Abstract
Recordings have gained increased attention as research sources in academic research for the
past two decades, prompted by both the growing presence of popular music in university
curricula and the renewed interest in early (pre-1945) recordings as sources for historical
performance practice. Indeed, historical recordings have been in use as exemplars of past
performance styles since 1980s, and they remain in the centre of diverse topic span, including
historical performance practices in context of violin playing. Various insights in the late
nineteenth- and early twentieth-centuries, in terms of traditions, musical approaches and
different performance-based aspects of early recordings in context of violin performance
have all been well documented by Brown (1999), Fabian (2003), Leech Wilkinson (2006),
Katz (2006), Gebauer (2017), Milsom (2003, 2020), Williams (2019). The methodologies used
in this body of work span from contextual, critical, cultural approaches, computational
methodologies, and practice-led research. The latest explorations of violin playing styles by
David Milsom (2023) presented early sound recordings as research tools in historical
performance practice. Milsom in his previous work combined early recordings with range of
other sources such as treatises and editions to research romantic performance practices, while
his 2023 chapter explores the issues of emulation and expressive playing during mechanical
recording sessions. Milsom’s focal point are Joseph Joachim’s recordings, and methodological
paths he developed in order to further understand how Joachim played. A different
methodological avenue was taken by Vollmer and Bolles in 2024, who digitally simulated
violin recordings basing them on Fritz Kreisler’s recording of Liebesleid (1911), in search of a
way to modify modern recordings to sound like historical ones. The current state of research is
indeed promising, with authors approaching the same historical sources though different
avenues, bringing fresh and compelling results.
The present chapter seeks to explore the primary challenge in studies of historical violin
playing styles captured on early acoustic recordings: what do these recordings actually capture?
The chapter begins by presenting a case study conducted at the University of Surrey in October
2023. Using the historically-informed recording method developed by Stanović and Stanović
(2021), this case study expanded the investigation into the relationship between performer and
mechanical technologies, and the differences between what was played, recorded, and heard
by the audience. By placing violinists in carefully controlled acoustic conditions, the
researchers were able to observe the players' reactions, draw conclusions about the mechanical
recording process, and identify aspects of early recordings that can be reliably studied as well
as those that remain speculative. Through analytical comparison of the recordings made in this
study, as well as audience questionnaires and discussions, the case study sheds light on the use
of mechanical recording technologies and their impact on performers. The chapter notes that
early sound recordings offer abundant evidence of performance styles from the turn of the 20th
century, including sparing use of vibrato, prominent portamento, tempo rubato, and other
stylistic elements. However, this case study represents just a first step towards understanding
the historical recording technologies and their influence on violinists. The results bring us
closer to comprehending how we interpret historical recordings and the capabilities and
limitations of the recording machines themselves. Due to the scope of the study, the chapter
focuses on wax cylinder recordings, leaving discussion of acoustic discs for another time.
Given the small sample size of participating violinists, this study should be considered a pilot
investigation, underscoring the need for further research in this area.