Fast / Major and Slow / Minor pairings in J . S . Bach ’ s Well

Some studies have found that subjects give higher preference ratings and clearer emotional descriptions to pieces with the tempo/mode pairs fast/major and slow/minor. If these findings are true, one might expect to find a predominance of fast/major and slow/minor pairings in musical literature. This study looks for just such a trend in the canonical work of J.S. Bach’s Well Tempered Clavier. Books 1 and 2 together contain 24 Major Preludes, 24 Minor Preludes, 24 Major Fugues and 24 Minor Fugues. We use four different measurements of fast/slow tempo based both on elements within the musical notation as well as the interpretive markings of four different music editors: 1. Attacks Per Beat (APB) – measures the number of subdivisions of each piece 2. Attacks Per Minute (APM) – is the APB of a piece multiplied by the editor’s metronome marking 3. Time Signature Distribution and 4. Editor’s Tempo Distribution. We find a significant amount of fast/major and slow/minor pairings in Book 1 but not in Book 2. Submitted 2011 August 25; accepted 2012 February 2. KEYWORDS: music, minor mode, minor key, tempo, sad, J. S. Bach TEMPO/mode pairings have often been studied in the context of musical preference and happy/sad judgments. There have been studies on the role of tempo and modality in the perception of happy vs. sad music (Hevner, 1935, 1937; Desroches, 1997; Dalla Bella, 2001; Gagnon et al., 2003), the role of tempo and modality in musical preference (LeBlanc 1983 and 1988, Flowers 1988), and the effect of tempo and mode on mood and arousal (Husain et al 2002). These studies have shown that tempo is considered the most significant musical dimension when subjects are asked to determine the emotional character of a piece or to make a preferential judgment between pieces. Mode also affects a person’s musical preference and their perception of the emotional character of a piece, but the importance of mode tends to be secondary to that of tempo. Several studies (Flowers, 1998; LeBlanc, 1983, 1988) found fast tempo in particular to be a distinctive feature of preferred musical selections. This finding was bolstered by another study (Husain, 2002) which found that fast tempo also seems to increase arousal in listeners. The primary focus of these studies was to understand musical preference or emotional character. In the process, though, they also found some interesting relationships between tempo and mode. Hevner found that music in a major mode (1935) or with a fast tempo (1937) is often labeled happy and that music in a minor mode (1935) or with a slow tempo (1937) is often labeled sad. In a more recent study (Gagnon, 2003), the happy/sad distinction was strongest in music that paired fast tempo with major mode and slow tempo with minor mode. Husain (2002) found [in a small subtest within her experiment] that listeners considered fast/major tunes most enjoyable and slow/minor tunes second most enjoyable out of the four possible tempo/mode pairings (fast/major, fast/minor, slow/major, slow/minor). On the other hand, Flowers (1988) found that listeners rated the fast/major tune first and the slow/minor tune last on a preference scale. In all of these studies, whether the focus was on emotional clarity, personal preference, or attention arousal, there seems to be something special about the particular tempo/mode pairs of fast/major and slow/minor. Post and Huron (2009) looked directly at these tempo/mode pairings within Western classical music of the common practice era. They based their research on work done in speech prosody, which also Empirical Musicology Review Vol. 6 No. 3, 2011 156 shows an association between slow tempo and sad emotion. Their hypothesis was that since slow tempo is generally associated with sad emotion in speech and the minor mode is generally associated with sad emotion in Western classical music, then one would expect to find a strong association between slow tempo and minor mode within Western classical music. They did a survey of the three periods of the common practice era (Baroque, Classical, and Romantic) and found that such an association did exist but only in the Baroque and Classical periods. Our study pursues a similar hypothesis to Post and Huron’s study and is consistent with their findings. Instead of a broad survey of the literature, however, we chose to limit our sample to J.S. Bach’s Well Tempered Clavier, which consists of two books of 24 preludes and fugues. Though hardly representative of Western classical music as a whole, J.S. Bach’s Well Tempered Clavier is arguably one of the most influential musical texts within the field. Bach systematically works through all 12 keys in each book writing a Prelude and Fugue for each of the major keys and each of the minor keys. The result – a neatly arranged set of 48 minor pieces and 48 major pieces within the same musical genre – is a nice sample for a study which wishes to look for significant tempo/mode pairings. Furthermore, the Well Tempered Clavier has long been considered a critical component of any pianist’s repertoire and is one of the most ubiquitous models used in basic composition and part-writing classes. It therefore holds a venerable place in the canon of Western classical music as part of the foundation of the common practice era. The notion of tempo is a central and important element of music. When trying to define tempo, however, the concept turns out to be rather slippery. One might simply rely on the words people use to describe tempo, such as fast, slow or moderate – in various languages, such as the Italian terms, allegro, adagio, and moderato. The history of music is replete with examples where such terms are regarded as highly ambiguous, and not especially useful. For example, “a walking pace” might seem clear, except that the Italian term for this (“andante”) has been interpreted dramatically differently in different periods. It is widely thought by musicologists that andante became slower in the late 19th and early 20th centuries. The invention of the metronome by Johann Mälzel in the first half of the 19th century was widely hailed as providing a way to clearly specify the tempo. Although metronome markings may allow performers to replicate a precise tempo, the metronome markings themselves tell us little about the sense of speed or feeling of musical rapidity. It is problematic to claim that 80 beats per minute is faster than 70 beats per minute without considering such aspects as the texture, density, and harmonic rhythm of the work. Even when the beat rates are identical, the manner of beat subdivision may influence how the tempo is perceived. So a 6/8 passage may be perceived as faster than a 2/4 passage even when the beat rate is identical. Furthermore, expressive performances often employ rubato, which can cause the perceived tempo to vary widely throughout the course of a piece. (See also London, 2004, 2011) Given the complexities involved, one might simply despair and conclude that tempo is ultimately enigmatic and impossible to characterize. While recognizing the slipperiness of this theoretical concept, we might nevertheless recognize that it should be possible to estimate the tempo under some circumstances. Hence, for example, although tempo terms are ambiguous, one might still reasonably assume that a passage marked "presto" might well be perceived as faster than a passage marked "allegro," that a passage containing a metronome marking of 80 bpm might well be perceived as faster than a passage marked 70 bpm, and that a passage containing 3.8 note onsets per second will be experienced as "faster" than a passage containing 3.1 note onsets per second. For the purposes of this study, we will attempt to estimate tempo using four operational definitions: (1) The number of attacks per beat or APB, (2) The rate of attacks per minute or APM, (3) Time signature and (4) Italian tempo descriptions such as allegro and andante. While none of these can be regarded as an accurate way of characterizing tempo, we hope that the results for these different estimates might converge on similar answers, and so help us address our hypothesis concerning the relationship between mode and tempo. Empirical Musicology Review Vol. 6 No. 3, 2011