Development and Character of Gothic Architecture

Front matter

DEVELOPMENT & CHARACTER

OF

DEVELOPMENT & CHARACTER

Contents

Contents

CHAPTER I

Definition of Gothic

I. Definition of Gothic

CHAPTER I

DEFINITION OF GOTHIC

Since the decline of Gothic architecture the ideas which have prevailed respecting it have been for the most part confused and incorrect. Indeed, until within the last fifty years only the most vague notions of it were entertained even by students of architecture. The very name Gothic, though not wholly inappropriate, originated in a spirit of contempt, which naturally precluded any disposition to study attentively enough to understand it, this splendid manifestation of human genius. The architects and amateurs of the schools of Sansovino and Palladio in Italy, where the revival of taste for classic forms of art had set in as early as the time of Brunelleschi, could not be expected to admire anything so far removed from the spirit of the art which was in fashion with them. The manicra Tedesca, as they called such Gothic as they possessed—supposing Gothic art to be of German origin, because their own pointed style was an importation from Germany—was regarded by them as barbaric and without principles, in comparison with their grammatical Vitruvian orders. It was not unnatural that such distaste for the pointed style should be felt in Italy, for the style was foreign to Italian genius and Italian traditions. It had been adopted merely as a fashion, and the very modifications which the Italians wrought in it show how little suited it was to their wants. It is, indeed, impossible for a people possessed of an art, which is a natural outgrowth of their wants and tastes, and hence proper to them, to adopt and practise rationally, and to make their own, another art which is an outgrowth of other and different needs and predilec . The Greek and Roman types of building were not only the natural inheritance of the Italians, but they were the best for them as being suited to their climate and as supplying all their demands of convenience and taste.

On the north and west of the Alps the case was different. Here the traditions of classic art were not, in the same sense, an inheritance. The ancient forms of building had here been an importation. They had never here been wholly understood, and they were not well suited to the conditions of climate and of race. But the Gothic style which gradually took form in France was a natural outgrowth and expression of the genius of the people, and it was as well suited to them and- to the local conditions as the classic styles had been to the people and the climate of the South.

Yet here, too, at length, the fashion of distaste for Gothic set in—following the lead of the more natural Italian reaction,—though the change did violence to much that in architecture was proper to Northern temperament and Northern needs.

This fashion had its root in the prevalent, yet often insincere, feeling characteristic of an artificial state of society, such as that which Northern Europe, and especially France, exhibited at the end of the fifteenth century,—a state of society in which display of private wealth and pleasurable indulgence became the chief animating motives of an art that found its main expression in vast and luxurious private dwellings. In the former time private dwellings, even those of the rich, had been comparatively unpretentious and plain in character, while the Church edifice, the great centre of social and communal interest, and the product of the joint energy and enthusiasm of all classes, had been enriched by generous expenditure of toil and public and private treasure, but now it was the dwellings of the rich that chiefly demanded the services of art. The ambition of Charles VIII to possess a palace equal in splendour to those which he had seen in Italy, indicates the early stage of a movement which, gathering force under Francis I, and greatly stimulated by the genius of De L'Orme, reached its height in the sumptuous architecture of the reign of Louis XIV.

The taste for the new style was long confined to the upper classes. This architecture could never become really an architecture of the people; and the cities and, here and there, the church held out against it. But with the growth of artificial conditions the new fashion at length prevailed, and under its influence it was not strange that the Gothic monuments of the country were not only neglected and despised, but shamefully, and often irreparably, disfigured.

In England the taste for the pseudo-classic orders, fostered by the genius of such men as Inigo Jones and Wren, was not less hostile to Gothic. Any feeling for mediæval forms which had lingered on through the Elizabethan period was soon effectually quenched. Germany, though not quick to accept the Renaissance style, was also at length conquered by it. Everywhere some form, though often a travesty, of the revived classic taste prevailed. Gothic art became everywhere extinct.

Fashion, however, began after a while to change. In the course of the eighteenth century an antiquarian interest in pointed architecture was awakened and received a considerable stimulus from the zealous but ignorant advocacy of Horace Walpole. The attention of amateurs began to be directed towards existing monuments, and the publication (1780-1795) of Carter's volumes with measured drawings, followed before long by the works of Britton and Pugin, created an extensive, though not a discriminating taste for the long-abandoned pointed style. So undiscriminating, indeed, was this new interest that it long remained unproductive of good results. No just notion of the nature of Gothic was anywhere entertained. That it involved principles beyond those which were revealed to a superficial view nobody yet imagined. The whole subject of the modifications and transformations which pointed architecture had undergone at different periods in its history was shrouded in obscurity. No correct classifications had been made, and attention was, for the most part, directed to the later and least excellent varieties. Before there could be progress toward a true understanding of pointed buildings, it was necessary that the different forms which they had assumed should be examined and classified.

But at length this progress began. In the year 1817 appeared Rickman's first essay—An Attempt to Dis the Styles of Architecture in England. This book did much to clear up the confusion that had prevailed, by pointing out that the differences of style which appeared in the English monuments might be broadly divided into three groups belonging, respectively, to three successive periods of construction. And, although Rickman's work was naturally imperfect and inadequate, its classifications were mainly correct, and it has served as a substantial basis for all subsequent study of the pointed architecture of England. So good was it, however, that the many other treatises which soon after appeared did little more than extend the field by bringing a larger number of buildings into notice. Professors Whewell and Willis, however, deserve to be mentioned as learned and able investigators who must always command the respect of students of architecture. Whewell, in his Notes on German Churches, did much to systematise methods of observation, and Willis, in his Architecture of the Middle Ages, and in his Essay on Vaulting, has given us a more thorough analysis of constructive systems than any other English writer, and has also rendered acknowledged service to the most able writers of the Continent. But neither of these authors succeeded in bringing out with clearness the essential principles of Gothic.

In the year 1851 was published Sharpe's Seven Periods of Church Architecture, which showed that Rickman's division of styles might be subdivided. But beyond this Sharpe did nothing to invalidate the correctness, in broad outline, of Rickman's work. As regards the true nature of Gothic, Sharpe himself, though a writer of much merit, did not possess a true conception. For he says, referring to the commonly received distinction between Romanesque and Gothic, which is merely that one employs round, and the other pointed arches, that he has "little hesitation in adopting this primary division as the groundwork" of his system. And in his various other works, excellent as they are in many ways, he everywhere treats the subject of Gothic design as consisting merely in this and other peculiarities of detail. Of the considerable number of more recent English writers on Gothic art few, if any, have contributed towards a more just apprehension of its principles. English writers generally have understood by Gothic merely a style of building in which pointed arches take the place of round ones, and in which mouldings and other details are treated in a peculiar way. Hence, in treating of the evolution of Gothic, English writers, with hardly an exception, confine themselves to the consideration of these subordinate things. Even Sir Gilbert Scott, who has shown more insight than most others, quite fails to lay hold of the ruling principles and to exhibit them with clearness. And his son, Mr. G. G. Scott, even describes incidentally some of these principles without emphasising them as fundamental.

A recent Belgian writer, regarding the subject from the English standpoint, has published a book, in which it is assumed that Gothic consists in a purely decorative modification of the component members of a building. These members—capitals, bases, mouldings, etc.—he examines without due reference to their mutual adjustments and functional offices, and considers that the more they differ decoratively from corresponding members in the preceding styles, the more Gothic they are. For standards of Gothic form he points to those buildings in which such details depart most widely from the Romanesque types. And others which, in these features, retain the older characteristics he affirms to be on that account not Gothic.

In France the history of the revival of interest in Gothic seems to have derived its impulse from an influence transmitted from England. One of the earliest French writers to show an intelligent interest in the subject was M. de Caumont, who, however, equally with the authors already named, misconceives the nature of Gothic art. Thus, in treating of the transition, he says: "La révolution architectonique qui s'opérait durant la période transitionnelle ne consistait pas seulement dans la substitution de l'ogive au plein-cintre, mais aussi, comme nous le démontrerons, dans l'adoption d'un système nouveau de moulures pour la decoration, et dans l'abandon de la plupart des ornaments usités aux XIe et XIIe siècles."

fig1 These conceptions of the nature of Gothic are inadequate. It is not by the consideration of such structurally unessential features that an understanding of the subject can be reached. By such approach it would be impossible to discover the principles of any art. The principles to be considered are constructive principles. They determine the nature and govern the entire fabric of every art. In architecture they are pre-eminently fundamental. In architecture mere forms apart from their functional offices and relations are not enough to enable us to apprehend the distinctive characteristics of styles. Semicircular forms instead of straight beams may, for instance, be used to bridge the spaces between the upright supports of a building without a result which would constitute between a building employing the latter form and one employing the former, a difference of architectural style. For an arch may be cut out of a single stone, as at A, Fig. 1, as it frequently is in the buildings of Central Syria, where the constructive principle is, of course, that of the plain lintel, as at B, Fig. 1. fig2 Or the arch may be built up in horizontal courses of small stones, and thus be what is known as the offset arch (Fig. 2), like the gate of Ephesus, which is still on the principle of the lintel. It is not until the arch is constructed of separate stones cut into the shapes of voussoirs, when it comes to exert a lateral thrust which requires to be met by some opposing force, that we have a new con principle, the systematic employment of which in architecture constitutes a new style.

In a secondary sense, indeed, it may be admissible to speak of differences of style where there are no differences of constructive principle. Egyptian architecture is, in this sense, a style different from Greek, and arched Roman is a style different from Romanesque. The Romanesque may be broadly divided into two styles—the Eastern and the Western; and the variety of Western Europe may be said to be of one style in North Italy, of another in Southern Gaul, of another in Normandy and England, etc. While of pointed architecture it may be said that there are differences of style, or rather that there are many varieties, some of which are nearer to, and some more remote from, the type which alone is strictly entitled to be called Gothic. But it is only in a secondary sense that it is correct to speak thus of styles in which there are no fundamental structural differences. Pointed architecture is not, in the strict sense, a style distinct from that which is round arched; for pointed arches in apertures do not much differ structurally from round ones. Gothic architecture differs from Romanesque far more fundamentally than by the use of pointed arches in place of round arches, or by the substitution of one decorative system for another.

In the midst of such imperfect apprehension as has thus far generally prevailed, and as preliminary to what is to follow on the nature and origin of Gothic art, it will be well for us to seek a clear and unmistakable definition of it, in order that we may have a standard whereby to estimate the degrees of Gothic quality that may appear in the pointed architecture of different countries and at different epochs. Such a definition is afforded in the monumental work of M. Viollet-le-Duc, the Dictionnaire Raisonné de l'Architecture Française. He has therein given a profound and exhaustive illustration of Gothic. He has shown that Gothic architecture came into being as a result of the development of a new constructive system of building. A system which was a gradual evolution out of the Romanesque; and one whose distinctive characteristic is that the whole character of the building is determined by, and its whole strength is made to reside in a finely organised, and frankly confessed, framework, rather than in walls. This framework, made up of piers, arches, and buttresses, is freed from every unnecessary encumbrance of wall, and is rendered as light in all its parts as is compatible with strength in a system whose stability depends not upon any inert massiveness, except in the outermost abutments, but upon a logical adjustment of active parts whose opposing forces produce a perfect equilibrium. It is thus a system of balanced thrusts, as opposed to the former system of inert stability. Gothic architecture is indeed much more than such a constructive system, but it is this primarily and always. And so fundamental and far-reaching is this mode of construction as the distinctive principle of Gothic, that it may be taken as a rule that wherever we find it developed there we have a Gothic building, even though the decorative system connected with it may retain many of the Romanesque characteristics. And, on the other hand, wherever this principle of thrust and counterthrust is wanting there we have not Gothic, however freely the pointed arch may be used, and however widely the ornamental details may differ from Romanesque types.

The evolution of the Gothic system was gradual, and the final results were unforeseen when the first steps were taken. This is evident from the characteristics exhibited by the transitional monuments which remain. The first steps were taken early. Indeed, the beginnings of Gothic antedate considerably the period which is usually assumed as that of transition. The earliest functional grouping of supports in the churches of Northern Italy was, it would seem, the real beginning; though it was a beginning that was destined to remain unfruitful in its own locality. But there is reason to suppose that the Lombard buildings of the early eleventh century were the sources from which the Normans derived a large share of their architectural inspiration. A comparison of the Church of San Michele of Pavia with that of the Abbaye-aux-Hommes of Caen would alone suggest this, even if we did not know that Lanfranc, who was Abbot of the Monastery when this church was building, had come from Pavia. The rudimentary principles of structure embodied in the churches of Lombardy seem to have been thus transmitted to the North, where they were rapidly developed, so that from San Michele of Pavia to the Cathedral of Amiens a logical and progressive series of changes may be traced.

Gothic architecture is thus in no sense an independent, though it is a distinct style. And hence it is not strange that the finest Gothic buildings should retain, as they do, many traces of the Romanesque elements out of which they were evolved. It is a mistake to suppose with M. Colfs that such lingering of older elements necessarily marks a building as wanting in Gothic character. On the contrary, such elements are natural to Gothic, which is an art not only derived from Romanesque, but which is Romanesque recreated. Every constructive member of a Gothic building exists, in rudimentary form, in a vaulted Norman building. Even the system which distinguishes Gothic exists there in potency. But the creative principle of the new style is not in Norman art yet recognised, and the structural system does not yet frankly and independently exercise its functions.

It may help us to gain a more complete idea of Gothic briefly to review some of the steps in the process by which the transition was effected. And we may advisably begin with the first principles of arched construction.

I have already alluded to the fact that a new principle was introduced into the art of building when the arch exerting side thrusts was first employed. The most effective way to meet such thrusts is by some kind of external abutments. But the thrusts of arches may also be neutralised by downward pressure upon the walls where they operate. Both methods were employed by the Romans and by the Romanesque builders, as well as by the Gothic architects. In the case of a simple arched opening in a wall the thrusts are, of course, stayed in both of these ways. The lateral masses of wall act as buttresses, and the superincumbent masonry tends to overcome the side pressure by its weight. Where a space fig3 between two parallel walls is roofed over by a barrel vault the continuous side pressures, which would tend to overthrow the walls, are, in Roman constructions, met by thickening the walls enough to resist the thrusts—the extra thickness given for this purpose forming a continuous abutment. The walls in Roman buildings are also sometimes weighted by heavy masonry above the springing of the vaults. In Roman buildings of several stories, such as the Flavian amphitheatre (section, Fig. 3), the walls of the lower stories are enormously thick, and the vault pressures are further stayed by the weight of the walls above. The top story has no vault, and the weight of its thinner wall helps to maintain the stability of the vault below. By such massive masonry employed in this double way, the pressures of Roman vaults are much more than met. In the case of Roman intersecting vaults, like those of the Basilica of Constantine, the thrusts, instead of being continuous, as in the foregoing instance, are concentrated upon those points (a, plan, Fig. 4) from which the groins spring, where they are met by walls set across the aisles and dividing them into separate compartments. These walls are, of course, true buttresses in disguise. The compartments of the aisles are covered by barrel vaults springing from the dividing walls, and thus, having their axes at right angles to the side walls of the building, they exert no thrust upon these side walls, and consequently no external stays are required. Thus the buttress employed by the Romans was not a buttress pure and simple, devised to meet a side pressure with economy, as well as efficiency, and openly confessed as a functional member. They always contrived to arrange the plans of their buildings so that some of the enclosing or dividing walls should act as stays to their vaults, or else they resorted to the methods before noticed of employing such vast thickness of wall as to secure stability by sheer inertia of material.

fig4

FIG. 4.

The Romanesque builders went a step farther in the development of the buttress, in accordance with their general progress in the art of construction. They at first placed a pilaster strip on the outside of the wall against the pressure that was to be met, treating it as a distinct functional member (Fig. 5). It is true that the fig5 Romans had employed engaged columns in the same positions, but they had employed them for a decorative purpose only. And even in early Romanesque constructions the pilaster strip had little more than a decorative value. It did, indeed, stiffen the wall somewhat, this was the reason for its use, the walls of Romanesque buildings not having the great thickness that was common in Roman walls, but it had not projection enough to bear much vault pressure. It had, however, rarely to meet such pressure, except in the aisles where the vaults were of no great span. But though it was of slight efficiency in vaulted constructions, it yet had great value as marking the place where, in such constructions, additional strength was required by the walls. And in the later Romanesque, as vaulting became more general, the pilaster strip was developed into a true buttress (Fig. 6).

fig6A beginning was made in the direction of further progress when the Romanesque builders began to vault their naves. It was then found that the pilaster buttress against the clerestory wall was not enough to stay vaults of so much wider span than those of the aisles for which a buttress like that shown in Fig. 6 had been adequate. Expedients to augment the resistance of the clerestory buttress were accordingly resorted to which were destined ultimately to yield unforeseen results.

In the Abbaye-aux-Hommes at Caen the forms of the vaults—which date from the early part of the twelfth century, and are among the earliest that were constructed over a nave—were such as to exert powerful thrusts. That is to say, the arches of their groins were curves of low sweep, such as the Romanesque builders had derived from Roman intersecting vaults, and consequently of enormous push. To stay these vaults the expedient was adopted of constructing demi-barrel vaults, springing from the top of the aisle walls, and butting against the wall of the nave under the aisle roofs (Fig. 7). These demi-vaults were in reality concealed continuous flying buttresses. But they were flying buttresses of bad form, for only a small part of their strength met the thrusts of the vaults, the rest being exerted against the walls, between the piers, where no props were required, and where their effect would have been disastrous had not these walls been of excessive massiveness. The level of the abutment was, moreover, so low that it failed to meet the points where the thrusts were greatest. The precise chronological sequence of buildings in which the successive improvements on this mode of buttressing were made cannot be traced. But an illustration of the next step is afforded by the Abbaye-aux-Dames (Fig. 8), which seems to have been vaulted somewhat later than the Abbaye-aux-Hommes. In this case, instead of

fig7

FIG. 7.

a continuous arch, or demi-vault, springing from the aisle wall, separate arches were established, springing from the aisle walls opposite the piers, and abutting against the piers only, where the thrusts of the vaults were gathered. But these arches still fell too low to be wholly effectual, and as the precaution was not taken to reinforce the buttresses of the aisle walls the supports have yielded, and the original vaults have been destroyed. But the same arrangement may be seen in the nave of the Cathedral of Durham where the original vaults remain. Their duration, however, is owing to the enormous massiveness of the construction rather than to the form of the support, which of itself is inadequate. Hence, though an important improvement was attempted in these instances, a satisfactory solution of the problem of the

fig8

FIG. 8.

buttress was not yet reached. The abutting arches of the Abbaye-aux-Dames and of Durham are indeed true flying buttresses, but they have not the character of such members in Gothic architecture, inasmuch as they are ill adjusted and are not externally apparent.

An advance of no less importance than this buttress development and an integral part of the growing system was the employment of independent arches, or ribs along the lines of the groins, projecting below the vault surfaces and in a measure sustaining them. Here, as in the case of the buttress, the Romans had employed a kindred principle, though in a most rudimentary way. They had constructed a framework of masonry to give strength to their vaults, but with them this framework was buried in the thickness of the masonry instead of projecting from, or even appearing upon, its surface. It consequently failed to possess the important use of the rib system we are now considering, which, being quite independent of the vaults, serves as a strong centring, and prevents any rupture that may by any chance take place in one cell or compartment of the vault, from communicating itself to others. In addition to this, the employment of an independent support for each rib and arch to be carried—which constitutes the functional grouping of supports above spoken of—completed the structural improvements devised by the Romanesque builders.

We are yet far from the Gothic system. The inert principle of construction, the massive walls, the small apertures, and the horizontal lines of the Romanesque architecture make it still closely akin to the old Roman style. But there are rudiments in it already quickening with latent life, which will completely transform the Romanesque building. The evolution of the Gothic system consisted in gradually perfecting the rudimentary skeleton, so as to make it an independent structure. To every part a complete and independent working efficiency was to be given, and an appropriate artistic, as well as a mechanical value. All this was rendered possible to a far greater degree than it could otherwise have been by the introduction of the pointed arch, not as an ornamental feature in doors and windows, but as a constructive device in vaulting.

The properties of the pointed arch which enabled the Gothic builders to overcome difficulties in vaulting that had before been insuperable are that it exerts a less powerful thrust than the round arch, and that with a given span its crown may be made to reach any level. Its employment in the transverse arches of the vault raised their crowns so that, with arches of full semicircular sweep for the diagonals, the thrusts were greatly diminished. The vaulting of oblong compartments had before been attended with difficulties resulting from the fact that the height of the crown of a semicircular arch is determined by its span. In vaults over oblong compartments the crowns of the round arches which spanned the narrow sides would not reach the level of those fig9 which spanned the longer sides. While if full semi-circular arches were used for the diagonals of such vaults their crowns would reach highest of all. Thus in Fig. 9 the height c d of the arch a c b is less than f e, the height of the arch a f g, which again is less than i h, the height of the arch a i j. A vault constructed upon such a system of arches must have an excessively domed form. To obviate this, in part, the expedient was adopted of stilting the narrow arches. The point of their springing was raised to a level considerably above the springing of the greater arches, so as to bring all the crowns nearer to the same height, and thus to reduce the amount of doming required. But even with this modification a vault over an oblong compartment, upon a system of round arched ribs, is excessively heavy, exerts powerful thrusts, and presents an awkward appearance. Oblong groined vaults, though sometimes constructed, were therefore usually avoided by the Romanesque builders, who, indeed, had rarely vaulted their naves, the portion of the building where, in North-Western Europe, oblong compartments most frequently occur. They contented themselves with vaulting the aisles whose compartments were commonly square, and where groined vaults, on round arches, were easily constructed and easily rendered secure.

The introduction of the pointed arch, however, obviated these difficulties. It now became possible to construct groined vaults over oblong compartments without either doming or stilting, since the crowns of all the arches could be readily brought to the same level whatever their differences of span (Fig. 10).

fig10 But it is important to observe that, in true Gothic, oblong vaults are never constructed upon ribs which all spring from the same level, and whose crowns all reach the same height. Other exigencies, which will be explained in the next chapter, stood in the way of so constructing them. True Gothic vaults are always, to some extent, both stilted and domed. But though full advantage of the pointed arch, in affording any height with any span, could not be taken, its introduction was a great help, and it gave a powerful stimulus to constructive invention.

With the reduced thrust of the vault, secured by the pointed arch, the volume of the external stays could now be much reduced, and new experiments soon led to their better adjustment. The flying buttresses were brought to bear more directly upon the points of greatest pressure, which were found to be at a higher level than those upon which the abutting arches of the Abbaye-aux-Dames and of Durham Cathedral had been brought to bear. In order to reach these points it was necessary to make the flying buttresses spring over the aisle roofs, and thus to become marked external features—as in St. Remy at Reims, St. Leu d'Esserent, St. Germain des Prés at Paris, and a few other early Gothic monuments which still retain their original flying buttresses. The vault ribs and the vaults themselves were now also made lighter. And finally the ribs were more closely grouped at their springing—being made to interpenetrate, more or less, by which means the pressures were concentrated to the utmost, while the sustaining shafts and piers, for which the best materials were selected, were reduced to a minimum of thickness. As this development of an independent framework progressed, the intervening walls, now no longer needed for the stability of the fabric, were also reduced in thickness; and the small apertures of the Romanesque style gave place to larger openings, which were gradually more and more enlarged, until they filled the entire space between the supports.