Classicism in Italian Architecture


At first sight a clear distinction can be made between Romanesque, Gothic and classical Renaissance buildings; correct use of the classical language of architecture in the last category should make it easy to distinguish from the rest.

Unfortunately, the position is much more complicated than this. We are not only badly informed about what the Trecento or indeed the Quattrocento knew or believed about ancient architecture, but are equally vague about the stylistic intentions of the architects of those times. The matter is further complicated, as always, by the revival/survival problem; Tuscan Romanesque, for example, can be shown to contain elements from the antique which were ,to be re-employed during the Renaissance.511 Occasionally it is apparent that architects were unclear about the exact period in which their models had been built; this is the case with Brunelleschi's imitation of elements of the Florence Baptistery in his work on the dome of the Cathedral precisely because he believed the Baptistery to be a Roman building and therefore worthy of imitation. Alberti used the same source, surely for the same reason, in his Holy Sepulchre for S. Pancrazio, Florence (c. 1455-60). Although it is self-evident that the classical style in architecture implies an interest in the remains of Antiquity and in the details of the Orders and their



A medal of Sigismondo Malatesta by Matteo de'Pasti, showing the dome with which Alberti wished the crossing of the Tempio Malatestiano at Rimini to be crowned.

decoration, it may well be that Brunelleschi's architecture, pace his own best intentions, is only pseudo-antique. A scholar has recently asserted that 'there is not a single major work of Brunelleschi for which a plausible and specific post-antique source (or sources) cannot be suggested'.511 He illustrates his point by comparing the very similar ground-plan and elevation of the Old Sacristy in S. Lorenzo, Florence, and those of the much earlier baptistery of Padua Cathedral. Other scholars have demonstrated how much the details of Brunelleschi's architecture depart from the example of the antique as we now know it to have been.513'517

We must therefore make a distinction between the basic principles of classical architecture, which control its rationalism and emotional temperature, and the motifs, which can only be assimilated by the study of antique remains. Brunelleschi, as we shall see, could work in a classical manner without access to a full range of antique motifs.512-533 Then again, our knowledge of Quattrocento opinion about what made a building look antique is slight. Obviously a work which, to us, looks respectably antique can be both antique and Romanesque in inspiration, like the lantern of Florence Cathedral, which resembles the lantern of the Baptistery. The same is true of Brunelleschi's unfinished centrally planned



Brunelleschi: portico of the Pazzi Chapel, Florence.

church of S. Maria degli Angeli, Florence, commonly believed to have been inspired by the antique Temple of Minerva Medica at Rome; a much more likely source is the doubling of the transept of Florence Cathedral. Certainly, Michelozzo's circular chapel at the east end of SS. Annunziata (begun 1444), which really is a copy of the Temple of Minerva Medica, met with much opposition from the critical Florentines.527

Brunelleschi is important to the classical tradition because he understood the basic principles of classical architecture;514 the dome of Florence Cathedral, together with its absidal chapels, forms in reality a centrally planned church. He may have taken the idea from the Baptistery or, as Vasari asserts, he may have studied antique remains in Rome in 1401, after losing the competition for the Baptistery Doors


to Ghiberti. His biographer Manetti reports, perhaps fancifully, that 'he decided to rediscover the fine and highly skilled method of building and the harmonious proportions of the ancients . . . Together [i.e. with Donatello] they made rough drawings of almost all the buildings in Rome and in many places beyond the walls ... In many places they had excavations made in order to see the junctures of the membering of the buildings, and their type . . ,'518 But even if some of his ideas can be easily related to the Romanesque, and even if he tends to conceive of architecture in two dimensions, as a pattern on a wall, rather than as the forming of space (except, perhaps, in S. Maria degli Angeli), his example nevertheless laid the foundations of the Renaissance style. The basis of his buildings was geometry, for he believed in the power of geometry to represent and evoke beauty. Sphere and cube, and exact proportional relationships based on mathematics, made his architecture clean and uncluttered. Decoration was kept to a minimum so that the vocabulary of antique architecture—columns, pilasters, entablatures —might speak for itself. To such a work as the Old Sacristy of S. Lorenzo, Florence, one can apply the same terms as for Masaccio's great frescoes in the Brancacci Chapel: reason, balance, proportion and harmony. Although architecture cannot 'speak' in the same way as painting because it lacks narrative subject, nevertheless Brunelleschi's forms, because they evoked the past, evoked also that flood of associations which were wedded to the antique during the Renaissance.509

Indeed, his understanding of the antique enabled Brunelleschi to adapt antique forms to the needs of the Christian religion, as well as to the technical problems raised by the construction of the enormous dome of Florence Cathedral. But while it is correct to maintain that his knowledge of Roman masonry led him to adopt a strong herring-bone formation for the masonry of the dome which did away with the need for centring,516 it is equally true that the skeleton of the dome, formed by the great ribs which require the weight of the marble lantern to keep them from buckling, is Gothic in inspiration.515 Brunelleschi must of course

have known this, but it is pleasant to imagine him thinking his work the equal of that most influential of all antique edifices, the Pantheon. It seems likely, from the little knowledge we have, that the form of the Pantheon dome conditioned that of S. Maria degli Angeli (1434^7), which he began immediately after another postulated visit to Rome, no doubt with Donatello (1432/3).510-519

The centrally planned church: a Renaissance ideal

In Rome, Brunelleschi surely met L. B. Alberti, who may well have inspired him to build a truly antique structure. S. Maria degli Angeli, however, is not a totally new departure in Brunelleschi's work, for it makes explicit a type of plan already broached in the Old Sacristy of S. Lorenzo (designed 1419, built

Brunelleschi: Pazzi Chapel, interior.

1421-8) and in the Pazzi Chapel adjacent to S. Croce (1430 or later). This type is the centrally planned church, able to fulfil better than any other form the Renaissance interest in geometry. Because of its 'perfect' shape—always some variant on the square or the circle—and patent descent from antique prototypes, it was to become the preferred shape of the ideal Renaissance church. But in practice this form was highly inconvenient. It was very difficult indeed to span such a great area without intermediate supports, and the placing of altar and congregation, as well as the planning of suitable routes for ceremonial processions, posed almost insuperable problems.523 Brunelleschi's S. Lorenzo (begun 1419) and S. Spirito (designed 1434 or later), with their basilican form and flat wooden ceilings, were much more practical and still allowed the architect to exercise his interest in geometry.


classicaltrad3-1.jpg classicaltrad3-2.jpg

Rome, the Pantheon. LEFT: A nineteenth-century photograph taken before the removal of the two bell towers. RIGHT: Sebastiano Serlio's plan. From the Architettura, Book HI, Venice, 1584.

The prototypes both formal and icono-graphical of the centralized plan of the Renaissance church were various.522'524 A clear Christian source was the tradition which connected the circular church with the Virgin who was frequently referred to in litanies as the 'Temple of the Lord';521 hence the appellation of the Pantheon as S. Maria Rotunda. The Virgin is often depicted in paintings with the Temple of Jerusalem, popularly believed to have been a centrally planned building, as in Raphael's Betrothal of the Virgin (and its source, Perugino's version in Caen, 1503/4). In Perugino's Donation of the Keys in the Sistine Chapel, the building is indicated as an IMENSV(M) TEMPLUM, and again the Temple of Jerusalem is clearly intended. In some respects, therefore, the Virgin was the Church. Equally, antique pagan tradition connected centrally planned buildings with certain gods,


ABOVE Brunelleschi's design for the east end of S. Maria del Fiore, Florence, approximates to a central plan.

particularly Diana, a virgin, who was on occasions compared to the Virgin Mary. During the Renaissance, antique circular temples (many of them, of course, tombs) were often called 'temples of Diana'.520

Again, the Church of the Holy Sepulchre in Jerusalem was centrally planned; and its frequent imitation in Italy during the Renaissance525 usually corresponds in form with the Early Christian martyrium, a monument commemorating the death of a martyr,31 and based on the pagan Roman forms of the mausoleum, These buildings are imitated in several Renaissance churches. The best-known example of the church as memorial is Bramante's Tempietto beside S. Pietro in Montorio; the above considerations make it obvious that its shape is not determined merely by mathematical aesthetics or by a desire to imitate the antique closely, for the work was intended to mark the spot where St Peter was martyred. Other examples would be Michelozzo's rotunda527 for SS. Annunziata, commissioned as a memorial to the Marquis of Mantua, and S. Maria della Pace in Rome (by 1483), a memorial church by an unknown architect. Of course, the greatest of them all is the new St Peter's, conceived (but not completed) as a centrally planned building, and planned by Julius II as a mausoleum for himself, as well as to shelter the relics of St Peter.

RIGHT Plan of S. Maria degli Angeli, Florence, by Michelozzo.


In the question of centrally planned churches, the revival/survival problem is again much in evidence. As with Brunelleschi and the Baptistery of Florence Cathedral, so Bram-ante had no way of distinguishing Early Christian (or later) from Roman work, because most of the serviceable structures had been converted into churches. For Bramante, the circular mausoleum of the daughter of Con-stantine, S. Costanza, must have been as important as the pagan mausolea of the Appian Way, the circular Temple of Vesta by the Tiber, or the Temple of the Sibyl at Tivoli. The circular churches of the Renaissance are a mixture of pagan and Christian: Raphael's centrally planned Chigi Chapel in S. Maria del Popolo is Christian, yet its position near the Porta del Popolo, and its shape, as well as the presence of Elijah-cum-Jove in the cupola, inevitably link it with the form and siting of pagan memorials, which had to be erected outside the City.

The final link between mausoleum and church, pagan and Christian, is that baptisteries are, by long tradition, often centrally planned. They are places of death and resurrection : the death of the old life and the birth of the new. A mausoleum, for both pagans and Christians, has similar overtones. With such a wealth of associations behind the


classicaltrad3-5.jpg classicaltrad3-6.jpg

LEFT Interior of S. Constanza, Rome. Until 1791 it held the porphyry sarcophagus of Constantia, Constantine's daughter.

RIGHT The Temple of Vesta (or of the Sibyl), Tivoli.

central plan its popularity in Renaissance architectural theory is understandable. We shall meet it again when considering the work of Alberti and Bramante.

Alberti, architect and scholar

Brunelleschi's S. Maria degli Angeli is the first separate centrally planned church of the Renaissance, although both the Pazzi Chapel and the Old Sacristy of S. Lorenzo are in effect separately designed buildings as well. They were the foundations on which Alberti, who dedicated his Delia Pittura to Brunelleschi, built his theories, aided by the only architectural treatise to survive from Antiquity, Vit-ruvius' De architectura, and by his own observation of antique remains.534'537

When Brunelleschi died in 1446, many of his projects were incomplete, and were to be completed in accordance with others' designs. Alberti's schemes were often to suffer a similar fate, but we know his intentions better than those of Brunelleschi because his theories are clearly stated in his imitation of Vitruvius, the De re aedificatoria, which was completed by 1452.535 The first printed edition appeared in 1485. These theories were of great importance, for they influenced not only Bramante and Raphael, but also the architectural textbooks of Serlio and Palladio and, of course, the latter's architectural style. Often dismissed as a mere theorist, Alberti's example is in fact crucial for the classical tradition because he demonstrated how to apply the vocabulary of antique architecture to contemporary needs.529'530

Alberti's scholarly knowledge of antique architecture was, we may be certain, much greater than that of Brunelleschi; the older man was a consummate construction engineer, for example, whereas Alberti was a scholar and apparently never a practical architect in that sense. His position in the Papal Curia entailed not only advising the Pope on the rebuilding of St Peter's (an idea broached long before the time of Bramante) but also supervising restoration work on Early Christian churches such as S. Stefano Rotondo and S. Prassede.541 It was in Rome, amidst that repository of antique solutions to the problem of how to rebuild the city in a grand manner,546 that he composed his book on architecture.538

In the Seventh Book of De re aedificatoria he lays down his requirements for the ideal church. Unlike Vitruvius (who has little to say about centrally planned buildings), Alberti ignores the basilican form in favour of the central plan. He wishes churches to be 'of great Use for stirring up Men to Piety, by filling their Minds with Delight, and Entertaining them with Admiration of their Beauty', and therefore recommends the central plan: the church is to be placed on a podium which will elevate it above the everyday world, and embellished by a portico such as the Ancients had used. The whole should be rich yet simple, in colouring as in form. High-placed windows


Leon Battista Alberti's self-portrait on a medal, c. 1430.

would concentrate the mind on devotions, for 'That Horror with which a solemn Gloom is apt to fill the Mind naturally raises our venerations, and there is always something of an Austerity in Majesty . . .' The building should be completed by a dome, and even the 'Composition of the Lines of the Pavement full of musicall and geometrical Proportions' will echo the perfect harmony of the whole structure (translation by J. Leoni, London, 1755). Beauty is conceived as being the result of a mathematical formula based on reason, in which nothing can be added or taken away without spoiling the whole. One scholar, denying that Alberti was a mere antiquarian, maintains that his passionate concern with the revival of antique forms arises 'from a conviction that the beauties and harmonies of the plastic arts correspond to a moral and spiritual equilibrium of human existence'.531

As we have seen, Alberti's ideas on architecture derive partly from his close study of Vitruvius. Although Vitruvius' text had been




Amphiprostyle Temple, from Caesariano's edition of Vitruvius, Como, 1521. Renaissance editions of Vitruvius contain more central plan buildings than the text warrants. RIGHT Diagram of the winds. Philibert de I'Orme: Architecture, Rouen 1648. This figure is similar to that in Vitruvius for ideal city plans, and to the well-known Renaissance scheme of the man within the circle as a measure of all things. It is another example of an obsession with the centralized plan.

known and quoted since Carolingian times,528 Alberti's examination is important because he was the first scholar with a good knowledge of antique remains to compare written word with surviving ruins, and to form a system of architecture from that comparison. His interpretation and misinterpretation of the text can stand as an example of the kind of difficulty faced by any scholar who looked to ancient texts for authority.

The problem with Vitruvius' text was the obscurity of its language. Alberti admitted (De re aedificatoria, VI. i) that, in the field of architecture, Vitruvius was 'the only writer, and without doubt very instructive, to have survived that great shipwreck [of Antiquity], but much ravaged by time so that in many places much is missing and, in others, you would wish for a fuller explanation'. His style was so bad, Alberti complained, that 'the Latins thought he wrote Greek and the Greeks believed he spoke Latin . . / Such obscurity meant, of course, that the commentator could read almost what sense he liked into the Vitruvian text and still, as it were, receive the blessing of Antiquity. The text must have been illustrated once upon a time, but no manuscript with illustrations survived. The scholar's


task was therefore to try to make sense of Vitruvius by comparing his words with the visible remains of Antiquity. (Fra Giocondo, who edited Vitruvius in 1511, may have similarly clarified Vitruvius' account of town planning by referring to what he could see on the ground.)532

It is no surprise to find that Alberti sometimes made mistakes, as with Vitruvius' account of the Etruscan Temple.536 Vitruvius describes a temple type with three cellae at the rear of the site and a portico taking up half its area. Actual examples of that form have now been uncovered, but Alberti knew nothing so strange. And so, in the De re aedificatoria, he describes the nearest thing he knows that will fit Vitruvius' description—a miniature version of the Basilica of Maxentius, which, indeed, was believed throughout the Renaissance to be the remains of a temple.

Unconstrained by Vitruvius' neglect of centrally planned temples, Alberti makes his own specification for the form: such structures must, he writes, have a portico leading to a room which must be vaulted, and should usually be circular or some elaboration on that shape. He obviously studied buildings like the Pantheon and the Temple of Vesta at Tivoli, but for the more complicated apse and chapel shapes that he prescribes Alberti must have closely examined the mausolea in and around the City of Rome. Krautheimer forges another link between temple and mausoleum by suggesting that the widespread Renaissance belief that antique gods were great men made immortal (euhemerism) may have made it clear to Alberti that such small buildings, precisely because they contained tombs, were in fact temples as well, raised to the glory of the newly created god. The same duality would, in Alberti's eyes, apply to Early Christian centrally planned buildings such as the Mausoleum of Theodoric at Ravenna, or S. Costanza in Rome.

Whatever the quibbles, Alberti it was who resurrected and interpreted Vitruvius for the Renaissance. He was also the first modern architect to pay thorough attention to the Orders of classical architecture,526'540 although he was never sufficiently interested in architecture as structure to conceive of the Orders as anything other than decorative motifs. In his work, they are not essential load-bearing elements. Nevertheless, it was his interest in the Orders, combined with his study of antique remains, which led him to propose two solutions, both of which have been extensively used, to the problem of how to make a Christian building look suitably antique. His church designs are therefore of crucial importance in the history of architecture.

He began with the remodelling of the old church of S. Francesco at Rimini (begun 1450), where he hit on the idea of using a triumphal arch on a high podium for the facade: the one large opening and two small openings could then correspond with the entries to the nave and aisles. Sometimes, however, the aisles of a church are very much lower than the nave; at S. Maria Novella, Florence (begun 1458) he solved the awkward connection between nave-facade and aisle-facade by joining them with elegant scrolls, another feature which established a tradition in church architecture. S. Francesco at Rimini is more commonly called the 'Tempio Malatestiano',541 because it was designed for the tyrant of Rimini, Sigismondo Malatesta, as a memorial to himself and his humanist court. It is indeed a temple—almost a temple of fame.542 The imagery of the complex scheme, quite possibly worked out by Alberti himself, is replete with classical references, being a Neoplatonic monument to Sigismondo's 'own apotheosis as a sun-god'.539'543 This pantheon has sarcophagi (containing members of his court) in the outside walls under arched openings which, in their articulation, derive from Roman aqueduct design, or a simplified Colosseum. The crossing was to have been completed by a dome, which would probably have resembled that of the Pantheon in Rome. The implications of the whole scheme must have shocked the Church: Pius II publicly excommunicated Sigismondo for his presumptuous vainglory. Yet the idea of the building bore fruit in future centuries, as in the Pantheon in Paris or the Walhalla at Regensburg.

An alternative to the triumphal arch, conveniently used to express Christian beliefs of resurrection, is the temple front, which Alberti was also the first to use in church design, in his much altered S. Sebastiano, Mantua (begun 1460). Alberti's own design, as reconstructed by Wittkower, shows a high podium bearing a Greek-cross church, whose facade has its austerity modified by a broken pediment from


Ground plan of S. Sebastiano, Mantua, by Alberti.



West transept ofS. Andrea, Mantua, by Alberti, 1470ff.

which is suspended a window. A precedent for this arrangement was the triumphal arch at Orange, in Provence.* Neither S. Sebastiano nor the facades of S. Andrea at Mantua (begun 1470), which give a better idea of Alberti's intentions, make any use of full columns; both use a wall architecture of pilasters. At S. Andrea, the facade motif of a massive wall pierced by a barrel vault is increased in majesty by the giant pilasters which, rising apparently through three storeys, frame the exact square of the facade, and divide it into four exact and smaller squares. Inside the church, the same facade motif is continued down the aisle-less nave in the chapels to left and right, and unified by the barrel vault of the nave itself. Never before, in form or decoration, had the grandeur of ancient Rome been so potently evoked. It was works such as S. Andrea that were strongly to affect Bramante's conception of ancient Rome, which was, for him, as much the Rome of the great

*The matter is contentious. Heydenreich and Lotz reject Wittkower's reconstruction, call the work 'a bizarre revival of late antiquity', and speculate on the connections of Alberti's design with Early Christian works, pointing out its two-storeyed similarity to the Mausoleum of Theodoric at Ravenna. From either position, the church is clearly in an antique spirit.


baths as the Rome of the little temples and mausolea. Unlike Brunelleschi, Alberti throws over delicacy in favour of strength, and sacrifices Brunelleschi's light arcading for the massiveness of barrel vault and pillars. It is essentially in a structure like S. Andrea that the Plato and Aristotle of Raphael's School of Athens hold sway, when their philosophy is magnified by the very scale of the architecture. In De re aedificatoria (VII. ii), Alberti prescribed the vault as an essential element of the temple: the dome, as in S. Sebastiano and as projected for the Tempio Malatestiano, is a variation on the vault.

Alberti was an expatriate Florentine, and left examples of his architecture in Rome as well as in Rimini and Mantua. Elements of his style took root in Florence,544 where he founded a type of palace facade with his Palazzo Rucellai (begun 1446), a building which applies the Orders, as pilasters, to its facade, and divides the three storeys horizontally and mathematically by bold entablatures. Later palaces owe much to this example: the Palazzo Pitti (begun 1485), unusual in its monumental scale, might even be by Alberti himself, while the Palazzo Gondi (Giuliano da Sangallo, begun c. 1490) and the Palazzo S^rozzi (begun 1489) take the easier course of no Orders, deriving their manner from Michelozzo's Palazzo Medici (begun 1444). The great breadth of Alberti's ideas on architecture and town-planning545 bears fruit in the work of Bernardo Rossellino, whose ideal city of Pienza, the birthplace of Pius II (Enea Silvio Piccolomini), was actually in part built (1460-2). The Palazzo Piccolomini there is a variation on the Palazzo Rucellai, which Rossellino had in fact helped to build. The various buildings of Pienza are arranged almost as a stage-set, but unfortunately Pius II died before the grand design could be completed. Then again, Alberti also provided the first barrel vault in Florence, in the Chapel of the Holy Sepulchre within the church of S. Pancrazio (c. 1455-60).

Alberti was not what we would call today a 'professional' architect, and presumably he did not take students. But the career of Giuliano da Sangallo (1443-1516), a Florentine architect, demonstrates the spread of Alberti's ideas. Giuliano was apparently the first Florentine architect to receive a thorough training in Rome; he arrived there in 1465, and worked until at least 1472. Some of his earlier designs—for example, La Madonna delle Car-ceri, Prato (begun 1484)—are an extension of the manner of Brunelleschi, but his mature work benefits from his study of antique Roman architecture. Such is the case with his villa for Lorenzo de' Medici at Poggio a Caiano, near Florence (begun c. 1482), and with his Palazzo della Rovere at Savona, of the next decade. If we examine these works together with the grandiose designs in his sketch-books (including one for an immense palace for the King of Naples), Giuliano's confident use of antique Roman forms becomes even clearer. Poggio, for example, is probably the first villa where a Roman temple portico is introduced, albeit


recessed; the salone displays a coffered barrel vault designed and constructed after the manner of the antique. The Naples project has a portico standing proud of the block, and is completely unlike any Florentine scheme; its whole organization is Roman, and shows a concern for balance and mathematical division which Palladio, drawing on similar sources, used in the next century. On the Palazzo della Rovere, furthermore, Giuliano takes up the classical Orders, and applies them, as Alberti did, as pilasters to the austere facade. In the Palazzo Scala, in Rome (1472/80), Giuliano attempted to recreate an antique palace, complete with reliefs in a totally antique manner. He might therefore be called a link between the delicate Florentine manner of Brunelleschi and, through the antiquarianism of Alberti, the grand manner of the High Renaissance in Rome.


Various ground plans for houses, using centrally-planned Studi es of round temples. Francesco di Giorgio: Codex Saluzzo, courtyards. Francesco di Giorgio: Codex Magliabechiano, fol. fol. 84r. Turin, Biblioteca Reale, MS Saluzzo 148. Francesco 20r. Florence, Biblioteca Nazionale, MS II.1.141. provides a link between Alberti and Leonardo.


classicaltrad3-12.jpg classicaltrad3-13.jpg classicaltrad3-14.jpg classicaltrad3-15.jpg
Leonardo da Vinci, drawings of central-plan churches. Paris, Institut de France.

Leonardo and Bramante: the High Renaissance in architecture

Another link between Alberti and the High Renaissance is forged by the extraordinary treatise of Antonio Filarete (1461/4). This not only copies parts of Alberti's De re aedificatoria (which was not printed until 1485, in Florence), but also adds an immense disquisition on an ideal city called Sforzinda, named after his Milanese patron Francesco Sforza (he had arrived there from Rome by 1456). The book is important because Filarete's enthusiasm for classical antiquity and his dismissal of the 'modern' style no doubt affected the outlook of both Leonardo and Bramante, both of whom lived for a time in Milan, the former from 1482 to 1499, the latter from 1481 to 1499.

Leonardo is certainly a crucial figure in the development of Bramante's art, but the extent of his actual involvement with real buildings is unclear.551'553 Leonardo drew centrally planned churches in his notebooks; these


derive from the Florentine tradition of Brunel-leschi, particularly the cathedral dome and the S. Maria degli Angeli project, and Giuliano da Sangallo's sacristy for S. Spirito. Furthermore, we know that Leonardo took an interest in ancient ruins: he describes an antique temple in the Codex Atlanticus (fol. 285) and from the anonymous Codex Bramantinianus it is certain that he possessed a book containing drawings of ancient buildings (which he might have drawn himself). Several links have been suggested between Leonardo and Bramante. They were, of course, working in Milan near each other, the one on the Last Supper, the other on the fabric of the same church and convent of S. Maria delle Grazie. Pedretti has recently shown that among Leonardo's drawings are details of the Grazie, and he has concluded that Leonardo took an active part in the actual process of design.555 Leonardo's proven interest in engineering problems makes the suggestion convincing, for he could have offered his friend much help in an area where subsequent events showed Bramante to be woefully inadequate (particularly the cracking of the piers of the new St Peter's). The same scholar has even suggested that, because of his centrally planned church designs, Leonardo might well have had a hand in the design of the Tempietto in Rome, which is now dated as post-1510.554 The supposition is supported by the close friendship of the two men, revealed by the Antiquarie Prospetiche Romane, dedicated to Leonardo, of which Bramante seems to have been the author.547 But it is possible to argue the matter the other way, as Heydenreich does, stating that Leonardo 'derived the values of ancient monumentality through the classic style of Bramante'.548

Bramante's architectural education must have begun in his home town of Urbino, where he would have seen the marvellous and sober buildings in the frescoes and panels of Piero della Francesca. We can assume that, on his way to Milan, he would have passed through Mantua, and admired the works of Alberti. Once in Milan, his manner shows connections with the spirit of Alberti's work. Bramante's only centrally planned building in Milan was the little chapel of S. Satiro next to the church of S. Maria presso S. Satiro (where he also worked). His inspiration here was the original foundation of S. Satiro, a ninth-century structure based on Early Christian examples, which it was his task to remodel. The ground-plan of the medieval church appealed to him, and he adapted it for his sacristy to the same church (1470-C.1482). As for the exterior elevation of S. Satiro, the clarity with which he indicates the cross-within-the-circle is completely Florentine, whilst the pilaster-and-niche motif probably derives from Brunelleschi's S. Maria degli Angeli. He was to use it again when he went to Rome.

Equally significant in Bramante's search for an antique style is his interior modelling of S. Maria presso S. Satiro. The trick perspective of the apse (he lacked the space for anything other than an illusion) is less important than the massive barrel vault which weighs down upon it and echoes the vaulting of the nave. The articulation of S. Maria, with its use of

The sacristy to S. Maria presso S. Satiro, Milan, 1470-c. 1482, remodelled by Bramante.

classicaltrad3-16.jpg classicaltrad3-17.jpg

Rome, Cloister of S. Maria della Pace, by Bramante, 1504.

pilasters and great pillars, is in the tradition of Alberti's S. Andrea at Mantua. And the architectural detailing of this beautiful church reveals a close study by Bramante of the Early Christian architecture in which Milan then abounded. But the greatest similarity with Alberti's style is shown in the tribune of S. Maria delle Grazie, which is very close in style to S. Sebastiano at Mantua. Not only is the Milanese work in effect a centrally planned design; a comparison may also be made with the Tempio Malatestiano at Rimini, since the tribune was intended as a mausoleum, in this case for the Sforza family. This links the building not only with Alberti and with antique prototypes, but also with Leonardo's projects for a Sforza mausoleum.549

In 1499 Milan fell to the French, and Bramante moved to Rome where, after making the cloister of S. Maria della Pace (finished 1504), he designed the Tempietto550'556 at S. Pietro in Montorio, the original design of which is preserved in Serlio's woodcut. This shows the temple as the centrepiece of a courtyard


which, also with columns and niches, echoes the structure itself, and provides a clear demonstration of Bramante's concern with geometry. The scheme is similar in plan to the Early Christian church of S. Lorenzo, Milan, which Bramante would have known well. The columns of the Tempietto actually are antique; Bramante combines them with detailing and forms of his own invention to make a building inspired by antique temples and, in its proportions, by a basically Greek notion of geometry. Yet in spite of these derivations it remains wholly original. The Tempietto as we see it today is not as Bramante designed it: the dome has been heightened from the original half-sphere, and the air of blank-eyed austerity would perhaps have been mitigated by statues in the niches. This small temple was to assume great significance during the Renaissance, for it was considered the equal of anything erected during Antiquity. Palladio, for example, in his Quattro Libri dell'Architettura (1570) refers to Bramante as the man who was the first 'to bring back into the light of day the good and beautiful architecture that had been hidden since the time of the ancients'. He, like Serlio before him, therefore includes the Tempietto among his accounts and illustrations of antique buildings. His ground plan is different from Serlio's; he makes the temple look more like the mausolea or Christian martyria from which it is derived. Palladio himself was a great sketcher of ancient ruins, so perhaps this is an example of wishful thinking.

The Tempietto had presented a relatively simple problem which could be solved by recourse to the readily available formal sources, ranging from Early Christian martyria to the rooms of Roman palaces like the Domus Aurea or Hadrian's Villa, and the numerous tombs and nymphaea scattered around the Campagna. But when, in 1503, Bramante was asked to design a new church of St Peter to replace the Early Christian basilica, problems abounded.552 Old St Peter's had consisted of a centrally planned martyrium with a long nave, which made Bramante's choice of a central


Bramante's Tempietto of S. Pietro in Montorio, Rome (after 1510), and BELOW in plan, as drawn by Sebastiano Serlio in the Architettura, Venice, 1584.




Bramante's project for St Peter's, Rome, redrawn from Caradosso's medal of 1506.

plan all the more likely. His idea, known from a drawing in the Uffizi, was to incorporate a Greek cross within a square, and to divide the interior symmetrically into a series of smaller, centrally planned areas. The entrances were to be through columned porticoes on each of the four sides of the square, and, as the foundation medal by Caradosso of 1506 shows, a tower at each corner would have set off the semicircular dome. It was the dome which was to cause the trouble. Bramante intended to support it on a drum decorated by columns, to build it up in 'steps' like the Pantheon dome, and then to crown it with a lantern. But Bramante's inexperience of work on such a huge scale necessitated his considerably enlarging and strengthening the supporting piers at the crossing. After his death in 1514, when Raphael took over, it was suggested that the central plan should be abandoned, and a whole series of nave designs based on the Latin cross were put forward, all of which had to incorporate the great piers for the crossing, which had already been built.

Had Bramante's first project been completed, how antique would it have looked ? * The answer must lie in the relation of the dome

*Perhaps we should ask the question with one eye on Bramante's part in the destruction of the old basilica: this was too rickety to survive, but what happened to most of the fittings and monuments?




The design for St Peter's by Antonio da Sangallo the Elder (redrawn).

to the four facades, and the only evidence we have for this is the foundation medal mentioned above. This, like some of the designs by Leonardo which surely influenced Bramante's conception, shows a set of four porticoes reaching to drum level, with the transition to the lower entrance porticoes being made by four smaller domes, each on its own drum, all providing an echo of the much larger central dome. Also, of course, the antique intention of the structure would have been made clear by its purpose: to shelter the tomb of St Peter (let alone that of Julius II!)—in other words, to act as a gigantic martyrium.

It is impossible to tell whether Bramante's facades would have been quite as austere as the medal implies: blank walls predominate, and nothing fights for attention with the dome. This was to have been a single-shell dome, like the Pantheon; a double-shell in the Brunel-leschi manner would have been much lighter. But whereas the Pantheon was supported to ground level by its thick walls, Bramante tried to maintain his dome on four piers, and misjudged his calculations. Some idea of how Bramante's St Peter's might have looked, but with a double-shell, not a saucer dome, can be


Bramante's project for St Peter's, Rome, from a drawing in the Uffizi, Florence.

gained from the two churches obviously modelled upon it. S. Maria della Consolazione at Todi, by Cola da Caprarola (and others, begun 1508), has similar giant pilasters on the piers supporting the dome, and looks even more like one of Leonardo's sketches. La Madonna di S. Biagio, at Montepulciano, by Antonio da Sangallo the Elder (begun 1518), is more impressive because in it a much bolder use is made of the classical Orders and therefore the work achieves an impression of antique weight


Plan of S. Maria della Consolazione, Todi, by Cola da Caprarola.

The plan might be for the east end, or, equally, half of a central design.

and seriousness lacking in the church at Todi, which is elegant rather than strong. La Madonna di S. Biagio, although built to a Greek cross plan, is directional: two towers, only one of which was completed, were to flank the main entrance. They are similar in design to those on the Caradosso medal, fully articulated by the Orders, and welded into the facade storey by storey, which presumably would have happened at St Peter's too.

In fact, La Madonna di S. Biagio might represent a slightly later stage of work on St Peter's, namely the period when Raphael was in charge and produced a model (now lost), plans and elevations. It is interesting to note that Antonio's son worked on the fabric from 1516 to 1520, under Raphael. Basically, Raphael's design changed St Peter's permanently from a Greek-cross martyrium into a nave church adapted for processions. A drawing by Raphael shows that the facade was intended to have a two-storeyed portico, the storeys linked by a giant Order (to echo the piers of the crossing?) and supporting a triangular pediment. This would have been the exact height of the barrel-vaulted nave, itself graced by giant pilasters. To such elements, evidently inspired by Alberti's S. Andrea at Mantua, Raphael added flanking towers of a festive nature, articulated by the Orders and decorated by niches and swags.



The scale of St Peter's is indicative of the tendency toward much larger buildings both civil and religious during the sixteenth century. A different architectural vocabulary from that perfected by Bramante in the Tem-pietto was required in order to cope with the increased scale of construction. This alone, however, cannot begin to account for the development of 'Mannerism' in architecture, which can best be understood as a reaction to the elements of classicism as erected in the High Renaissance canon. Basing themselves on the example of Antiquity, architects like Alberti and Bramante had evolved a structural system which was rational, calm and dignified: columns supported entablatures, the design of window openings was consistent with antique practice, and the whole system of architectural detail found its justification in Antiquity. Architecture, for such men, was a pleasure for the mind, which could delight in both the calm perfection of its mathematics and the associations it evoked.

By contrast the Mannerist architects, particularly Giulio Romano and Michelangelo, looked to Antiquity not as the seat of all



Michelangelo's mannerism: The vestibule of the Laurentian Library, Florence, begun in 1525


In Giulio Romano's mannerist facade for the courtyard of the Palazzo del Te at Mantua, all is restless confusion. Note the dropped 'keystones' breaking the architrave.

authority but rather as a vast mine of examples which could be drawn upon to satisfy an approach to art which is emotional, not rational. Nevertheless, it is obvious that Mannerism is a development of High Renaissance classicism, and in order to appreciate the effects at which the Mannerists aim, a knowledge of the syntax of classicism is essential. Giulio Romano's buildings create an impression of emotional instability and cap-riciousness only because of this underlying contrast with classicism, and we can therefore say that they play on the notion of classical order and reason for their effect.

All Giulio's important architectural works—he was also Raphael's chief assistant in painting—are in Mantua, where he ruled as architect, designer and painter from 1524 until his death in 1546. He built himself a strange house (c. 1540) which, like the Cortile della Cavallerizza of the Palazzo Ducale (c. 1539) breaks all the rules of classicism. But the most amazing achievement was the Palazzo del Te, a villa on the outskirts of Mantua for which Giulio was almost totally responsible: it is filled with his frescoes, some of which are pseudo-antique, others shockingly illusionis-tic. As with all Mannerist creations, at the

Palazzo del Te the total effect is one of unease and confusion. The eye is frequently made to doubt what it is in fact seeing, and nowhere more so than in the courtyard where the deviations from the classical canon are quite deliberate. For a start, none of the four sides match, nor do they appear to be completely finished, thanks to Giulio's startling use of heavy, rusticated blocks in contrast with expanses of smooth blocks. Occasionally, stones of the architrave have slipped and left gaps in the frieze, where the triglyphs have slipped as well. Strangest of all is the section where a pedestal supports both the high Order

A corner of the Medici Chapel (begun 1519) in S. Lorenzo, Florence, where complicated motifs jostle one another for breathing space.


bearing the entablature, and a lower Order bearing the pediment over the central opening: the point of the pediment struggles for room and the supporting columns teeter in the midst of their more robust fellows.

Michelangelo's desire for emotional force is as great as Giulio's, but his greater inventiveness assures his work of a larger impact. The work of both men was extensively imitated: Mantua, that antiquarian paradise, was much visited until the fall of the House of Gonzaga, but Michelangelo's buildings were easily accessible in Florence and Rome. The New Sacristy of S. Lorenzo, Florence (the Medici Chapel, begun 1519) and the vestibule of the Laurentian Library attached to the same church (begun 1525) present the epitome of Michelangelo's maltreatment of classicism for expressive ends. Instead of making the library vestibule a light and airy room he presents the visitor with a combination of antique elements, all altered, which are almost threatening because of the way in which they are combined. The staircase (by Ammanati after a design by Michelangelo) threatens to flood the room with steps. But it is the wall architecture which is so strange, principally because the exact level of the wall surface and the relation of the columns set within it to that surface cannot be worked out by logic or by eye alone.(*The fact that Michelangelo had to work within an awkwardly thin set of foundations cannot afiect our reactions to the design, although that design depends partly from such considerations.) If the columns support the entablature, why are they recessed into the wall and themselves supported only on gigantic volutes, which never knew this task in Antiquity ? If they do not, why does the thick wall between them support nothing other than the evidently paper-thin area of the window-openings ? This system of solid above void and void above solid is contrary to reason, particularly when everything described, except those volutes, is far above the head of the spectator, who can find nothing of human scale with which to relate. Essentially the same effect is created in the Medici Chapel: architectural elements lead a wilful life of their own, pushing and squeezing. A comparison with the


Old Sacristy, by Brunelleschi, demonstrates the wilful misuse of the vocabulary of classical architecture.

Michelangelo cannot, however, be cast as the villain of the piece. Certainly, the attitude to the 'correct' use of classical elements was never the same again, but it is equally true to say that Michelangelo broadened the emotional vocabulary and thereby extended the life-span of the classical language of architecture. One device he invented was widely used in succeeding generations: the giant pilaster. In his redesigning of the Capitoline Hill (perhaps begun 1539), Michelangelo wished to give a unity to the piazza, and attained this by endowing the facing Palazzo del Senatore and Palazzo dei Conservatori with identical facades. Crowned by a balustrade, and including a loggia at ground level, the two-storeyed buildings are articulated by the Corinthian Order, but, unlike any secular building hitherto erected, these palaces have their storeys united by a giant Order which supports the imposing entablature. The ground storey has its own scaled Order as well. Although some triumphal arches use a similar scheme, nothing quite like this had been built in Antiquity; however, Alberti's S. Andrea at Mantua, with its giant triumphal arch, provides a similar solution. In other words, Michelangelo solved the problem of linking two or more storeys without recourse to the usual 'Colosseum' solution of using one Order per storey, each with its own apparatus. The solution provides a new way of balancing horizontals and verticals, but represents a totally new departure in scale and monumentality. This can best be seen by comparing Bramante's plan for the new St Peter's with the design of Michelangelo (1546); while adhering to Bramante's general scheme, Michelangelo greatly increased the size of the supports for the dome, and added a portico with giant columns, and a giant Order of Corinthian pilasters both inside and outside.

It is at this point that problems of nomenclature intervene to demonstrate that 'Mannerism', while possibly a suitable concept in

Palazzo del Museo Capitolino, Rome 1539ff., designed by Michelangelo.

classicaltrad3-30.jpg classicaltrad3-31.jpg

Mantua Cathedral, rebuilt in 1545 by Giulio Romano, and inspired by the classicism of Early Christian basilicas. Compare this with his more fractured Palazzo del Te,p. 141.

painting, has little to recommend it in architecture, unless we understand it as simply a different approach to Antiquity. Perhaps the work of Michelangelo, Vignola and Pirro Lig-orio founds a new orthodoxy in architectural design. Thus Vignola (1507-73) is not only the architect of the Palazzo Farnese at Caprarola (begun 1559), the Villa di Papa Giulio (begun 1551) and the church of the Gesu (begun 1568), but also the author of the Regola delle Cinque Ordini (1562) which, as he says in the preface to this collection of illustrations, is derived from his examination of antique architecture. Pirro Ligorio, builder of the Villa d'Este (c. 1565-72) and the Casino di Pio IV in the Vatican (begun 1559), would also have been outraged had anyone told him that he was mangling the very spirit of antique architecture. He would have pointed to his painstaking study of Antiquity, and shown how his own work depended on it. Nomenclature is also difficult for seventeenth-century Italian architecture: in painting, it is easy to make a precise distinction between Mannerist and Baroque, but such is not the case with, for example, the standard comparison between the facade of Vignola's Gesu and Carlo Maderno's S. Susanna (1597), both in Rome. Any differences are only of degree, not of type. The Gesu facade, in the ordering of its constituent parts, begins a tradition at least as important as Alberti's Tempio Malatestiano, which we shall meet again not only in seventeenth-century Rome, but also in Paris, particularly in Francois Mansart's Val de Grace (begun 1645). Or, to be more accurate, the Gesu restates a theme initiated by Alberti's S. Andrea at Mantua, both in its facade and in the arrangement of the interior.


Perhaps the best way of underlining the fact that the classical tradition in architecture is both more continuous and, indeed, more long-lived than it is in either painting or sculpture is to study the career of Andrea Palladio. The facade of the Redentore, Venice (begun 1576),558 a mature work, is clearly an extension of the temple-front idea of Alberti, but, equally clearly, the multiplication of pediments one behind the other produces an excitement and

Section west-east of the Redentore, Venice, by Andrea Palladio. From Bertotti-Scamozzi, Lefabbriche . . . di Andrea Palladio, IV, 1783.

effects of light and shade which the sober Alberti did not seek. Inside, Palladio moulds the space of his white vaults by techniques derived from his study of the great Roman baths; the screen of columns in the chancel, half-revealing a room beyond, are from the same source. Palladio uses walls as the Romans of the Empire used them, as elements to be gouged out by arches and niches, and decorated by columns and pilasters. Undeniably, therefore, ancient Rome is the very foundation of his architectural vocabulary, enhanced by a good knowledge of his Renaissance forebears.560

Balancing the splendour of Palladio's forms is another quality which is the main reason for his great popularity, and for the ease and frequency with which he was imitated: mathematical clarity in both plan and elevation, stemming from the Renaissance belief that beauty could be attained through the use of geometry and measurement. His fame was spread by his textbooks. Le Antichita di Roma (1554) was to become the most popular of all guides to the antiquities of the city. His illustrations to Daniele Barbaro's edition with commentary of Vitruvius (1556) underlined

Baths of Diocletian, Rome. Etching by Alo Giavonnoli, early seventeenth century.


his involvement with the antique. Above all, his own / Quattro Libri dell' Architettura (1570), the most popular of all architectural treatises except Vitruvius', constantly reminds the reader that the author's own buildings were largely intended as reconstructions of antique architecture, based on his reading of the Vitruvian text. Such an avowal was obviously good publicity, perhaps not to be taken too seriously; but Palladio does find authority for his multi-pedimented facade to the Redentore in Vitruvius' description of his own basilica design at Fano, which refers to the 'double arrangement of gables'. His villas, the most influential part of his work, also use the temple front, because Palladio believed that this was also a feature of antique domestic architecture. As well as worshipping in antique temples, the classically minded intelligentsia of Europe could henceforth live in them as well, choosing the variation which pleased them most from the great range of alternatives illustrated in / Quattro Libri.

How far Palladio based his villas on actual classical remains is not clear, but he did produce a practical form of country house, well fitted to be the centre of a farm, and



Elevation and section, and plan of the Villa

Rotonda by Palladia, from his I Quattro Libri


capable of adaptation by the landed gentry of northern Europe, particularly those of eighteenth-century England, after the emotionalism of his typically Venetian effects of light and shade had been chastened.

Much of Palladio's early manner depends on the achievements of the High Renaissance in Rome. The palace type represented by Bramante's House of Raphael (after 1510) is manifestly the source for Palladio's Palazzo Porta Colleoni in Vicenza (c. 1550). His church facades are in the same tradition, partly because they go back to Vitruvius and to the Pantheon (which also has intersecting pediments). True, he sometimes does violence to correct classical usage (the eighteenth-century critic Francesco Milizia wrote of his work as 'bizarre' and of 'impure taste'),557 but the exuberance of some of his motifs is less important than his general feeling for symmetry and clarity. He conceived of architecture as something rational, which obeyed rules: if a work is created according to rules, it can be imitated and taught, assuming that the same basic precepts are imparted. This / Quattro Libri did for classicizing Europe. Palladio's feeling for clarity is boldly in evidence in some of his earlier buildings, like the Villa Godi, where he dispenses completely with antique details, and cannot therefore rely on their cosmetic qualities. With most of his structures, we can strip away the Orders and their accessories in the mind's eye, and produce an armature very similar to some of the buildings of the Neoclassical period. That such a process is possible demonstrates the classicism of Palladio's style. His later work, following his visit to Rome in 1554 when he would surely have learned of Michelangelo's project for the Capitol, becomes much more monumental, the giant half-column or pilaster is now used, following Michelangelo's example, as at Vicenza in the Palazzo Valmarana (1566) or the Loggia del Capitanio (1571). In the Loggia, the festive decoration with which Palladio covers the whole design is deliberately reminiscent of a triumphal arch, for the building was intended as a symbol of victory against the Turk. Even more powerful is the Palazzo Thiene in Vicenza (facade 1556-8) where he used the favourite motif of Giulio Romano (whose work he must greatly have admired), namely the rusticated base pierced by plain windows. On the piano nobile, the windows are articulated by great projecting blocks of semi-hewn stone which almost mask the architectural frame of col-onettes and pediment. The whole composition, held down by a powerful cornice, has much greater power than, for example, his casing of the Basilica in Vicenza (1546/9) which, with its succession of 'palladian' windows (actually invented by Bramante and popularized by Sansovino in Venice, and by Serlio's book), is much lighter and less imposing. We might say that the Palazzo Thiene is more personal and hence less imitable, at least as far as palaces are concerned, because the forms which Palladio borrows from Rome ancient and modern are specifically urban in intention; they have to be surrounded by other buildings to give them scale. A hundred years later, we find Bernini proposing a design for the rebuilding of the Louvre (the 'First Design') which clearly has close connections with the Loggia del Capitanio, just as his piazza for St Peter's has touches of Palladio in its light and movement.559

With the coming of the seventeenth century and architects like Maderno, Borromini, Bernini and Rainaldi, this chapter of Italian classicism draws to a close. Classicism— simple, measured, intellectual—gives way in the seventeenth century to the Baroque, a style which is exuberant, expansive and sensuous. As I have already emphasized, it is not, for all that, estranged from the example of Antiquity. Antiquity is modelled by each age after its own image, and to follow the classical tradition in architecture we must pass to France (pp. 177ff., below), whose artists and architects, at the end of the sixteenth century and the beginning of the seventeenth, look to the style of the High Renaissance in Italy, and adopt and develop it into a national style of their own.


Main facade of the Palazzo Thiene, Vicenza, by Palladio. From Bertotti-Scamozzi, Le fabbriche . . ., I, 1776.