JOHN SUMMERSON The Classical Language of Architectt1re

THE M.l. T. PRESS Massachusetts Institute ofTechnology Cambridge, Massachusetts I. The Essentials of Classicism

I must begin by assuming some general knowledge. Like, development of the classical language of but for instance, knowing that St. Paul's Cathedral is a classical about its nature and its use - its use as the common archi­ building while Westminster Abbey is not; that the British tecturallanguage, inherited from Rome, ofnearly the whole Museum is a classical building while the Natural History civilized world in the five centuries between the Renaissance Museum at South Kensington is not. That all the buildings and our own time. round Tra&J.gar Square - the , St. Martin's­ Very well- from now on we can be more precise. Let us in-the-Fields, Canada House and even Gust) South Africa look at this word 'classical' as applied to architecture. It is a House are classical, that all the public buildings in White­ mistake to try to define classicism. It has all sorts of useful hall are classical; but that the Houses of Parliament are not. meanings in different contexts and I propose to consider two Elementary distinctions, and you may think at once that I meanings only, both of which will be useful throughout am going to deal in superficialities. When is a classical build­ these talks. The first meaning is the most obvious. A classical ing not a classical building ~ Does it really matter? Are not building is one whose decorative elements derive directly or the important qualities ofarchltecture deeper than and inde­ indirectly from the. architectural vocabulary of the ancient pendent ofsuch stylistic nomenclature? They are. Neverthe­ world - the 'classical' world as it is often called: these ele­ less, I cannot reach all the things I want to say in these talks ments are easily recognizable, as for example offive without first isolating all the buildings which are, prima standard varieties, applied in standard ways; standard ways jacie, classical from all the others. I shall be talking about of treating door and window openings and gable ends and architecture as a language and all I want to assume at the standard runs of mouldings applicable to all these things. moment is that you do recognize the Latin of architecture Notwithstanding that all these 'standards' are continually when you hear - that is, see - it. departed from they do remain still recognizable as standards The Latin of Architecture - that brings me to another throughout all buildings that may be called classical in this general knowledge assumption. has its sense. roots in antiquity, in the worlds ofGreece and Rome, in the That, I think, is one fair description ofwhat classical archi­ temple architecture ofthe Greek world and in the religious, tecture is, but it is only skin-deep; it enables you to recognize military and civil architecture of the Romans. But these the 'uniform' worn by a certain category of buildings, the talks are not going to be about the architecture of Greece category we call classical. But it tells you nothing about the and Rome - they are not going to be about the growth and essence ofclassicism in architecture. Here, however, we have

7 got to be r~ther careful. 'Essences' are very elusive and are ify as a 'classical' building? The answer must, I think, be 'no'. often fOWld, on enquiry, not to exist. Nevertheless, embed­ You can say, in describing such a building, that its propor­ ded in the history ofclassical architecture is a series ofstate­ tions are classical, but it is simply confusing and an abuse of ments about the essentials of architecture and these are in terminology to say that it is classical. The porches ofChartres agreement over a long period, to the extent that we may Cathedral are, in distribution and proportion, just about as say that the aim ofclassical architecture has always been to classical as you can get, but nobody is ever going to call them achieve a demonstrable harllJony ofparts. Such harmony has anything but Gothic. And one could cite plenty of other been felt to reside in the buildings ofantiquity and to be to examples of the Gothic system being closely analogous to a great extent 'built in' to the principal antique elements ­ the classical. It is, by the way, a great mistake to think of especially to the five'orders' to which we shall come pre­ Gothic and Classic as opposites; they are very different but sently. But it has also been considered in the abstract by a they are not opposites and they are not wholly unrelated. series oftheoreticians who have demonstrated that harmony It is nineteenth-century romanticism which has made us put analogous to musical harmony in a structure is achieved ~y them in totally different psychological camps. People who proportion, that is to say by ensuring that the ratios in a say they 'prefer' Gothic to Classic or Classic to Gothic are, I building are simple arithmetical fWlctions and that the ratios suspect, usually the victims of this nineteenth-century mis­ of all parts of the building are either those same ratios or interpretation. The fact is that the essentials ofarchitecture ­ related to them in a direct way. A vast amoWlt of preten­ as expoWlded by the Renaissance theorists - are to be fOWld tious nonsense has been written about proportion and I have expressed, consciously or Wlconsciously, throughout the ar­ no intention ofgetting involved in it. The Renaissance con­ chitectures of the world. And while we must incorporate cept ofproportion is fairly simple. The purpose ofpropor­ these essentials in our idea of what is classical we must also tion is to establish harmony throughout a structure - a accept the fact that classical architecture is only recognizable harmony which is made comprehensible either by the con­ as such when it contains some allusion, however slight, how­ spicuous use ofone or more ofthe orders as dominant com­ ever vestigial, to the antique 'orders'. Such an allusion may ponents or else simply by the use ofdimensions involving be no more than some groove or projection which suggests the repetition of simple ratios. That is enough for us to go the idea ofa or even a disposition ofwindows which on with. suggests the ratio ofpedestal to , column to entabla­ There is, however, one point about this rather abstract ture. Some modem buildings - notably those of the late conception of what is classical and it may be put as a ques­ Auguste Perret and his imitators - are classical in this way: tion. Is it possible, you may ask, for a building to display that is to say, they are thought out in modem materials but absolutely none of the trappings associated with classical in a classical spirit and sealed as classical only by the tiniest architecture and still, by virtue ofproportion alone, to qual­ allusive gestures. In the last talk. in this series I shall have

8 more to say about all this. In the meantime the thing which kind to have survived from antiquity and for that reason has it is quite essential for us to understand before we go any been accorded enormous veneration. Vitruvius was not him­ further is this question of the orders - the 'Five Orders of self a man of any great genius or literary talent or indeed ­ Architecture'. Everybody has heard of them, but what ex­ for all we know - of architectural talent. The thing about actly are they~ Why are there five and not four, or sixteen his treatise is that it rounds up and preserves for us an or three hundred and twenty-six~ immense quantity oftraditional building lore - it is the code One thing at a time. First, what are the orders~ On the of practice of a Roman architect of the first century A.D. endpaper of this book you will frod a very clear diagram of enriched with instances and historical notes. the . It consists, you see, of a temple column In the course ofVitruvius' third and fourth book~ he des­ standing on a pedestal and carrying on its head the archi­ cribes three ofthe orders - Ionic, Doric and Corinthian - and trave, and cornice, those elements which are collect­ gives a few notes on another, the Tuscan. He tells us in which ively called the . Then, in Plates I and 2 you see part ofthe world each was invented. He relates them to his the Doric order again, with its four companions; it is the descriptions oftemples and tells us to which Gods and God­ second from the left, with the Tuscan to the left of it, the desses each order is appropriate. His descriptions are by no Ionic, Corinthian and Composite to the right. There are two means exhaustive, he gives no fifth order, he does not present sets here - one of 1540 on the left of the page (plate I), the them in what we think ofas the 'proper' sequence (Tuscan, other more than a hundred years later (Plate 2), but they are Doric, Ionic, Corinthian) and - most important - he does in principle the same thing. An 'order' is the 'column-and­ not present them as a set of canonical formulae embodying superstructure' unit ofa temple colonnade. It does not have all architectural virtue. That was left for the theorists of the to have a pedestal and often does not. It does have to have an Renaissance. entablature (columns are meaningless unless they support In the middle of the fifteenth century, fourteen hundred something) and the cornice represents the eaves ofthe build­ years after Vitruvius, the Florentine architect and humanist, ing finishing offthe slope ofthe roo£ Leon Battista Alberti, described the orders, partly with refer­ Now, why are there five orders~ This is a little more diffi­ ence to Vitruvius and partly from his own observations of cult and it is necessary to glance back to some origins. The Roman remains. It was he who added, from observation, a earliest written description of any of the orders is in Vitru­ fifth order - the Composite - which combines features of vius. The name ofthis Roman author will crop up frequently the Corinthian with those ofthe Ionic. But Alberti was still in these talks and this is the moment to introduce him. He perfectly objective and Vitruvian in his attitude. It was Sebas­ was an architect ofsome consquence in the reign ofAugustus tiano Serlio, nearly a century later, who really started the and wrote a treatise in ten books: De Architectura, which he orders - the five orders now - on their long career of canon­ dedicated to the Emperor. This is the only treatise of its ical, symbolic, almost legendary, authority. I am not sure

B 9 that Serlio quite meant to do this but that is what he did. verbs in the grammar ofthe Latin language. Serlio was a man ofthe High Renaissance, an exact con­ This very effective and, quite literally, dramatic gesture of temporary ofMichelangelo, a near contemporary ofRaphael Serlio's was not lost upon his successors, and the fIve orders and an associate ofthe architect-painter Baldassare Perruzzi as a 'complete set', all deviations from which were question­ whose designs he inherited. He built a few quite important able, passed from hand to hand. Nearly all seventeenth and buildings but his greatest service to architecture was to com­ eighteenth century primers of architecture start in the same pile the fIrst full-scale fully illustrated architectural grammar way, with a plate ofthe five columns and ranged ofthe Renaissance. It came out as a series ofbooks. The fIrst side by side - Bloem in Switzerland, De Vries in Flanders, two appeared in Venice, the later books in France under the Dietterlin in Germany, Freart and Perrault in France and, in patronage of Frans;ois Ier. The books became the architect­ England, Shute, Gibbs and Sir William Chambers. George ural bible ofthe civilized world. The Italians used them, the Gwilt's edition ofChambers carries us up to 1825 and ifyou French owed nearly everything to Serlio and his books, the go from this to the same author's Encyclopedia ofArchitecture Germans and Flemings based their own books on his, the and follow this work to its latest edition in 1891 you will Elizabethans cribbed from him and Sir Christopher Wren fmd it still being stated there that 'in the proper understand­ was still fInding Serlio invaluable when he built the Shel­ ing and application of the orders is laid the foundation of donian at Oxford in 1663. architecture as an art'. As little as forty years ago when I was Serlio's book on the orders starts with an engraving - the a student in School at University College, Lon­ very fIrst ofits kind (plate I) - in which all fIve orders are don, it was taken for granted that one's first task as a student shown standing side by side like ill-assorted nine-pins ranged was to draw out in great detail three ofthe classical orders. according to their relative slimness - that is to say according Now there are two important points about aU this. The to the ratio oflower diameter to height. A11 are on pedestals. fIrst is to realize that although the Romans clearly accepted The stubby Tuscan is on the left; then the similar but slightly the individuality ofDoric, Ionic and Corinthian, and knew taller Doric; the elegant Ionic; the lofty, elaborate Corinth­ about their historical origins, it was not they who embalmed ian; and finally the still more elongated and further enriched and sanctified them in the arbitrary, limiting way with which Composite. In the text accompanying this plate Serlio ex­ we are familiar. The second point is to realize the immense plains himsel£ He says that just as the ancient dramatists used importance, for the whole ofarchitecture since the Renais­ to preface their plays with a prologue telling audiences what sance, of this process of embalming, of canonization. The it was all going to be about, so he is putting before us the orders came to be regarded as the very touch-stone ofarchi­ principal characters in his treatise on architecture. He does it tecture, as architectural instruments of the greatest possible in a way which makes the orders seem as categorical in the subtlety, embodying all the ancient wisdom of mankind in grammar of architecture as, say, the four conjugations of the building art - almost, in fact as products ofnature herself

10 And this is where the modem eye must often confess itself and sheer personal invention on the other. Somewhere be­ defeated. Unless you really know your orders and can recog­ tween the extremes have been the types composed and pub­ nize, at a glance, a Tuscan according to Vitruvius, a Cor­ lished by the great theorists - Serlio, ofcourse, :first in IS37, inthian from the temple of Vespasian or an Ionic from the then Vignola in Is62, Palladio in IS70 and Scamozzi in 161S. temple ofSaturn or the rather odd Composite concocted by These have had a normalizing effect all over the world. But Serlio from the Colosseum you will not appreciate all the in all the centuries there have been instances where architects refinements and variations which, from time to time have have taken pride in quite literally copying specific antique been lovingly and assiduously applied to them. Nevertheless, examples. For instance, Jean Bullant at Ecouen, the great even an '0 level' understanding of the orders is something, house near Paris, derived the Corinthian with most of its for it is by no means only in the handling ofthe orders them­ ornaments from the Temple ofCastor and Pollux; that was selves that the character ofclassical architecture lies. It is also in IS4O. at Covent Garden in 1630 reconstructed - even more (much more, in fact) in the way they are de­ the Tuscan on the basis ofVitruvius' text (plate 19) - almost ployed; but that is a subject for another talk. an archaeological exercise. Then Sir in 1793 Meanwhile, let us be quite clear about how variable or borrowed literally from the Temple ofVesta at Tivoli for the how invariable the orders are. Serlio puts them before us Bank ofEngland. On the other hand, there have always been with a tremendous air of authority giving dimensions for daring innovators. Philibert de L'Orme invented a new each part as ifto settle the profiles and proportions once and 'French' order for the Tuileries Palace; Wendel Dietterlin's for all. But in fact, Serlio's orders, while obviously reflecting orders in his book of 1594 are phantasmagoric variations on VitrUvius to some extent, are also based on his own obser­ Serlio; Borromini's orders are outrageous and extremely ex­ vation ofancient monuments and thus, by a process of per­ pressive inventions, entirely his own. So it is a mistake ever to sonal selection, to quite a considerable degree his own inven­ think ofthe 'five orders ofarchitecture' as a sort ofchild's box tion. It could hardly be otherwise. Vitruvius' descriptions of bricks which architects have used to save themselves the have gaps in them and these can only be filled from know­ trouble of inventing. It is much better to think of them as ledge of surviving Roman monuments themselves. The grammatical expressions inlpOSing a formidable discipline orders as exemplified in these monuments vary considerably but a discipline within which personal sensibility always has from one to the other so it is open to anybody to abstract a certain play - a discipline, moreover, which can sometimes what he considers the best features ofeach in order to set out be burst asunder by a flight ofpoetic genius. what he considers his ideal Corinthian, Ionic or whatever it Now, at this point I am going to ask you to look again at is. All through the history ofclassical architecture speculation the Doric order in the drawing on the endpapers. Because :. iill as to the ideal types of each of the orders has continued, I think it may still puzzle you that the entablature has so d£ oscillating between antiquarian reverence on the one hand many curious bits and pieces, all with names but with no

II particular decorative or symbolic value that you can see. back from the formalized Doric to its last timber prototype. Why mutules? Why and metopeSl! Why the taenia Their guesses are worth more than mine, but guesses they and those odd little tassels called guttae? You may well ask. are and are likely to remain. All that matters for us now is And I can only give you a very general answer. It is quite that in the process of time a system of timber construction, certain that the Doric order derives its forms from a prim­ copied in stone, crystallized into the linguistic formula which itive type oftimber construction. Vitruvius tells us as much. Vitruvius knew, and so we know, as the Doric order. This When you are looking at a Doric order executed in stone crystallization has a very obvious parallel in language. Words. . you are looking, in effect, at a carved representation of a expressions, grammatical constructions have all at some time Doric order constructed of wood. Not a literal representa­ had to be invented to meet particular needs ofcommunica­ tion, ofcourse, but a sculptural equivalent. The earliest tem­ tion. Those immediate needs are long since forgotten, but ples in the ancient world were ofwood. Gradually some of the words and their patterns still form the language we use these temples - those, doubtless ofspecial sanctity and which for a thousand purposes including poetry. That is how it is attracted wealth - came to be rebuilt in stone. It would be with the five orders ofarchitecture. felt imperative to preserve in the more permanent stone One more word about the orders. They are always sup­ version the actual forms round which so much sanctity had posed to have something resembling personalities; Vitruvius gathered. Hence, the carpentry devices of the wooden en­ was perhaps responsible for this. The Doric he saw as exem­ tablature, already, no doubt, somewhat stylized, were copied plifying 'the proportion, strength and grace ofa man's body' in stone or marble. Later on, no doubt, stone temples on - presumably an average well-built male. The Ionic, for him, new sites copied the copies, and so it went on till the whole was characterized by 'feminine slenderness' and the Corin­ thing became a static and accepted formula. thian as imitating 'the slight figure of a girl', which may Look at the Doric entablature again in the light ofthis and seem not very different from the last. But Vitruvius having it does, to some extent, explain itsel£ The mutules seem to be opened the door to personalization ofthe orders the Renais­ the ends of cantilevers jutting out to support the eaves and sance let in a lot more - often V€ry contradictory. Thus while to carry the eaves, from which the rain drips, well away Scamozzi echoes Vitruvius in calling the Corinthian 'vir­ from the columns. Then the triglyphs could be the ends of ginal', Sir Henry Wotton, a few years later, distorts him by cross beams resting on the . The taenia looks like calling it 'lascivious' and 'decked like a wanton courtezan', some kind of binding member and it appears to be secured adding that the morals ofCorinth were bad anyway. Never­ to the triglyphs by theguttae, which are not tassels, ofcourse, theless, the Corinthian has always been regarded as female but pegs. I say 'seem to be', 'could be', 'looks like' because and the Doric as male, with the Ionic in between as some­ all these things are my own rough guesses. Some archae­ thing rather unsexed - an ageing scholar or a calm and gentle ologists have devoted much ingenuity to trying to work matron. Serlio's recommendations are perhaps the most

12 - ~~~------­

:ype. specific and consistent. The Doric, he says, should be used character. And there is the fascinating case ofInigo Jones and they for churches dedicated to the more extraverted male saints - the Tuscan at Covent Garden which I shall come to in another ,w is St Paul, St Peter or St George and to militant types in gen­ talk. Tuscan and Doric are the two most primitive orders tion, eral; the Ionic for matronly saints - neither too tough nor and architects have tended to use them when they wanted hich too tender and also for men oflearning; the Corinthian for to express roughness and toughness or in the case of the This virgins, most especially the Virgin Mary. To the Composite Doric what is called a 'soldierly bearing'. At the other end )rds, Serlio awards no special characteristics, while the Tuscan he of the scale the Composite is sometimes quite obviously ~e finds suitable for fortifications and prisons. chosen because the architect wants to lay it on thick -luxury, rlca­ Now there is no need to take any of this too seriously. opulence, no expense spared. but Certainly, there is no need when you are looking at the Anyway, the main point is this. The orders provided a ~use Corinthian columns of, say, the Mansion House in London, sort ofgamut ofarchitectural character all the way from the it is to wonder if the Lord Mayor who commissioned them rough and tough to the slim and fine. In true classical de­ thought of them as virginal or the other thing. The fact is signing the selection ofthe order is a very vital point - it is sup­ that the orders have mostly been used according to taste, a choice ofmood. What you do with the order, what exact vius according to circumstances and very often according to ratios you give its different parts, what enrichments you put em­ means - building in plain Tuscan or Doric being obviously in or leave out, this again shifts and defines the mood. xly' less expensive than building in richly carved Corinthian. Well, so much for the Five Orders of Architecture - five rim, There are cases where the use of an order has a deliberate basic elements in the architectural grammar of Antiquity. Irin­ symbolic significance. I think, for instance, that Wren must But what can you do with the orders~ How does the gram­ may have used Doric at Chelsea Hospital because ofits soldierly mar worb That I shall try to explain in my next talk. nng lais­ ,bile vir­ 1by, :an, 2. The Grammar of Antiquity ~er­ pale ~e­ So far, I have devoted nearly all my time to the Five Orders I shall take your familiarity with them for granted and talk btle and I hope you are not tired ofthem because they are going less about the orders themselves than about how they are ~ost to be with us more often than not. From now on, however, used. Just look for the last time at Plate I. These orders, what I3