Florence yesterday. I’ve settled into a weekday pattern in Bologna that is proving productive, but the weekends this month are absolutely about being a tourist (or, excuse me, for my Chair’s benefit here–getting good slides for next Fall’s ARCH 423 course). I know from experience that going to Florence without a work plan and sharp elbows is a recipe for disaster, especially in the height of the spring tourist season, so I hung yesterday’s itinerary on one story from Vasari that floored me last year while reading up on Brunelleschi for class:
WITH HIS OWN HAND he made…the model of the house and loggia of the Innocenti, the vaulting of which was executed without framework, a method that is still followed by all in our own day. It is said that Filippo was summoned to Milan in order to make the model of a fortress for Duke Filippo Maria, and that he left this building of the Innocenti in charge of Francesco della Luna, who was very much his friend. This Francesco made an architrave-ornament running downward from above, which is wrong according to the rules of architecture. Wherefore Filippo, on returning, reproved him for having done such a thing, and he answered that he copied it from the Church of San Giovanni, which is ancient. ” There is one sole error,” said Filippo, “in that edifice, and thou hast followed it.”
So, for those of you not up on your classical elements, the architrave in Brunelleschi’s Loggia deli Innocenti (1419-1427) is the strong horizontal band that runs straight across the top of the arches. Arch- is (roughly latin) for ‘top’ or ‘main,’ and trave is a straightforward word meaning ‘beam.’ So the architrave started life as the main beam supported by columns, and migrated over time to any strong horizontal band that is visually supported by piers, columns, or pilasters. It’s a good way of separating stories, emphasizing a horizontal grain, or just tying together, say, a bunch of arches.
What beams don’t do, at least in classical language and structure, is suddenly make a 90° turn and dive into the ground, which is what Francesco della Luna did in Vasari’s (undoubtedly exaggerated) story:
This is the detail at the very end of the loggia, and it’s safe to say it doesn’t show up in many history textbooks, because it just looks odd (one exception–an essay by Tim Benton in a collection titled Making Renaissance Art from 2007). A structural rationalist would point out that the column is already doing the “work” of holding up a beam in this allusion, so why on earth would you also need a beam/column holding up the corner? Brunelleschi’s right to call it an ‘error,’ at least as it appears here.
But Francesco was also right, if Brunelleschi told him to copy the Baptistery, (1059-1128) because the one place a true architrave exists is on the upper story, where it crosses over four columns and then, at the edges of each wall, it dives down exactly like the detail on the Ospedale. (If the builders of the Baptistery had been good classicists, of course, they would have made the horizontal line above the arches, on the lower story, a true architrave, too…)
So, what gives? First, Vasari was notorious for repeating gossip, whether substantiated or not, and it seems unlikely that, no matter how busy Brunelleschi was, such a detail would have been left in the hands of a subordinate. He was notorious for giving verbal orders instead of drawing things out, but still, this was a howler. And, of course, if he had just said “copy the Baptistery,” any good second-in-command would have copied the lower story, which had arches similar to those of the Loggia, right?
The problem, I think, is the confusion that was evident in medieval building–and for that matter in the first couple of generations of the Renaissance–between the structural and the graphic languages of classicism. The architrave is a horizontal punctuation in its purest form–a graphic that’s designed to cast a deep shadow line on a facade and thus to drive home that there’s a distinction between what’s above and what’s below. But the builders of the Baptistery faced a curious problem that results from its octagonal shape. How do you make visually strong corners–necessary if you want to connote the monumental presence of one of the most important religious and political structures in the city–with a language of columnar, point elements? On the lower story they resorted to the green and white stripes that are familiar from medieval building all over town–effective visually. But the top is designed to be visually lighter, and it’s apparent that here the builders reverted to a graphic trope that borrowed the multiple lines of classical architraves and simply repeated it at the obtuse corners. There’s a bit of a vertical emphasis there that let them finish off the intercolumniation, and that lets one facade ‘breathe’ a little without being jammed into the corner of the next. It’s not overly elegant, but in white stone it’s also gentle.
In grey pietra serena, however, it’s a pretty violent clash, and whatever the actual story one can imagine Brunelleschi being desperate to blame the results on anyone handy. Whatever the actual story, the lesson was clearly learned. Here’s Alberti, who learned every trick he could from Brunelleschi, employing a very Baptistery-like architrave to help distract from the fact that the graphic ‘columns’ on the facade of Santa Maria Novella (1448-1470) don’t actually line up (note that he repeats the ‘mistake’ of pairing the column with, in this case, a pier…but the pier has those horizontal stripes that make it at least seem like it’s part of a different system):
And sometime later, here’s Alberti with a more archaeologically correct architrave on the Rucellai Palace (1446-1451, and maybe more by Rossellino than Alberti):
That’s doing it right, at least by classical rules. The horizontal line makes an absolute distinction between stories, and it carries all the way around the corner, making it seem that the columns at the corners are doing what columns do, namely carry the beam above them. It’s free and clear, too, of the arches, which also makes an important distinction that Brunelleschi never quite understood–that if the columns carry the beam, the arches really aren’t necessary except to carry the bit of wall between them and the beam itself. It makes sense to disengage these, and to put the arches on separate supports, to show that there are two different orders of structure going on here.
Splitting hairs? Sure. But this is the fun of teaching Renaissance architecture in between research on concrete and steel, etc. There was an awful lot of extra money and time being thrown around in Florence in the 15th century, which meant that for the first time in centuries architects got to throw ideas around and think about how ornament, or detailing, could help tell a story–or could make an otherwise legible facade trip over its own feet. That lesson is worth learning, and the subtleties are, if nothing else, fun to ponder. Over a doppio or two, in the midst of a 30,000 step, half-million tourist kind of a Saturday.