The historian John Keegan was one of the first to ask the simple yet searching question: what actually happens in combat? It is well known that English footsoldiers received a charge by French knights at the battle of Agincourt in 1415, but what took place when men and horses collided? Keegan gives his answers in The Face of Battle, and it may be time for musicologists to modulate the sonorous questions that he poses there for their own purposes. What actually happened, for example, when a motet by Johannes Ciconia was performed in northern Italy c 1400? When friends and associates gathered together to hear such music, what was the nature of their various musical aptitudes and interests? Did women participate in the performances? What was the role of instrumentalists? Some of these questions, no doubt, will never find an answer; there are no medieval chronicles devoted to musical gatherings as there are chronicles – and many other writings – devoted to battles like Agincourt. None the less, literary and iconographical sources are among those which may still have something to reveal about ‘the face of performance’ (to coin a phrase after Keegan's own), and the purpose of this article is to examine the contents of one that has been unjustly neglected: the Tractatulus de differentiis et gradibus cantorum by Arnulf de St Ghislain. This brief treatise classifies the kinds of musicians who performed or admired polyphonic music and is therefore quite exceptional among the works loosely classified as medieval music theory.

Few extant collections of late-medieval polyphony have been linked definitively to composers whose works they transmit. A dearth of documents that are literally ‘authoritative’ – in the sense that they come directly from the hand of the author – is not surprising in light of the small survival rate of fourteenth- and fifteenth-century manuscripts and the likelihood that pieces were not written down for the first time in versions for presentation. Most surviving manuscripts appear to be several generations removed from written originals. Nevertheless, presentation manuscripts and other anthologies were not necessarily compiled by scribes in isolation. When composers were nearby they probably were consulted, either directly or through scribal access to authoritative written exemplars. Therefore, despite a paucity of documented relationships, some extant manuscripts may have been compiled in close association with the composers whose music they contain, perhaps even under their guidance.

The extensive repertory of English anthems composed between the passing of the Act of Uniformity (1549) and the cessation of church services precipitated by the Civil War during the 1640s has been the focus of such concentrated attention in recent years that, on first sight at least, few important facets appear to have languished in neglect. Amongst those aspects of the anthem which have been subjected to detailed scrutiny are the genre itself,' the associated printed and manuscript music sources, the many vexing problems of performance practice, and the anthem settings of most of the principal composers. One conspicuous omission from this research profile, however, is any survey of anthem texts, a subject of such fundamental importance as to refute any suggestion that the present understanding of the anthem genre is complete.

Our understanding of the long-term tonal relationships that grow out of seventeenth-century harmonic language is at an elementary stage. An enormous gap seems to exist between our ability to deal with, on the one hand, a sixteenth-century compositional language that basically adheres to a contrapuntal technique and tonally abides by the rules of modality and, on the other, an eighteenth-century tonal language for which we quietly assume harmonic functionality. That scholars have largely avoided an investigation of harmonic language in the seventeenth century is perhaps surprising. Problematic and somewhat enigmatic features of seventeenth-century music are, in some cases, not very different from the characteristics Harold S. Powers attributed to late sixteenth-century modal music: ‘sometimes faintly exotic, often charmingly vague and undirected to our ears, but hardly alien’. Benito V. Rivera appeared to corroborate Powers's perception when he characterized some of the idiosyncrasies of seventeenth-century harmonic practice as ‘isolated harmonic quirkiness’.

In a study of Verdi's debt to Bellini published over 20 years ago, Friedrich Lippmann set the course for subsequent investigations of phrase structure in ottocento opera by using letters to describe the archetypal shape of many melodies written by Italian composers after 1830. He identified the most common pattern in the music of Bellini and the young Verdi as a1 a2 b a2, in which each letter represents a four-bar phrase in a prototypical 16-bar melody. The method has obvious risks of turning uncomfortably antiseptic for the analyst, but there is much to be said for its virtue of concision. It served Lippmann for his subsequent work on Bellini as well as Julian Budden for his three-volume study of Verdi's operas. With a minimal addition of bulk, Joseph Kerman and Scott Balthazar subsequently formulated a more elaborate alphanumeric system. Both replace Lippmann's use of arabic numerals to demonstrate musical variation with primes (AA' BA’ instead of al a2 b a2; Kerman uses lower-case letters, Balthazar upper-case) and they communicate the number of bars in each phrase, including those longer and shorter than the four-bar norm, with subscript numbers (for example, A4A‘4B3C6). The relationship of each phrase to the standard double quatrain or sestet of Italian poetry is also accounted for by a parenthetical tracking of verse lines for each musical phrase; for example, A4(S11–2)A'4(S13–4) illustrates how the first quatrain is distributed across the first eight bars of a normative melody. More important, Kerman and Balthazar recognize the musical function of phrases within the prototypical 16-bar group. Balthazar, in particular, identifies the first two four-bar phrases as an ‘opening thematic block’, the B phrase as both ‘contrasting’ and ‘medial’, and the last phrase as a ‘closing’ unit ('some version of A or C, depending on whether motives from the opening phrases are present).