The first known collection of musical terms is one of the oldest extant musical documents. Nabnitu (‘Creature’), a compendium of the Old Babylonian period (c1800 BCE), treated all areas of human activity and is typical of Mesopotamian scribal instincts. Book XXXII, one of a small corpus of cuneiform musical texts, preserves intact the nine canonical string names and their arrangement, followed by a long and fragmentary list of tunings, instruments and instrument parts, most of which remain unidentified. Many of the Akkadian terms are given their Sumerian equivalents, implying a still older tradition of which Nabnitu XXXII was merely the codification. In fact, some of the Sumerian terms appear in musical contexts as early as the 24th century BCE. Besides Nabnitu, which continued to be copied through the neo-Babylonian period (c300 BCE), the first important non-Western treatise is Bharata's Natyashastra, a Sanskrit work giving a comprehensive account of dramaturgy in all its aspects, including music. In its present form, the Natyashastra is generally agreed to date from the 2nd or 3rd century CE, but its jumbled inconsistency reveals the incorporation of much earlier material of indeterminate date and origin. The problem of chronology and sources plagues Sanskrit literature in general, not least the Dattilam of Dattila, a compendium of ancient Indian music first published in a translation by Wiersma-te Nijenhuis in 1970. Itself dating from around 700 CE, one of its verses attests the existence of earlier terminological dictionaries: ‘One should understand the words from common practice with the help of manuals of terms and other [books]’.

The first two centuries of Greek musical literature, between the Peri mousikês of Lasus of Hermione (late 6th century BCE) and Aristoxenus’ extant Elementa Harmonica, have been lost. In any case it was not until the 4th century BCE that the first true musicologists – Glaucus of Rhegium, Heraclides of Pontus, Phaenias of Eresus, and Aristoxenus himself – began to make systematic collections of Greek musical history. These treatises, now lost, were key sources for Imperial compilers such as Athenaeus and pseudo-Plutarch, and ultimately supplied much material for the late lexica. Such compilations reflect the Greek cosmopolitanism, with its more generalized forms of language, literature, art and music, which was the hallmark of the Hellenistic age.

In such a form Greek music and musical literature came at length to Rome, although it remained a somewhat exclusive art. Vitruvius, the master builder of Augustan Rome, warned that ‘harmonikê is a dark and difficult subject, especially for those who do not read Greek’. In Book V of De architectura, Vitruvius reports and defines many Greek musical terms which are not strictly relevant to his theme; chapter IV in particular is, in effect, a detachable musical dictionary. Earlier still was Varro’s Disciplinae (1st century BCE), a comprehensive textbook based on the Greek system of education. Although Varro was eccentric in defining nine subjects rather than the canonical seven of later antiquity – nothing remains of the section on music – the work was an important prototype of encyclopedic form in its organization of material into such broad categories. With the addition of voluminous notes, a practice apparently begun by Pliny the Elder in his encyclopedic Historia naturalis of c77 CE, the Trivium and Quadrivium provided the basic format for most encyclopedic works up to the 17th century.

Although Walther did not mention Pliny’s work (which contains virtually no musical material), he gave a description of Julius Pollux's Greek lexicon, the Onomasticon (c180 CE), an excellent example of the format pioneered by Varro and Pliny. This work defines many musical terms relating to form and style, instruments, dances and drama; both it and Pliny’s Historia were known to Brossard. Pollux’s contemporary, Athenaeus, compiled the Deipnosophistae (‘Learned Banqueteers’, c200 CE), the most famous example of that curious genre in which arid stockpiles of trivia are made more palatable through being served as witty dinner conversation. Lichtenthal amplified Walther’s description of the Deipnosophistae, noting the musical terms and topics which appear in it. This gastronomic glossary is of special interest for its origins in previous collections, particularly the huge lexicon of Pamphilus of Alexandria W (c50 CE), itself derived from many earlier sources now lost. Both the Deipnosophistae and the coeval De musica attributed to Plutarch, which features a less imaginative party of diners, are of still greater importance for preserving extensive fragments of the 4th-century musicologists; indeed, these works are little more than digests of Aristoxenus and his colleagues.

Some time between the 1st and 4th centuries CE, Aristides Quintilianus produced his De musica, the closest any ancient writer came to a comprehensive encyclopedia of music. The author begins by decrying earlier piecemeal discussions of music, then presents a systematic account of harmonics, metrics, composition, instruments, notation and acoustics, including extensive treatments of the physiological, psychological and cosmological aspects of the art. Some of this material may well go back to the Archaic period (6th century BCE). About 420 Martianus Capella, again borrowing freely from many predecessors, produced his curious De nuptiis Philologiae et Mercurii, an allegorical fantasia on the liberal arts. The ninth book (De harmonia), drawing heavily on Aristides, was printed in Marcus Meibom's Antiquae musicae auctores septem (1652) and Gerbert's Scriptores (1784). Walther will have known the work from his reading of Meibom. About 500 CE Stephanus of Byzantium compiled his De urbibus, noted by Forkel, Becker and Lichtenthal, and about the same time Boethius wrote his well-known De institutione musica. The latter exerted a great influence on Cassiodorus, compiler of the notable De artibus ac disciplinis liberalium litterarum of c560; chapter 5 of book 2, Institutiones musicae, is printed in Gerbert and was apparently well known to Walther.

One of the most authoritative reference books of medieval Christendom, the Etymologiae of Isidore of Seville, was completed about 600. Isidore was especially concerned with the words used by his clerical brethren, adding to his systematically arranged encyclopedia two dictionaries of terms, one alphabetical, one topical. Among the most widely-used books of the Middle Ages (there are about 1000 surviving manuscript copies, according to Collison, 1964), the Etymologiae became a source for countless subsequent lexicographers and exerted a profound influence on learning for a millennium. Eggebrecht (‘Lexika der Musik’, MGG1) believed that the 11th-century Vocabularium at Monte Cassino was an epitome of the work; Bartholomeus Anglicus in his 13th-century De proprietatibus rerum repeatedly stated that things were ‘As Isyder sayth’; Walther and Brossard both noted it. In the Handwörterbuch der musikalischen Terminologie (1972–) it is cited as an authority for meanings; there are studies of its musical importance by Tello, Fontaine and Avenary; excerpts are translated in StrunkSR1.

The Myriobiblon, sive Bibliotheca librorum, written in the 9th century by Photius, a patriarch of Constantinople, is important even today for its extracts from and comments on the writings of nearly 300 authors whose works are otherwise lost. Such syntheses were often more popular, more frequently reproduced and more enduring than the originals from which they were drawn, a tendency already visible in the various anthologiae or florilegia of later antiquity. Also from the 9th century come two Chinese encyclopedias: the T'ung-tien (‘Complete institutions’) by Tu Yu (735–812), chapter 5 of which is devoted to music, and the Yüeh-fu tsa-lu (‘Miscellaneous notes on music’, c890–900) of Tuan An-chieh. The latter is a remarkable work made widely available in a translation and study by Gimm. Chapters 13–26 cover instruments and their masters, 27–40 individual compositions; chapter 41 is on theory and 42 on institutions, including floor plans for music schools.

The relationship of Arabic musical works to the Western classical tradition remains largely unexplored, involving both the adaptation of Greek authors and the later reintroduction of this material to the West via Spain and in Latin translation. From the 10th century come the Mafa¯hti¯h al-’ulu¯m (‘Keys to the sciences’, 975–7) of al-Kwarizmi¯, with three chapters in book 2 devoted to music, and the compendious Kita¯b al-agha¯ni¯ al-kabi¯r (‘Great book of songs’) by al-Isaha¯ni¯, which includes biographies of musicians. The latter has been frequently translated and reprinted in whole or in part, but the place of both works in the history of music lexicography remains unclear.

Popular throughout the Middle Ages, the 11th-century Suda lexicon circulated widely in manuscript before its first publication (Milan, 1499) and was still in use in the 17th century. It was exceptional in eschewing the conventional systematic arrangement of medieval encyclopedias, by book and chapter based on the seven liberal arts. Instead, it was arranged alphabetically, and its nearly 30,000 entries embrace proper names of people and places as well as terms. Although it draws mainly on Greek literature, some Roman scholarship is also included. Forkel, Lichtenthal, Becker and others remarked that it contained many musical terms and ‘historical notices’ about music, some of which are now seen to belong to pre-Aristoxenian musical traditions deriving ultimately from the lost works of the 5th century and even preserving traces of an oral theoria of the Archaic period. Hugh of St Victor, writing in about 1127, did employ the traditional categories of systematic arrangement in his Didascalion, adding new ones as well: ethics, crafts and physics. He avoided the catholic secularity of the Sudalexicon, keeping to more monastic concepts and drawing heavily on Boethius, Augustine of Hippo, Cassiodorus, Plato and others.

Another widely known encyclopedia of the Middle Ages and Renaissance, and one of those most readily available to musicians because Hawkins reprinted it in his General History of Music (1776), was De proprietatibus rerum by the English Franciscan friar Bartholomeus Anglicus, who graciously acknowledged his indebtedness to Isidore. It was written about 1230 and translated frequently (into French, Spanish, Dutch and English in the 14th century) before the first edition was printed in 1472. An English translation made by John of Trevisa in 1397 appeared about 1495 and was republished frequently, several times in an edition enlarged by Stephen Batman. Brossard noted it among those works he would have liked to study; Walther provided a biography of Bartholomeus in his Lexicon, admitting that he saw the name in Brossard's list. Imitating Bartholomeus and also dependent on Isidore (to the point of obvious plagiarism) was Vincent de Beauvais' Speculum maius, compiled about 1244. In spite of its borrowings from Isidore (or perhaps because of them), it became one of the major encyclopedias of the Middle Ages, and now furnishes ideas about tastes, prejudices and intellectual concerns in the 13th century more clearly than many of its counterparts. Walther mentioned it, but apparently did not see it, although he does seem to have known Johannes Balbus's contemporary Catholicon, which, like the Suda lexicon and many later works, included proper-name entries.

Brunetto Latini's Li livres dou trésor of about 1264 was probably the first major exception to the tradition of writing in Latin. His departure from it was more than just a change of language; it indicated a change of audience. This was the first encyclopedia compiled for laymen, and it became immensely popular. Between Guglielmo da Pastrengo's De originibus rerum libellus (c1350) and Johannes Tinctoris's Diffinitorium (c1495), there were few lexica of importance. Eggebrecht (‘Lexika der Musik’, MGG1) discussed several general terminological dictionaries, including Gerardus de Scheieren's Vocabularium (1477) and Wenceslaus Brack's Vocabularius rerum (1483), which, although unrelated to the work of Tinctoris, exemplify the growing demand for translations of Latin terms into other languages. These Latin–German dictionaries show clearly the derivative nature of most lexicographical work at the time; Brack, for example, derived most of the definitions for his 80 musical terms from Isidore via Hugh of St Victor (Eggebrecht, ‘Lexika’, RiemannL12).

Tinctoris's Terminorum musicae diffinitorium, published in the 1490s but written before 1475, is certainly the most important musical incunabulum. Nearly 300 terms from a wide range of musical matters are defined with almost cryptic conciseness. Parrish noted in the preface to his excellent translation (1963) that many of the definitions reappear verbatim in some of Tinctoris's later treatises. The sources that Tinctoris used for these definitions cannot be clearly determined. It is unlikely that he relied on any of the general encyclopedias then in circulation; nor does he appear to have used the 11th-century Vocabularium in Monte Cassino, with definitions (unlike Tinctoris's) derived mainly from Isidore. He may have relied solely on earlier music theorists, for the Diffinitorium includes only theoretical terms: there are no proper names, aesthetic considerations or descriptions of musical instruments. Although eight copies of the printed book and three early manuscripts (one from the 15th century and two from the 16th, with some 19th-century copies of them) survive, there is no evidence that the Diffinitorium had any impact on subsequent compendia, and most musicians continued to use the works of Pollux and Isidore, the Sudalexicon and others.