POGGIO BRACCIOLINI, AN INSCRIPTION IN TERRANUOVA, AND THE MONUMENT TO CARLO MARSUPPINI: A THEORY

The strangest Renaissance inscription is the dedication plaque of Poggio Bracciolini in the church of S. Maria in Terranuova Bracciolini. Over the course of eighteen lines, its letters morph from Florentine sans serif capitals to Imperial Roman capitals. The author theorizes that the gradual change was the result of Poggio Bracciolini coaching an untutored lettercutter in the subtle differences between the two styles of letters. Furthermore, there is a visual link between the letters of the Terranuova inscription and those of the inscription on the monument to Carlo Marsuppini in S. Croce that suggests Poggio played a role in its design.

The strangest Renaissance inscription is the dedication plaque of Poggio Bracciolini in the church of S. Maria in Terranuova Bracciolini, a small Tuscan town located between Florence and Arezzo. The lengthy (18 line) text, describes the story of a «certain Roman citizen, needy and in poverty» who, upon complaining to God about his situation, had a series of dreams in which an apparition told him to demolish the altar in a church built by Sixtus II in order to find treasure (Fig. 1). When he did this he discovered two boxes of marble, one of which contained «a small glass vase» which in turn housed extremely precious relics, among them a bone of St. Lawrence. The man brought the relics to Rinuccio da Castiglione (Rinuccio Aretino), a papal secretary and humanist, and asked for help. Rinuccio gave the relics to Poggio who, after helping the man with his poverty, placed them in a new chapel he had built in Terranuova, his birthplace, in 1438 1 . The events of the story occurred in 1433, in the fourth year of the pontificate of Eugene IV, according to the text. The inscription gives the date for the deposit of the relics in the chapel (built in 1429) as «anno aetatis meae LVIII» or in the 58th year of his life. Given that Poggio was born in 1380, the deposit is dated to 1438 and the inscription is presumed to have been made the same year 2 .
The Terranuova inscription is 59 cm (23.25 inches) high by 138 cm (54.33 inches) wide. The text is eighteen lines long, flush left, ragged right, with the last line centered. The letters vary in height from 23 mm to 25 mm depending upon the line 3 . Word spacing is so tight that it often seems non-existent. The massing is enabled by the liberal use of a variety of Medieval space-saving strategies: ligatures (including a few three-letter combinations), nested letters, overlapping letters, tall letters, abbreviations, and a Z-shaped Tironian et. (Fig. 2). Beginning with line 10, puncti, a feature of classical inscriptions, are present between most, though not all, words 4 . Light guidelines for the tops and bottoms of letters are visible (e.g. line 11). All of this suggests an inscription struggling to escape the medieval world and enter the ancient one. That sense of transition permeates the letterforms.
What makes the Poggio inscription so odd is that its letters metamorphose over the course of the full eighteen lines from a contemporary Florentine sans serif to a very close recapitulation of the capitalis monumentalis of Ancient Rome. I know of no other inscription from the Quattrocento (or even any other era or place) that contains letters that shift as these do. While that alone would make the inscription worthy of study, it is the nature of the transformation that is truly as intriguing as the author of the text. B.L. Ullman has called Poggio Bracciolini (1380-1459) the «inventor» of the scrittura umanistica or humanist bookhand c.1402/1403. He has also suggested that Poggio's manuscript majuscules played an important role in the Renaissance revival of Roman capitals. Although his evidence for this claim is not convincing, his idea of Poggio's broader importance is strongly supported by the Terranuova inscription 5 . Manuscripts copied out by Poggio have pen-Section 7. The inscription has been transcribed by Fubini. See Fubini, 1966: 861 (note) and 863 (transcription). Debra Pincus has called my attention to a discrepancy in the dating of the events described in the text. The text says they occurred in May 1433, yet the fourth year of the pontificate of Eugene IV was 1434. I do not think the discrepancy has any bearing, however, on the making of the plaque. See Rorimer, 1955Rorimer, -1956. The lettering on the reliquary has been described as Florentine sans serif of the Donatello school style by Christine Sperling, but is closer to medieval lettering in my view and has little relevance to the capitals of the Terranuova inscription. 3 The overall dimensions come from Sperling, 1985: 166 Script (1960). Ullman (pp. 54,56,and 63) argues that Poggio's ma-made letters with minimal or no serifs; and those copied by scribes at his behest often contain initial letters similar to the inscriptional letters found in the sculptural work of Ghiberti, Donatello, Michelozzo, Luca della Robbia and others from 1412 through the 1440s. Ghiberti called his capitals «lettere antiche» but they, and Poggio's initials, were actually indebted to Carolingian and Romanesque capitals rather than anything from Ancient Rome. These letters are marked by thick-tothin stroke contrast, terminals that are either wedge-shaped or flared, and an absence of serifs. The latter characteristic led Nicolete Gray to dub them Florentine Sans Serif (1960). Such letters play a role in the Terranuova inscription.
In comparing Ancient Roman inscriptions carved in Imperial Roman capitals and Renaissance inscriptions carved in Florentine sans serif capitals, there are six letters whose form is especially significant: E, G, M, N, Q, and R. In this discussion, I am using the influential inscription on Trajan's Column and others in the Trajanic mold as a model for the capitalis monumentalis and Luca della Robbia's Cantoria as an exemplar of the Florentine sans serif.
The first line of the Terranuova inscription has Florentine M and R; Imperial E, G, and Q; and an ambivalent N. The M has straight sides, a short vertex and flat apices. Its thick/thin distribution of weight alternates from stroke to stroke. The R has a curved leg (Fig. 3). Both letterforms can be found in Luca della Robbia's Cantoria (1431-1437). The E has horizontal strokes that terminate in brackets or serifs; the G has a seriffed jaw stroke; and the Q has a long, curved tail. The N has three evenly weighted strokes, but the apex is flat. The ambivalence displayed here becomes more pronounced as the inscription progresses with the N and R fully evolving into Imperial forms; the M changing only partially; and the E (and its relatives the F, L, and T) shifting back and forth from a Florentine form to an Imperial one before settling on the latter. The G and Q remain Imperial throughout. By line eighteen the only letter in the alphabet that has not fully shed its Florentine shape is the M that becomes splayed but retains its short vertex (Figs. 4 and 5 top).
The Terranuova inscription (especially lines 12-18) can be described as the first instance of revived Roman Imperial capitals in the Renaissance, predating such well-known claimants for that title as the tomb of Martin V (now dated to the early 1440s); Donatello's signatures (OPVS DONATELLI / FLO) on the statue of Gattamelata (c.1453) in Padua and the statue of Judith and Holofernes (c.1457-1464); the painting of S. juscules were closely based on Roman inscriptions but Nicolete Gray disagrees, correctly pointing out that they are pen-made and not related to carved letters (Gray, 1986: 122 (Fig. 6 bottom).
Although Poggio's Terranuova inscription is unusual, it is not anomalous. Surprisingly, there is an extremely close visual connection between it and the inscription on the monument to Carlo Marsuppini in S. Croce (Fig. 7). All of the Marsuppini letters, with the exception of the K, closely match the «final» ones of the Terranuova inscription in their proportion and key features. The A begins with a flat apex (lines 1-3) but eventually becomes pointed (lines 4-6); the E is narrow with strokes ending in clearly bracketed serifs; the G has an overhang; the M is splayed with a short vertex; the N is Trajanic; the Q has a long, curved tail; the R has a slightly curved diagonal leg; the S struggles to stay upright; and dots are used as puncti (though only at the end of lines). There is also a Tironian et, though it is the standard 7 form. What principally separates the Marsuppini inscription from Poggio's inscription is the consistency of its serifs, the higher quality of its carving and its airier layout (abetted by a much shorter text of only six lines).
How can the similarities of these two inscriptions, carved two decades apart, be explained? I believe the link is Poggio. He was a friend and humanist colleague of Marsuppini and his successor as chancellor of the Republic of Florence (Martines, 1963). Although he is not mentioned in the literature as being involved in the plans for Marsuppini's monument, I believe that upon returning to Florence from Rome to take up the post of chancellor he took an active role in its epitaph. The authorship of the epitaph is a mystery and the debate between the Martelli and Medici families over it has led to conflicting opinions among contemporary art historians as to the completion date of the monument 7 .  , 1789), II, 219». Anne Markham Schulz also cites this letter in arguing for a dating of the monument to the summer or later of 1459 (Schulz, 1992: 180-81). However, I am not suggesting that Poggio wrote the epitaph -though such an act would have mimicked Marsuppini writing the epitaph of Leonardo Bruni, his predecessor as chancellor -but that Poggio was involved in its visual appearance.
There is no indication as to who carved the Terranuova inscription, though the constant mutation of letterforms points to a single individual 8 . I theorize that Poggio, who had carefully studied Roman inscriptions during his tenure as a secretary for the Church and had begun to gather his research into a sylloge towards 1430, wanted the inscription to be cut in true Imperial Roman capitals rather than in the Florentine sans serif that had become popular. The difficulty facing Poggio was how to convey the form of these ancient Roman letters to a Tuscan sculptor or stonecutter who had probably never been to Rome.
In the 1420s Poggio successfully taught scribes to write the scrittura umanistica, in some cases so well that scholars have mistaken their manuscripts for those copied out by Poggio himself (see Ullman, 1960: 49-51). But, judging from his sylloge, he lacked the drawing skills necessary to accurately render the subtleties of Imperial Roman capitals, something that even stumped Jacopo Bellini 9 . Furthermore, his eyesight was deteriorating after 1425 and that would have made it more difficult to draw precise model letters. Instead, Poggio probably made rudimentary sketches and showed the carver some of the Roman inscriptions he had collected for the garden of his country house in Terranuova. But the carver must not have fully grasped the importance of the details (e.g. serifs) of the letters or lacked ancient Roman models for some of them. For the latter he turned to the Florentine sans serif, especially as found on the recently installed Cantoria in the Duomo. This would explain the M and R.
Poggio must have continually looked in on the carver to gauge his progress and, unhappy with the Florentine sans serif letters, begun to coach the correspondence of Francesco Aretino only indicates that he proposed two epitaphs for Marsuppini (neither of them quoted) to Piero di Cosimo de' Medici, not that the epitaph as carved was his. Tommaso Mozzati challenges Schulz's interpretation of the document: «[…] the text inscribed on the tablet, despite the elegance of the characters, is not centered or uniform, thus suggesting that the letter-cutter was obliged to adapt the epitaph to a space that had been planned for those years» (2007: 118). I disagree with his conclusion. Centering of a text is a decision that would not be affected by the limitations of a predetermined space. Any difficulties in adapting the epitaph to the space would have appeared either horizontally as crowded letters, as in the Bruni monument, or vertically as crowded lines. 8 Fubini concluded that the change in lettering indicated that two men carved the inscription (1966: 861). Sperling argued that it was the work of one man, though she mistakenly did so on the grounds that the letters -other than M and R -are consistent throughout, which they are not (1985: 168). 9 See the sketchbook of Jacopo Bellini in the Louvre (Accession number 401484), dated 1430-1460; Golubew, 1908: plates 43 and 44. him on the proportions, forms, and features that distinguished classical capitals. Under Poggio's tutelage the letters of the epitaph slowly moved toward being true Imperial Roman capitals in nearly all aspects, including the presence of bracketed serifs. This progress was painful at times as the carver was instructed to fix some letters, and to do so he had to recarve strokes, thus making them heavier and, in some instances, clumsier (e.g., RELLIQVIAS in line 17). Despite the difficulty of the work, the unknown carver, with Poggio's guidance, managed to achieve the first credible letters in the manner of the capitalis monumentalis in the Renaissance.
The close similarity between the letters of the Terranuova and Marsuppini inscriptions cannot be a coincidence. The latter has no other precedent and no successor. Although the monument to Carlo Marsuppini has frequently been compared to the monument to Leonardo Bruni (1449-1452), their inscriptions have not been. The only discussion of either has been by Millard Meiss who praised the Bruni inscription for having letters of «impressive symmetry and balance», a view which I would challenge 10 . The Bruni letters are unique among Renaissance inscriptions for their extreme lightness, a feature that gives them an elegance that disguises their Florentine sans serif roots. But individually and en masse they are inferior to their Marsuppini counterparts (Figs. 8 and 5 bottom).
Although there is no documentary evidence of his involvement in the planning of the Marsuppini monument, the letters of the inscription strongly suggest Poggio's influence. As a close friend of Marsuppini as well as his successor as chancellor of Florence, Poggio would have had an interest in seeing the epitaph set out in Ancient Roman capitals as both a fitting testament to Marsuppini's classical erudition and as a worthyif not superior -companion to the Bruni monument 11 . He would have 10 Pages 98-99 of Meiss's essay Towards a More Comprehensive Renaissance Paleography (1960) contain his assessment of the Bruni inscription which unfortunately ignores letter-and word-spacing and individual letters such as the P with its large, open bowl and the peculiar G. The P and the triangular puncti indicate some classical influence on the carver of the inscription. An interesting aspect of the Bruni inscription that escaped Meiss's notice is the presence not only of horizontal guidelines for the base and top of letters, but also of vertical guidelines for their width. 11 It is surprising that the author of the Marsuppini epitaph is not known. Marsuppini wrote the epitaph for Bruni, his friend and predecessor as Chancellor. It would have made sense for Poggio to have done the same for Marsuppini. This is the text and English translation from Pope-Hennessy, 1996: Who was the carver of the Marsuppini epitaph? Its high quality of execution -the forms are well balanced and consistent, the V-cut is clean and crisp -rules out the anonymous carver of the Terranuova inscription, unless he had matured during the intervening decades. Although Desiderio da Settignano (c. 1430-1464 is acknowledged as the sculptor of the monument, it is unlikely that he had a hand in the inscription. The lettering on other works attributed to him, such as the bust of Olympias, Queen of the Macedonians (c. 1460Macedonians (c. -1464, is squarely in the Florentine sans serif tradition. Since Settignano had apprenticed with the bottega of Bernardo Rossellino, he may have hired someone from there to do the lettercutting. If so, that carver would have required oversight from Poggio to avoid lapsing into familiar Florentine sans serif forms since those, with the notable exception of the Bruni inscription, were the stock style of the Rossellino workshop well into the mid-1460s 13 . The completion of the monument to Carlo Marsuppini, a few months before his death, meant that Poggio lived to see the classical Roman capitals he had studied and copied over a half-century earlier finally reappear in public in Florence. It would be more than another decade before they would be surpassed with the completion of Verrocchio's tomb for Piero and Giovanni de' Medici (1472).
Stay and see the marbles which enshrine a great sage, / one for whose mind there was not world enough. / Carlo, the great glory of his age, / knew all that nature, the heavens and human conduct have to tell. / O Roman and Greek muses, now unloose your hair / Alas, the fame and splendour of your choir is dead. (305) The text is deliberately not centered, but flush left/rag right with two lines indented to accentuate Marsuppini's name and his fallen status. In the fifth line, the «7» before «GRAIAE» stands for an ampersand, read as «ET». 12 The only letters in the Marsuppini inscription that deviate significantly from classical models are the K, which is Greek in origin, and M. Other than the K and    One of the puzzling aspects of the Terranuova inscription when I first encountered it in 2007 was the seemingly shallow carving of the letters. It made rubbing the letters difficult. And for some reason many of my photographs of the letters were in soft focus. On a return visit in 2014 with calligrapher Monica Dengo of Arezzo we discovered that the surface of the inscription was coated with wax from the nearby votive candles. The wax was responsible both for the softness of the outlines of the letters and for the appearance of shallow carving. A third visit in 2015, with Ms. Dengo, Prof. Carlo Fabbri, and several town officials, was intended as an effort to melt the wax and thus reveal the true nature of the letterforms. But a test attempt of a single word in the lower right corner quickly put an end to that idea. Once the wax melted the letters were no longer visible to the naked eye. Before there had been different tones of gray, but now the true dark gray color of the stone, including the incisions, had been revealed. In the dimness of the church, there was insufficient light to make the word visible. However, it was detectable by touch since the original sharp V-cut had been restored. This suggests that the Terranuova inscription may have been originally painted or gilt to make it visible.
Because of worries over the effect of heat on the stone as well as uncertainty over the wisdom of making a visible inscription invisible, we halted our attempt to remove the wax. Since the summer of 2015 I have made several inquiries of conservationists about how to best treat the Terranuova inscription, but since no one has been able to provide a clear recommendation nothing further has been done with it.