordained route in the emergence of Western “rationality” expressed by perspective as a symbolic form. I am less sure. As regards England, Pächt was surely right to call this phase an “episode” rather than something larger or more continuous. When Italian or Byzantine models were considered and adapted by Gothic artists it is also possible that they were consulted not for their modernity but for their authority, the sense of access they gave to a newly fathomed, distant but authoritative world of “truth”, not “progress”.

A second reason for viewing modernist arguments sceptically brings me to a final point which recapitulates an issue mentioned earlier: that to some modern art-historical mindsets, rhetoric and the discursive and persuasive powers of art and architecture continue to be objects of suspicion. I have emphasized the issue of invention and persuasion explicitly in response to this model because I am interested in the relation between invention and experience. The English tradition of empirical analysis of Gothic architecture is powerfully scientific and often, within the terms of its own remit, correct to the point of self-affirmation. But for the present writer at least it suffers too from a deficiency especially in regard to architecture. It does not seek out the eloquence of buildings or artefacts, what social, intellectual, and affective experiences they engendered. Indeed, it seems not to consider such things to be proper objects of historical enquiry at all. It is as though two cultures are confronted, that of science, and that of the humanities. We cannot pretend that the engendering of experience is easily graspable by means of critical or historical analysis: it is in fact exceedingly difficult. But we should not be discouraged. The poetry of things matters and provides an important reason for studying and liking them. A lack of curiosity about experience, about the outcome of arts, may also impoverish the discussion of invention itself, about what goes into a work of art as well as what comes out. In my view, the arts all had their own special areas of responsibility; but we have more to gain than to lose by considering the arts together and by asking what common factors linked the brightest heaven of invention in the verbal and non-verbal arts.