Premise

[3.1.1] “The constructions of a given language do not form an unstructured set. Relationships between and among constructions are captured via a default inheritance network” (Goldberg 2013, 21).

Paraphrase

[3.1.2] Smaller constructions can be contained within larger ones. When this occurs, all or some of the meaning of the smaller construction is inherited. Inheritance can also occur from a common to a less common pattern of similar size. The order of constructions in discourse is flexible but not random. Implications from one pattern can suggest dependencies that constrain possible continuations by other patterns. And the learned statistics of construction successions can similarly affect what follows what. Constructions are not, however, rigid building blocks—they may overlap, be incomplete, and be recognized by some listeners and not by others. Constructions in language and music are not immanent in phonemes and tones but are the products of learning and, in particular, human memory.

Networks of Patterns

[3.2.1] Consider the idiomatic statement “He finally kicked the bucket.” Though the full meaning—“He finally died”—may be opaque to non-native speakers, the idiom does inherit both its form and a partial meaning from the transitive verb to kick, which in turn inherits features of the class of regular transitive verbs, like -ed endings for the past tense (Croft and Cruse 2004, 264). Once learned, that idiom makes the parallel idiom “She finally kicked the habit” relatively easy to understand. Fluency in a language requires that one learn to master a host of these “structured relationships.”

[3.2.2] Anyone opening a manuscript of elementary partimenti from one of the eighteenth-century conservatories of Naples will almost always see three basic cadences among the very first patterns presented. Their names, translated into English, are the Simple Cadence (in Roman numerals, V–I, though the students never saw a Roman numeral in connection with chords), the Compound Cadence (V4–V3–I), and the Double Cadence (V5/3–6/4–5/4–5/3–I). Young apprentices would then learn the Rule of the Octave (how to harmonize an ascending and descending scale in major and minor) and a number of sequences, each named by the intervallic movement of the bass. One such sequence was called “down a fourth, up a second,” today better known as the bass of Pachelbel’s Canon in D, or as part of the old basso ostinato called the Romanesca. With these few schemes in mind, we will examine the network of patterns in the opening measures of a partimento (Gj 244, ca. 1740) by Francesco Durante, one of the greatest teachers during the era of Bach and Handel.

[3.2.3] As shown in Example 6, the Romanesca was typically an eight-count pattern, where the last two counts support a cadence. The cadence “slot” permitted some variation: Pachelbel chose a Simple Cadence, Durante chose the Compound Cadence. Durante then repeats the Romanesca (mm. 3–4) but with scalar diminutions. Scalar passages, marked scala on the example, could inherit the thoroughbass norms of the Rule of the Octave, especially if the tempo were slow enough. In m. 4, Durante “switches gears”: the Romanesca’s cadential slot fails to develop, and the Passo Indietro (“a step to the rear” or V4/2–I6/3; see Gjerdingen 2007a, 167) redirects the key toward G minor and a Simple Cadence. This diversion takes two extra beats, with the result that the ensuing Romanesca begins in the middle of m. 5 (tenor clef). At the level of discourse, this third Romanesca initiates a return of the opening theme, but a fifth higher (G minor).

Layers of Inheritance

[3.3.1] As students worked their way through collections of partimenti with their masters (Sanguinetti 2012), they would see the same constructions used over and over in different instantiations and contexts. Take, for instance, a Romanesca presented by Durante in a more advanced type of partimento (see Example 7). Even without annotations it is not difficult to recognize an expansion of this construction (the quarter notes of Example 6 have, in Example 7, been stretched to half-note spans with scalar and octave-leap diminutions). With each new partimento the young apprentice learned new collocations, new diminutions, and which slots in a construction could be expanded or contracted. Expansion within a schema, however, is limited by our working memory. The Romanesca of Example 7, shorn of all its diminutions, would still be intelligible and recognizable as a series of half notes. A Romanesca where each core tone was separated by perhaps 30 seconds of diminutions would likely not be recognizable as a phrase, or recognizable only as a sequence of modulations by descending thirds. That would be at the level of discourse, and it is at that level that the return of the theme (Example 7, m. 6, tenor clef, here in C major) is clearly recognizable even after considerable intervening material.

[3.3.2] For these advanced partimenti Durante provided “styles” (modi) as guides for right-hand realizations. The style provided for Example 7 is shown in Example 8. It details how the octave leap in the second half of each measure cues a 4–3 suspension and resolution (i.e., the high A5 resolving to G5 in m. 1), a form inherited from the similar bass pattern of the Compound Cadence (cf. Example 6, m. 2).

[3.3.3] Inheritance and analogy can continue to branch out as styles change and different combinations lead to new configurations. Example 9 shows a partimento in C major by Nicola Sala, a student of Durante’s contemporary Leonardo Leo and later an important master himself. Sala has taken over Durante’s styling of the Romanesca schema, inheriting the 4–3 suspension of beats 3 and 4 and then, by analogy, creating 9–8 suspensions on beats 1 and 2 (beginning in m. 2).

[3.3.4] When the Paris Conservatory was founded in 1795, “the Italian School” was adopted as the classical model for instruction and emulation (Choron 1804). That decision profoundly affected the course of compositional training in nineteenth-century France. “Harmony” remained an artisanal praxis and not a theoretical subject (Masci 2013). If we examine the many four-voice harmonic models in the textbook of François Bazin (1857, 192; see Example 10), we will find an unadorned Romanesca bass in C major whose soprano and tenor parts follow the exact course of Sala’s partimento realization (though an octave lower). This is just one of many pieces of evidence confirming that the Paris Conservatory continued to transmit to new generations of French musicians the “classical” artisanal praxis of eighteenth-century Italy. The long subtitle of Choron’s book (1804) plainly sets out its focus on Italian exemplars: “Extracts from the best Authors: Leo, Durante, Fenaroli, Sala, Azopardi, Sabbatini, Padre Martini, and others. A Classical Work serving as an Introduction to the study of Composition, translated from the Italian and arranged in an order most suited to facilitate intelligence and practice.”

[3.3.5] French musicians did, however, adapt and modify this Italian patrimony. The French prodigy Henri Fissot studied with Bazin at the Conservatory, winning a first prize in harmony at age thirteen. Decades later, when he needed a simple descending passage for his Phantasie-Stück (op. 2, no. 8, a “genre piece” dedicated to a German friend, ca. 1878) he recalled one of the constructions he had learned as a child—an old-style Romanesca with parallel descending thirds in the treble and alternating fourths and seconds in the bass (see Example 11). He gave a more modern tint to the “down-a-fourth” bass tones by adding 6/5/3 sonorities (the third beats of mm. 21 and 22), thus subordinating an equally well-known and smaller construction, one where the first tones of rising steps in a bass often took 6/3 or 6/5/3 chords, especially when, as in m. 22, the ascent is by half step (the old “mi-rule” of thoroughbass practice). These changes are small but they collectively put the passage squarely into the sound-world of Fissot’s own time. He was not merely regurgitating an Italian heritage but speaking it creatively as a first language, though with a strong French accent.