Lectures on the rise and development of medieval architecture; vol. 2

by Sir George Gilbert Scott

Norman architecture, then, judging of it from its principles, and throwing aside imperfections resulting from its development occurring in comparatively rude times, may, in the first place, be said to be an almost perfect carrying out into a style of art the arcuated system of construction; using, also, the simplest and most obvious form of arch, the semicircle. Roman buildings, when divested,—as in the case of aqueducts,—of Grecian or trabeated accompaniments, displayed often a perfect system of arched construction; but, in such works, one cannot say that it had been developed into a style of art.

To effect this, both the arch and its supports and accompaniments must be moulded into artistic elements; their natural crudenesses softened; their mere normal character relieved; and each part subjected to a system of decoration suited to its proper character and conditions. The parts, too, which have been thus dealt with, must be studied as to their grouping. They must not be viewed as isolated objects, but as parts of an architectural work; each contributing to the beauty and consistency of the whole; and that also, by such combinations as are dictated by the varied suggestions arising from the purposes and demands of the buildings of which they form parts.

All this required time; and the length of time was, no doubt, increased by the rudeness of the ages during which the process had been going on.

Among the earliest approaches to so reasonable a result, the Lombard style had taken, perhaps, the lead in Western Europe; and, during the days when the three Othos governed Northern Italy as well as Germany, the good seed had spread from Lombardy into Germany, and it there grew into an almost perfect development.

Somewhat similarly, a well-considered development seems to have originated in Central France, and spread towards the north. Probably these two varieties may have come in contact, and in some degree influenced each other; for the early Norman architects, though mainly developing upon French models, appear to have been acquainted with those of the Rhine. However this may be, it is certain that they developed for themselves a variety of Romanesque at once eminently reasonable, and susceptible of highly artistic treatment and combinations.

The elements of such a style are often not, as taken singly, peculiar to itself, but may be found in other and in earlier works: it is the aggregation of many such elements, and their judicious and artistic utilisation that constitutes the merits of a style.

Among the most important of these may be placed the sub-ordination of arches, by means of which, instead of going square through the thickness of a wall, they recede in orders or arched rims, each narrower than that above it, so as to give the entire arch or section of alternate salient and receding angles. This is the primary element; and it at once produces the second,—the breaking of the section of the bearing pier into a similar form to that of the arch. This, in the Anglo-Saxon style, was hardly known; while in the Norman it is the key-note.

At St Alban’s, where the unmanagable nature of the material,—the Roman brick,—rendering finished architecture unattainable, we find these two principles supplying all architectural requirements, and producing results certainly rude, but not unpleasing in their effect. This building is often said to partake of “Saxon” character. I think the very reverse of this; for the one thing to which it trusts for effect is that which scarcely exists in Anglo-Saxon buildings, while it is the leading principle in Norman ones. This error is the natural result of looking to rudeness of workmanship and homeliness of material, instead of the principles of design, as the evidences of early style.

The next principle is merely the resultant of those already named. It is the decoration by mouldings of the salient divisions of the arch and the substitution of decorative shafts for those of the pier.

These principles do not necessarily accompany one another. An arch-order may be moulded or otherwise decorated, while the corresponding pier-order may remain square, the two being parted by an impost; or the decorations of the arch may, without the intervening impost, be continued through the pier; or, again, a shaft may be substituted for the pier-order, while the arch-order remains plain. The above principles, thus variable in their application, supply the most marked features in the perfected Romanesque style, nor can any arched architecture be perfect without them. To illustrate their effects let us take a doorway of the older English period, and contrast it with a Norman doorway.