Unheighted notation taunts us like a riddle written in unfamiliar shorthand. Functioning as a reminder for medieval singers who already knew the chant's melody by heart, it begs the perennial question of today's scholars: ‘What did chant sound like?’ For more than a century now, scholars have examined neumes as graphic clues which triggered melodies in the memories of singers. One of the latest scholars to take on this task is Dirk van Betteray, who has recently turned his attention to the possible interpretation and functions of liquescent neumes in the oldest layer of the St Gall notational tradition. His book argues that liquescence in notation is not simply the result of phoneme combinations in the text, but that these liquescents were used to emphasise the meaning of a word in a chant, under the requisite phonetic conditions. This work brings the semiology of Gregorian chant and medieval theology together in a new way. In 2005, van Betteray submitted a version of this study, under the same title, to the Kunstuniversität Graz as his doctoral thesis.

Van Betteray begins by placing his work in its theoretical lineage, providing enough background on the topic for those of us who are not experts in semiology to follow his subsequent arguments. Here, he surveys medieval writings which address liquescence, including those of Hucbald of St Amand (840–930) and Guido of Arezzo (992–1050), and several other fragments whose authors are unknown. He then offers a cursory examination of studies on liquescent neumes by Mocquereau, Pothier and Freistedt, published in the late nineteenth and early twentieth centuries, before addressing the importance of semiotic studies by Eugène Cardine (Semiologia gregoriana, Rome, 1968), Luigi Agustoni (‘I neumi liquescenti’, Musica sacra 83 (1959), 34–42), and many others. Narrowing his focus, van Betteray draws parallels between his findings and those in Andreas Haug's article ‘Zur Interpretation der Liqueszenzneumen’ in the Archiv für Musikwissenschaft for 1993, where liquescent neumes are understood as the graphic representation of an aural, phonetic phenomenon. This conflicts with Mathias Bielitz's hypothesis, as articulated in his Zum Bezeichneten der Neumen, insbesondere der Liqueszenz (Neckargemünd, 1998), which states that liquescent neumes have their own textually independent musical interpretation.

As an important centre of Western plainchant with a well-known scriptorium tradition, St Gall produced many sources which survive today, containing not only chant but also information about medieval Latin pronunciation in that region. In order to ensure that his corpus is as homogeneous as possible, van Betteray selects the contents of St Gall mass books for his study, excluding sequences, tropes and newly composed chants. He arrives at a list of fourteen St Gall manuscripts, dating from the tenth to the thirteenth centuries, and along with these he uses Einsiedeln 121, Laon 239 and Benevento 34 as statistical controls. He uses an exegetical approach for his comparison, reasoning that Gregorian chant is medieval theology made audible, and notes the particular significance of the Psalms, upon which most of the mass texts are based.

Since liquescence has most often been seen as the result of certain phonetic conditions, it is important to know how Latin was most likely pronounced in early medieval St Gall. Van Betteray asserts that the clear distinction between Roman and Germanic pronunciation is a theoretical construct and that each region probably differed, according to its own vernacular. An earlier theory, promoted by Agustoni, suggested that Roman pronunciation held sway in St Gall because liquescent neumes were found on certain consonants which could be considered liquid in Roman pronunciation, such as ‘g’ in ‘reges’. However, van Betteray's examinations show that ‘g’ is pronounced differently when followed by a bright or a dark vowel and ultimately refutes the claim that Roman Latin was spoken in (Allemanish) St Gall. Liquescence, he asserts, is not simply a guide to pronouncing Roman Latin for singers north of the Alps; it is a system of vocalised, closed sounds – linguistic occlusives – which give auditory emphasis to the phonemes they affect.

The problem with the explanation for liquescent neumes as a phonetic phenomenon is that they do not always appear under the conditions suitable for them. In order to show just how optional the use of liquescent neumes seems to be, van Betteray selects ten sources (mostly of St Gall provenance) and identifies four locations in which they may be found. Overall, the rate of occurrence is fifty-five per cent, which discredits the idea that liquescence occurs where pronunciation dictates. Now the stage is set for the main thrust of the argument. Since liquescence occurs only half the time when the necessary textual conditions are present, and since it is neither an aid for pronunciation nor a musical phenomenon, it must exist for another reason: exegesis.

The evidence for this new proposal is found in van Betteray's fifth chapter, under the heading ‘Die Liqueszenz als Schlüssel zur Textinterpretation in St. Galler Handschriften’. Following some preliminary notes on methodology, the reader is presented with ten individual case studies. Each proof begins with a table, where the instances of liquescence are noted and compared between sources. In every case, van Betteray suggests the exegetical reasons which might lie behind the presence of a liquescent neume in one instance and its exclusion from another. For example, sections of one introit and two alleluias are shown where the introit and the first alleluia have the same text, ‘Dominus dixit ad me’, and the second alleluia, Dominus in Sion in sancto, sets ‘in sancto’ to the same melody as ‘ad me’ in the first alleluia. The liquescence occurs on ‘ad’ and ‘san-’ in the text. Using quotations from Augustine's commentary on Psalm 2 and Amalar's Liber officialis, van Betteray explains why these particular syllables might receive special aural emphasis for theological reasons. He submits that liquescent neumes are used on these syllables because they are especially important for a correct understanding of the text, as illustrated by medieval theological writings: ‘Die Liqueszenz steht in jedem Fall heir an einer Stelle grosser spiritueller Dichte. Dort finden sich, wie sich noch zeigen wird, oft Liqueszenzen; vielleicht weil sie in bestimmten Kontexten zu grosse Lautstärke vermeiden oder zurückführen und so die Intensität der Aussage erhöhen können’ (p. 129). The remaining case studies make a strong argument for the exegetical use of liquescence, including an interesting case of double meaning and many clear textual connections between biblical commentaries and the sung texts of the liturgy. When manuscripts differ in their use of liquescence, he attributes this to divergent interpretations on the part of the scribe.

Further pursuing textual interpretation as a key to liquescence, van Betteray then isolates two especially instructive categories for examination. The first occurs when the same text is set to a different melody, and the second takes place when a different text is set to the same melody. The first category bolsters the textual interpretation theory; liquescent neumes are not found in the same locations (in the text) when they are set to different melodies. This means that these signs were used to indicate something other than pronunciation, and van Betteray again explains why singers would want to emphasise these words for theological reasons. However, the second category, where the same melody sets different texts, does little to support the exegetical hypothesis. In the case of typical melodies, liquescent neumes are found in the same locations consistently, seeming to disregard the meaning of the text entirely. Undaunted, van Betteray explains how other factors, such as melody, rhythm and articulation, contribute to textual interpretation as well.

Ever mindful of clarity, van Betteray offers a brief sixth chapter as a summary of his argument in eleven main points. After this, he examines broader connections between the core St Gall group and those in other groups of sources. For example, liquescent neumes are found less frequently in the later St Gall sources, especially after a quilisma, on an augmented liquescent pes, or on an accented syllable. He also notes that liquescent neumes are often replaced by episemas on single-note neumes in these later manuscripts. The ‘control’ manuscripts from Einsiedeln and Laon show only between 0.7 per cent and 3.4 per cent divergence, which is about the same rate found between the core comparison group and the younger sources. Benevento 34 represents the largest departure from the St Gall group, where divergent liquescent usage amounts to 23 per cent. Van Betteray explains this difference by recalling that Latin pronunciation would have been different between St Gall and Benevento and that substantial aesthetic and interpretational differences probably existed as well.

This study concludes with another summary of the findings concerning Latin pronunciation at St Gall, the conditions under which liquescent neumes are used, and the specific situation at St Gall. An additional fifty pages provide an appendix made up of charts which record the presence or absence of liquescence throughout the comparison. The appendix bears witness to very thorough analysis, although the reader may choose to rely mainly on the figures given in the body of the text.

Semiological studies can be intimidating for the uninitiated, but van Betteray's clear organisation and frequent itemisation makes his complex argument straightforward and accessible. The connection between the meaning of a text and its performance is, of course, central to our understanding of the liturgy in the Middle Ages. Van Betteray illustrates this connection in many interesting examples and sparks important questions about the influence of medieval theology on plainchant. Specialists in the scribal tradition at St Gall will have an especially keen interest in this study. Images used throughout the book are usually visually clear and large enough to decipher, even without additional arrows for emphasis. However, the reader might have benefited from references to specific figures and tables within the text.

The idea of liquescence as the key to the semantic meaning of a given text is fascinating, if difficult to prove objectively; yet it inspires interesting questions. Could the proposed exegetical function of liquescence possibly be restricted to the St Gall tradition, or will future research find that other scribal centres used liquescent neumes in this way as well? Could liquescence have been used didactically as well as mnemonically, that is, to point out the theological significance of a word to singers who may not yet understand it? And does the presence of necessary phonetic conditions for liquescents influence the scope of the kind of exegesis van Betteray proposes? This study has provided a solid basis for further investigation.