Timbre Background

Timbre refers to the perceptual quality of sounds. Qualities, as opposed to quantities, pertain to the kind of differences in sensation as opposed to the amount of differences in sensation. Timbre, therefore, has to do with the kind of sound being heard, which is often specified in terms of what it sounds like. As a perceptual attribute, timbre is something we as listeners ascribe to the world; it is a way of characterizing a sound, or more importantly, it enables listeners to differentiate between sounds. Timbre then is not an inherent component of a sound but rather, as an overall quality it is an attribute of sensation (1). The linking of timbre to sensation has been discussed by many authors throughout
the century (2).

When defining timbre, the term tone is generally used instead of sound because tones are sounds capable of evoking auditory sensations. In the context of timbre experiments, tone is further refined to mean sustained, audible sensations that are (i) longer than impulses, (ii) quasi-periodic, eliciting a sense of pitch, and (iii) produced by musical instruments or the voice. In summary, viewed as an attribute of sensation or as a quality of tone, timbre functions by enabling us to differentiate tones.

The idea that timbre functions to differentiate tones has been expressed in many different ways. Around the turn of the 20th century, the term timbre referred specifically to different types of tones (3). Later in this century, however, the definition narrowed, i.e., that timbre allows a listener to judge tones with the same pitch and loudness as dissimilar (4). Even more specific definitions state that timbre is a property that allows a listener to distinguish between steady-state complex tones having the same pitch and loudness (5). This is roughly the definition provided by the American Standards Association (ASA). The ASA defines timbre in the following manner: “Timbre is that attribute of sensation in terms of which a listener can judge that two sounds having the same loudness and pitch are dissim-ilar” (Erickson 1976, p. 4). Depending on the type of experiment being performed, some authors do not define timbre as enabling a listener to differentiate tones but rather define timbre as an attribute that enables a listener to recognize which musical instruments produced a particular tone (6). Including this recasting of timbre in terms of recognition and identification, all of these definitions are forms of the ASA definition: timbre is the way in which tones, having the same pitch, loudness and duration, differ.

Historically, timbre was not defined as a perceptual attribute enabling listeners to differentiate or recognize tones and did not obtain this sense of the word until the mid 18th century. The first recorded use of timbre appeared at the end of the 14th century: “There was flu many a tymber bete. And many a maide carolende” (OED 1971, p. 37). At one point, timbre referred to a kind of bell (i.e., a tabular bell) but also was used in the context of heraldry to refer to the crest over the shield. The later sense of the term eventually changed from referring to a crest on a coat of arms to a crest stamped on a document, and finally, to a postage stamp. The musical sense of the word evolved from meaning a bell, then the sound of a bell, then the sound quality of an instrument in general, and finally to the quality of sound in general. Later, as mentioned above, a function was ascribed to tim-bre: to differentiate between types of sound.1 An early example of how the term was used in its modern sense can be seen in the mid 19th century in C. Brontë's novel Shirley: “Your voice has another timbre than that hard, deep organ of Miss Mann’s” (OED 1971, p.37). This sentence captures the common modern conception of timbre as a quality of tone.

Pitch and loudness are perceptual qualities of tones, but defining timbre as a per-ceptual quality is not especially informative. Pitch is an attribute of auditory sensation that can be ordered on a scale from low to high, while loudness can be arranged on a scale from soft to loud. Although pitch and loudness are not necessarily uni-dimensional, they are often reduced to one dimension: correlating loudness with intensity and pitch with fundamental frequency. Timbre differs from pitch and loudness in that it is multidimensional, incapable of being reduced to one dimension.1 Timbre cannot then be correlated with any one physical dimension of a sound. This multidimensionality contributes to the definitional vagueness of the term itself.

Definitional Vagueness

Most definitions of timbre are descriptive. The definitions explain what timbre is and how it functions. With respect to what timbre is, timbre is frequently defined analytically, such as in an Aristotelian manner (defining a term via its genus and differentia). In an analytic framework, a term is defined by generalization. The term is first placed within a class, its genus, and then distinguished from other members of this class. Two terms are used to provide a general framework when defining timbre: quality and percept. For example, timbre is frequently described as a sound quality. The term quality in this context refers to how a sound comes across to a listener, i.e., the general character of the sound. Timbre can also be defined as an auditory percept. The term percept refers to sensual experiences. After establishing timbre as a member of a class (quality or percept), it is typically contrasted with pitch and loudness in an attempt to further specify its meaning. For example, one main difference between loudness and timbre is that loudness is unidimensional and timbre is multidimensional. Stating that timbre is multi-dimensional can cause confusion because, although timbre is dependent upon numerous physical correlates, each physical correlate does not necessarily translate directly into a single perceptual dimension.

The second part of defining timbre is its functionality. In most definitions, timbre is said to enable a listener to differentiate between two or more sounds. More specifically, timbre enables the listener to identify the source of the sound – the particular musical instrument making the sound. Timbre also enables the listener to classify the sound according to some type – such as a brass instrument. The identification and classification of sounds depends upon subjective constancy (8). Subjective constancy is the consistent perception of one or more aspects of a sound, i.e., an attribute of a sound that seems the same each time the sound is heard.

Hadja et al (1997) discussed several methods in which timbre is defined, but focused on two types: the operational and the constitutive. The terms, operational and constitutive are taken from Kerlinger’s Foundation of Behavioral Research. According to Kerlinger, an operational definition is a “definition that assigns meaning to a construct or variable by specifying the activities or ‘operations’ necessary to measure the construct or variable” (Kerlinger 1973). Operational definitions either state a set of procedures for measuring the variables (constructs) or outline the method of manipulating them. Bridg-man (1934) used the term length to explain operational definitions. The actual length of an object is determined only when the set of operations used to measure the length of the object is determined. Specifying the length of an object necessarily presupposes that the act of measurement has been performed upon the object. Hadja et al (1997) suggested that there are three types of operational definitions: categorical, continuous and hybrid. A cate gorical definition involves verbal labels which are assigned to particular instruments. In this case, the timbre of a particular sound is measured by having a listener identify the sound of an instrument. A continuous definition incorporates scaling methods in which a numeric value is assigned to a sound that indicates the degree to which some adjective applies to a sound. For example, if a sound receives a 1 then it is not very sharp, whereas receiving a 10 means indicates that the sound is extremely sharp. The hybrid combines the definition of the categorical and continuous.

In contrast to an operation definition, a constitutive definition specifies a set of constructs in terms of other constructs. A construct can be defined as a concept that has been designed for a specific purpose. According to Kerlinger, the constitutive definition of the term, weight, for example, is that weight is the heaviness of objects (9). Hadja et al (1997) claimed that no actual constitutive definition of timbre has been proposed. Rather than define timbre directly, a number of researchers have attempted to find individual features that are possibly subsets of timbre. As subsets of timbre, individual features can be rigor-ously defined without having to rely upon a generalized definition of timbre, i.e., defining a set of terms without making reference to their superset.

Typically, timbre studies either involve determining what features are important to particular timbres, i.e., studying one sound in isolation, or understanding category formation, i.e, the relationship between more than one sound. Categories facilitate the comparison of different types of sounds. When we as listeners describe timbre, we often specify what something sounds like. This something can either refer to an object that produced the sound – such as a trumpet – or to an adjective used to specify a dominant characteristic of the sound – such as shrill. In either case, the qualities ascribed to sounds specify kinds of differences as opposed to amounts of difference. Simply put, different kinds mean differ-ent categories.

1. Authors that define timbre as a quality or characteristic of tones include Diderot and D'Alembert (1758), Zahm
(1900), Barton (1922 ), Sir Jean James (1968), Seashore (1938), Bartholomew (1942), Licklider (1951), Olson
(1952), Groves (1952), Encyclopedia de la musique (1958), Backus (1969), Wedin and Goude (1972), Howe (1975), Plomp (1976), Wessel (1979), Groves (1980), Campbell and Greated (1987), OED (1991).

2. See Fletcher (1934), Erickson (1975), Risset and Mathews (1969), Rosch and Plomp (1982), Dowling and Harwood (1986), Pellman (1994), Handel (1995).

3. See Zahm (1900)

4. See Riemann (1913), Barton (1922), Fletcher (1934), Sir Jean James (1937), Bartholomew (1942), Wedin and Goude (1972), Roederer (1975), Slawson (1975).

5. See Helmholtz (1957), Campbell and Greated (1987), Plomp (1976).

6. See Fletcher (1934); Olson (1952); Saldhanha and Corso (1964); Risset and Mathews (1969); Howe (1975).

7. See Diderot and D'Alembert (1758); Zahm (1900).

8. See Hadja et. al. (1997)

9. See Kerlinger (1973)