Linda Page Cummins

Correr 336, Part 4: A New Compendium of Late Medieval Music Theory

Saint Bonaventure, on the ways of making books: Someone … writes the materials of others, adding but nothing of his own, and this person is said to be the compiler.

(A. J. Minnis, Medieval Theory of Authorship)

The manuscript Venice, Biblioteca del Museo Correr, Correr 336, part 4, is devoted principally to a compilation of Latin music theory texts dealing with hexachords (including coniunctae) and mutation, intervals, and modal theory, all presented without attribution.[1] The compiler claims to have made the collection for his own use and that of pupils; thus it represents the interests of a musician who was also a teacher and who chose material that he considered practical, organized in a way he intended to be useful. Because I have been able to find concordances for almost all of the texts included in the longer of the compendium’s two main sections, comparison with these sources has made it possible to determine what portions of pre-existing material the compiler included, what he omitted, what he changed, and how he imposed a design on his borrowed material. The comparison thus reveals his attitudes, concerns, and prejudices, as well as his perception of his role as compiler. The work also provides an example of a compendium or compilation that is far more than merely a series of copied sources.

The present MS Correr 336 is a composite volume consisting of four separate manuscripts. At some point after these were bound together, the folios were numbered sequentially 1-456 (with some errors) at the bottom center of the rectos; earlier foliations are evident in the last three manuscripts (this will be explained for part 4 below). According to a table of contents on the second front flyleaf, the first manuscript contains the Tractatus de admirandis et secretioribus philosophiae arcanis by Giovanni Mariano Buri, the second the Opera spirituale of Bartolomeo Mozzi da Seravalle, dedicated to Clement XI. The third manuscript contains a hand-written copy, dated 1502, of the 1496 print of Franchino Gaffurio’s Practica musice, followed by a treatise in Italian on mensuration (Perche in ogni cosa la brevita et expeditione de quella e cosa degna como se dice in Brevitate moderni gaudent …), the latter not mentioned in the table of contents.

The table of contents makes no reference whatever to the fourth manuscript, which contains the music theory compilation; this omission may in part explain why this fourth manuscript was long overlooked. Giuliano di Bacco identified it as a source for Divina auxiliante gratia in his 2001 study of the transmission of the Ars contrapunctus of Johannes de Muris.[2] A description of the manuscript first appeared in RISM’s Theory of Music 6 (2003).[3]

Correr 336, part 4, consists of 32 paper folios (215 x 157 mm) gathered in two fascicles of eight bifolios each; there are two visible watermarks: a crown somewhat similar to Briquet 4752 (Carignola 1453) on ff. 425-441 and three mountains surmounted by a double cross somewhat similar to Briquet 11778 (Padua 1490-1502) on ff. 443-454, both signalled in RISM Theory of Music 6. Though the folios of part 4 are now numbered 425-456, the original numbering of the bifolios (1-8 for the first half of the first fascicle and 9-16 for the first half of the second) is still visible in the lower right corners of the rectos; thus we know that this document was intended as a work, or at least as the opening of a work. The collection of music theory was copied onto the first twenty-three folios (ff. 425-447) in medium brown ink in a not particularly refined cursive humanistic hand; paragraphi and some captions are in red, as is a good deal of underlining of text within the writing block. A collection of Psalm and Magnificat tones (ff. 448-456), copied in a more careful book hand using darker ink, seems to be the work of a different scribe, and I do not consider it in this essay.

The crudeness of the handwriting in the collection of music theory and the sometimes fanciful but rarely carefully drawn paragraphi in red ink lead a reader at first sight to consider this a carelessly produced document—an impression not ameliorated by a crowded writing block (unlined, but with the four margins drawn in pencil) in which the number of lines varies between 32 and 44, initials that appear to have been placed thoughtlessly, a confusing array of captions in both brown and red ink (captions in red sometimes duplicating those in brown, sometimes centered but more often placed in margins or at the ends of lines or squeezed between them), and a repetitious and seemingly haphazard sequence of topics—intervals, for instance, are discussed four times, modal theory five times, hexachords and mutation six times; the various texts offer differing and sometimes contradictory views. This first-sight impression is far from correct, but discerning the compiler’s design depends on both the identification of the sources from which he borrowed and the unraveling of visual cues, particularly initials and captions.

The largest initial in the treatise, and the only one written in red ink and within space left for it in the writing block, marks the source identified by RISM’s Theory of Music 6 (pp. 629-30) as a digest of plainchant (on ff. 434r-444r), drawn "in part" from Marchetto’s Lucidarium; this is, in fact, Divina auxiliante gratia, drawn not in part but entirely from the Lucidarium and known from five other manuscripts.[4] This copy of Divina includes the chapter titles present in four of the five other versions of the text;[5] they are written in the brown ink, given adequate space and often centered, and appear to have been part of the first layer of copying. Though the entire caption (Capitulum de plus the topic) appears in the first entries, the label Capitulum is soon dropped, leaving only de plus the topic. (The only four red captions found within the Correr Divina were squeezed into space underneath examples, and do not appear as captions in the other versions of that text.)[6] Initials that mark subdivisions of the Divina text—much smaller than the opening of Divina but as large or larger than any others in other sections of the Correr text—are also in brown, placed within the writing block, and often decorated with red; this also fits the pattern of other Divina copies. Thus it would appear that the Correr scribe copied the Divina text from an exemplar not very different in layout from most other known sources.

If the Divina text is excluded from consideration of the physical appearance of the Correr document, two things become clear: only two initials of considerable size and only two captions that are centered and in brown ink remain for consideration.

The two large initials, the opening N (Nota quod sunt tres modi cantandi) and the S (Sciendum quod mutatio) on folio 426v—both written in brown ink, in the left margin outside the writing block, and decorated in red—mark the points at which the compiler began to copy from his other two main source texts. As the compiler in neither case began at the beginning of the source texts as they are known from other manuscripts—that is, did not begin with the incipits that usually identify such texts—these initials are the only vestiges that remain to signal the beginning of something new. The RISM Theory 6 entry indicates material corresponding to chapter 8 of Berkeley’s First Treatise ("p. 84 et passim," referring to Ellsworth’s edition), beginning in Correr on f. 429r (De motetis et aliorum cantuum iudicium iudicandi); but Berkeley material actually begins on f. 426v (marked by the initial S just mentioned) and corresponds to the second through the eighth chapters, that is, the bulk, of Berkeley’s First Treatise. This extract includes not only the well-known reference to mode in motets and ballades noted by RISM, but also the discussions of mutation, mode, and very significantly coniunctae (hexachords built on notes other than C, G, and F).[7] RISM does not mention a third substantial concordant source, Rome, Vallicelliana C.105, 119r-123v, from which La Fage published excerpts.[8] The Correr compiler takes material from near the opening of the C.105 treatise as his own opening statement, following that text through the end of the discussion of plainchant, breaking off where C.105 begins a discussion of counterpoint.

Outside Divina, with only two exceptions, all other captions in Correr (excluding labels for musical examples) were written in red in spaces too small for them to have been allocated in advance, and thus were most likely part of a second layer of copying; their inclusion may in fact have been as much an afterthought as their physical appearance suggests. Perhaps the addition of red chapter titles, almost all beginning with de plus the topic in the manner of the Divina captions, occurred to the scribe after he had copied the Divina text.

The two remaining brown captions signal the large-scale structure of the Correr treatise: first, the title Ars cantandi at the beginning, followed by the later addition of scilicet de modis cantandi in red; second, Manus on folio 444v. These brown captions introduce the two major divisions of the Correr compendium: Ars cantandi is a comprehensive survey of the art of plainchant; Manus is at once a review and an exhortation to pupils to commit the information presented to memory. These sections having been established, it becomes possible to address the seemingly haphazard and repetitious sequence of topics. I will consider the two sections in order.

ARS CANTANDI

The discussion of hexachords and mutation in Ars cantandi is drawn principally from unknown sources concordant first with Vallicelliana C.105, then with Berkeley; but the compiler seems to have tailored each to create a more logical sequence of material. He begins (without prologue) with C.105’s discussion of the properties of hexachords ("There are three manners of singing, square , nature, and soft or round b"),[9] and continues by identifying the three initial letters, C, F, G ("In every place where G is found, there is the beginning [of the hexachord] of square , that is, ut; and [when ut is] C, there [is the hexachord of] nature; and [when] ut is F, there is the beginning [of the hexachord] of round b, that is, soft b.")[10] He omits two short passages on mutation from C.105 (items I/A/1 and I/A/4), as they anticipate material better introduced later. Berkeley will provide the bulk of his discussion of mutation, but as the Berkeley section 1.2 (in Ellsworth’s edition) begins with a definition of disiuncta ("A disjunction is a violent transition from one hexachord to another, without whatever mutation of syllables might be possible there"),[11] the Correr compiler omits it, probably deeming it inappropriate for opening a discussion of mutation. Instead he inserts a straightforward definition of mutation and a discussion of its fundamentals drawn from the tradition of Johannes de Garlandia (I/A/7/a).[12] Only after this does he turn to Berkeley’s exhaustive discussion, encompassing not only the conventional hexachords (item I/A/7/b) but coniuncta hexachords built on EE^b, FF, A, B^b, D, E^b, a, b^b, d, e^b, and aa as well (item I/A/7/b). After a discussion of mode, which the Correr compiler includes (to be discussed below), the Berkeley text returns to examples of hexachords, mutations, and coniunctae (item I/A/9/a-e), which the Correr compiler supplements with examples and a further discussion he either invented himself or drew from unidentified sources (items I/A/9/f-g). Thanks to the compiler’s judicious cutting and pasting, material on mutation from the three sources follows in a logical and progressive manner. (The same is true of the treatment of mode he constructed, as we shall see below.)

Though he omitted Berkeley’s definition of disiuncta, the compiler of Ars cantandi retained its examples of the practice (item I/A/9/c), with a significant modification. He replaced the diminished fifth with the less objectionable minor sixth (by definition a disiuncta since it cannot be negotiated within a single hexachord). He also replaced Berkeley’s caption in which these disiunctae are called "notable" (notabiles), replacing that term with "good" (bone)—the strong implication being that he did not find the diminished fifth "good." And in fact he omits the diminished fifth both from the list of intervals and their definitions (I/A/3)—even though the concordant texts in C.105 include it[13]—and from the examples of intervals he provides in item I/A/9/h; this list, the definitions, and the examples replace Berkeley’s chapter 9, whose discussion of intervals includes the diminished fifth.[14]

The compiler introduces modal theory on f. 425r (item I/A/2), in simple statements drawn from C.105: "Note that there are four letters that are finals, low D, E, F, and G... Note that there are eight tones, four authentic and four plagal."[15] He omits a discussion of the notes of the gamut as well as material on register and mode duplicated in the previous passage, but continues with the classification of mode by species of diatessaron and diapente, the five classes of tones (perfect, imperfect, pluperfect, mixed, intermixed), rules for choosing between b natural and b flat in the first tone, and the doctrine of employing the cord (the note a third above the final) to determine, in ambiguous cases, whether a mode is authentic or plagal. These statements present the fundamentals of the modal theory of Marchetto of Padua, pervasive in Italian music theory treatises of the fourteenth and fifteenth centuries.[16]

At 427v (item I/A/8), Berkeley provides a more conventional, but a more complete, doctrine of plainchant: from Berkeley 1.5 (Ellsworth, 66-75) the definition of mode, the "original four authentic modes of the Greeks" (protus, deuterus, tritus, tetrardus) expanded by the addition of four plagal modes to create eight, the ethnic names of the modes, and ranges of the modes (regular and irregular; the latter with ranges greater than a tenth); from Berkeley 1.6 (74-81), the four finals and the modes that use each, then the initial letters of the various modes; from Berkeley 1.7 (80-85), examples of the psalm tones. Significantly, given his lack of hesitation to omit or change material from his sources, the Correr compiler does retain Berkeley’s doctrine of mode in polyphonic music.

The culmination of the modal theory is the full-blown treatment of Marchetto’s doctrine in Divina auxiliante gratia, ff. 434r-444r (item I/A/10). Divina includes material from treatises 9 and 11 of the Lucidarium nearly complete (that is, the bulk of Marchetto’s modal doctrine, and the preliminary treatment of intervals), treatise 12 on the cord, treatise 13 on rests, and treatise 14 on clefs and registers. Divina developed its own manuscript tradition, appearing in six sources (see Contents, note 32); as only one of them contains the Lucidarium itself (I-PIu, 606), and that one is devoted in large part to Marchetto’s works, one may say that in the other five manuscripts Divina functions as a primary representative of his modal theory, standing in for the longer treatise. Thus the compiler’s discussion of mode progresses from the simple statements of C.105 to more sophisticated treatments, covering both the traditional doctrine of Berkeley and Marchetto’s more radical approach.

Following the Divina text, which does not include Marchetto’s subdivision of the tone, the scribe appends a short passage (f. 444r-v, item I/A/11) dealing with the tone, semitones, and the smaller dieses taken from Martianus Capella’s Nuptiae Philologiae et Mercurii. This may be a nod to one of the most controversial features of Marchetto’s theory, his division of the tone into five dieses. Perhaps the compiler of the Correr manuscript only knew Marchetto’s reputation but did not have a text of the Lucidarium to copy and so used another text in which the word diesis appears (though these dieses are not those Marchetto had described); another Marchettan digest, Sciendum quod antiquitus, mentions Marchetto’s division of the tone at the end of a work otherwise devoted entirely to modal theory.[17]

MANUS

The second section of the Correr compendium (item I/B) opens with the heading Manus and closes with a diagram of the information so often projected onto the hand: the scale, the letters, hexachords and their syllables, registers, etc. Here the compiler turns to practicality, issuing an exhortation to pupils to learn terms, "since the beginning of all knowledge is to know terms, as the Philosopher [Aristotle] says in the first book of the Posterior Analytics,"[18] and to commit to memory the material here presented [item I/B/1]. He begins appropriately with gamma, the first letter of the hand, and with an explanation of its name; this corresponds to the subject matter he had deleted from the beginning of Vallicelliana C.105, and also from the beginning of Divina auxiliante gratia. (The opening of Manus is similar to, but not a literal repetition of, the Divina opening; see below.) He continues with a review of the Ars cantandi material, repeating the introductory paragraph on mutation literally, several times referring the reader back to information given "previously" (antea); but in this section, information is presented in formats more conducive to memorization.[19]

Mnemonic verses for remembering the finals of modes appear on f. 447r (item I/B/7):[20]

A note introducing the verses calls them "a certain rule for recognizing the tones, briefly presented—the rule placed at the beginning notwithstanding";[21] the discrepancy signalled by "notwithstanding" (non obstante) is surely between the verses given here—easily memorized in that all the cofinals lie a fifth above the corresponding finals even though the high D, often cited as cofinal of the seventh and eighth modes, is really unsuitable for them, as the third above it is minor—and the more accurate listing at the beginning of Ars cantandi, where the cofinals are given as A, square B, round , and high C.[22]

An exhaustive list (item I/B/5) of the notes of the gamut with their letter names, syllables, hexachords, and mutations fully spelled out is, thanks to its repetitions and formulaic presentation, also in a form conducive to memorization. The following are typical of the entries, which extend from Gamma-ut all the way up to E-la:[23]

After this list, the compiler cautions: "And all these you are to commit to your memory,"[24] then he advises readers to turn back to examples of mutation "notated on folio 5 and following"[25] (i.e., ff. 429r-430v of the current numbering), where such examples do, in fact, appear. I shall return to this last point.

CONCLUSION

Many of the compiler’s text manipulations have been noted; one more remains. In every other copy of Divina auxiliante gratia, the preface reads (minor variations aside):

In place of this, the Correr compiler introduced Divina only with the phrase "Divina auxiliante gratia, etc." Did he do this because he regarded the opening as too well-known to bother copying? Because he intended to use the remainder of that first sentence later (which he did)? Because he deliberately wished to suppress the attribution? Though he relocates the words that in other manuscripts follow his "etc." to the opening of Manus (and it is this relocation that proves his copy of Divina included the full preface), he never names Boethius or Marchetto. There is, in fact, nothing from Boethius in Divina; Boethius seems to have been mentioned only auctoritatis causa by its compiler. The compiler of the Correr compendium seems to have felt little need of auctoritatis causa or indeed of auctoritas, whether real or invented. Of the three major texts he chose, only Divina includes an attribution in the extant sources; the concordant sources for his shorter excerpts (Martianus Capella, for example) may include an attribution, but it is not present in the passages he selected.[26] Whether by choice or by chance, the Correr compiler needed only the deliberate omission of the Divina attribution to create a text that lacks any attribution at all. In fact, the Correr text lacks any appeal to authority other than to God and Aristotle. The following excerpt gives the opening of Manus.

The reference to Aristotle is not an appeal to authority to lend credibility to the information presented; rather it is an appeal to students, urging them to learn, to memorize. Words from the opening of Divina have been relocated and refocused to serve what may very well have been the compiler’s primary purpose—though, as taken from Divina, it occupies second place: "for the profit of pupils."

The deliberate excision of material from a source drawn on for Ars cantandi and the relocation of it into Manus suggests that whoever compiled the two thought of them as a unit; but can Correr 336, part 4, be seen as the original version of the compilation? That is, did the creator of Correr 336, part 4, compile his text from separate documents, or simply copy a text that already existed as a unit in some other manuscript? Consider the treatment of initials in Ars cantandi. In the part of the text that corresponds to material in Vallicelliana C.105 there is only one large initial; it is placed in the margin and decorated with red tracery. In the part that corresponds to material in the Berkeley treatise, the initials are smaller and undecorated but still placed in the margins. In the part that presents Divina auxiliante gratia and the text from Martianus Capella, initials are relatively large and placed within the writing block; the first of them, a D that begins the Divina text and is the largest initial in the manuscript, is uniquely in red ink. The first initial of Manus is again in the margin. These variations in style strongly suggest that the scribe of Correr took not only the words he copied but their visual style from various exemplars. If so, both the relocation of text from Divina to Manus and the folio-number reference in Manus back to a passage in Ars cantandi represent efforts of the compiler of Correr 336, part 4, to unify the disparate texts he had borrowed from various sources and arranged in a feasible order.

In the first sentences of the opening of Manus (above), the Correr compiler echoes St. Bonaventure’s definition of the compiler: he has written "not in my own words but in those of others," but with one caveat: "almost completely"—and he echoes this definition using the words of another (the Divina compiler)—"almost completely." As Parkes and Minnis have shown, much of the contribution of compilers in the later Middle Ages consisted of imposing order on the material they copied, often by dividing a large work into sections, providing new headings, making tables of contents, even reorganizing material—all to make the information more accessible to the reader.[28] The Correr compiler, working with multiple sources, has performed this same function. He has made the material of music theory accessible by ordering from simple to more complex, by omitting material (and perhaps by adding headings) to make smoother and more logical transitions from one topic to another, and by dividing his material into two units: knowledge on the one hand, and on the other the practical methods of absorbing that knowledge.

The Correr compiler has also described his working method; he chose his sources as though "plucking flowers"; but it is clear that he has not created this bouquet willy-nilly. It is a meticulously crafted arrangement of carefully chosen flowers that the compiler has pruned, selected, separated, and rearranged to create a bouquet whose design is logical, practical, and beautiful.

For the profit of modern scholars, the Correr compiler offers another glimpse into the concerns of the medieval musician, concerns that center around the selection of appropriate material and the dissemination and teaching of useful and practical information; the Correr treatise also illustrates the liberties a compiler may feel free to take with his source texts, presents a compilation that exhibits a unity that simple collections often lack, and thus offers an expanded view of the variety of forms a compilation may take; and the manuscript itself should alert scholars to the possibility that physical appearance may veil, rather than reveal, a logical underlying plan.

Appendix – Contents of Venice, Biblioteca del Museo Correr, Correr 336, part 4: