172A. Cognitive Psychology of Music (Introduction) Undergraduate, non-major course Weeks 9 & 10

Summary of lectures

 

i s s u e s   c o v e r e d
Meaning and Emotion in Music	Music & Film

The notes will not include description and analysis of the film examples discussed in class.

Can meaning and emotion in music be measured?

   Kate Hevner (1930's) was the first to operationalize meaning in terms of descriptive adjectives. In a series of rigorous studies using check-list descriptions, subjects listened to pieces of music checking the appropriate words. Although there was no isomorphic relationship between musical variables and adjectives, Hevner's work revealed some basic relationships between musical variables and groups of adjectives.
    The adjective-check-list idea was extended by Osgood, Suci, &Tannenbaum (1957) in their book "The Measurement of Meaning", where they introduced a technique called semantic differential (first discussed on week 4.) According to this technique, meaning and emotion become quantifiable in terms of responses of listeners on scales (initially with 5, 7, 9, or 11 subdivisions; later with up to 100 subdivisions) outlined by pairs of polar adjectives (look at your final-project handout).

As was the case with Hevner's work, this technique has not been successful in recording fine distinctions between musical variables. R. Brown used the semantic differential technique in a study where groups of musicians and non-musicians rated 12 variations of sadness (operationally defined in terms of 12 musical pieces incorporating musical aspects that had been identified by previous studies as sad). The data indicated no consistent relationship between musical and verbal variations.

In the semantic differential method words are grouped into 3 large categories (factors) of meaning:
a) Evaluative (i.e. good-bad), b) Potency (i.e. strong-weak), c) Activity (i.e. active-passive)
Depending on the question asked, researchers usually pick a few pairs of polar adjectives from each factor.

 

Do the adjectives used in the semantic differential technique actually represent the internal states of the listeners?

    As previously discussed (week 4; music & productivity) claims that musical variables cause physiological, psychological, or emotional responses are not scientifically justified.
It is more reasonable to argue that music suggests meanings or emotions. Operationalizing meaning and emotion through words is one way to illustrate music's suggestive potential. It essentially allows us to link musical experience with other forms of experience and can be successful in examining the relationship between intended and received message/mood in music.

Lipscomb & Kendall (1995) conducted an experimental study using the semantic differential technique in an examination of the relationship between music and image in films that explored this possibility. Five scenes out of the 1993 film Star Trek IV (music by Leonard Rosenman) were presented to subjects
a) with their original score, b) with the score from the other scenes, c) without image, & d) without music. Starting out with the larger theory that the meaning of a scene can change with changes in the music the study tested the specific hypothesis that listeners/viewers will be able to identify as most fitting the music that was intended for each specific scene. According to the results, the majority of the subjects matched the intentions of the composer, although alternative choices were also possible. This last observation, combined with the fact that the most likely alternative choices had accent structures closely related to the accent structures of the images, suggested that:
In communicating musical messages the syntactical component of music (accent structures, contours) may play an important role but it is not the only musical communication cue.

 

What do we mean when we say that music suggests (signifies) meaning or emotion? How can music suggest anything?

Music: Temporally organized sound and silence, a-referentially communicative within a context.
The term 'a-referential' in the above definition is one of the main aspects of music that distinguish it from language. Music can, and often is a-referential (that is, without a referent outside itself). But, since the meaning of language is its referent (: an idea, object, or event that the sounds of language point to), if music is a-referential, is it meaningless? What do we mean when we say that music has "embodied meaning" that arises from within?

Back to our definition, we have said that music is temporally organized sound and silence. According to Leonard Meyer (1956, 'Emotion and Meaning in Music'), the term 'embodied meaning' refers to this organization. In other words, while the meaning of language is its referent (semantic meaning), the meaning of music is its temporal organization (syntactical meaning); the way its patterning unfolds with time.
(A similar point is made by other disciplines arguing that the referent of music is the human experience of time.)
This is reflected in the fact that, while we describe a sentence by outlining its subject, we often describe a musical phrase by outlining its organization. When listening to music, meaning arises out of our efforts to organize a series of sound events into patterns that effect data reduction and reduce uncertainty.
[This is not to say that there can be no other types of meaning in music. What we are concerned with here, however, is explaining the potential of music to be meaningful without necessarily having a specific referent outside itself].

 

What is the role of the above points in explaining emotion in music?

    As we have seen, the isomorphic mapping of musical variables to physiological responses and emotional responses (Cartesian approach) favored by early behaviorists does not hold. A more Humean approach is presented below.
    J. Dewey's 'conflict theory of emotion' (1894), was applied to music by Meyer (1956) who suggested that affect (emotion felt) is aroused when an expectation (a tendency to respond) activated by musical organization is temporarily inhibited or permanently blocked. Different levels of expectation and inhibition are experienced in the form of ambiguity, suspense of surprise, leading eventually to some kind of resolution. Music is seen as a game of expectations, a playful manipulation of the inherent human tendencies to identify patterns, exercise control and achieve certainty, with its meaning and value depending on the success of this very game.
The term 'conflict' refers here to patterns of tension and release. Alterations between uncertainty (high cognitive load) and redundancy (low cognitive load) that build expectations and make 'games' with them possible.

What are the expectations that musical organization activates?
We may distinguish 3 types of expectations:
a) Expectations based on cultural or other associations that are often arbitrary in nature.
b) Expectations based on explicit historical and formal musical knowledge: rhythmic and melodic contours as general stylistic features (typical rhythmic patterns or melodic lines) or as specific features established during the unfolding of a piece of music in time.
c) Expectations based on implicit knowledge of the structural organization of music in time (i.e. gestalt principles of organization). Implicit rules allow for prediction based on the fact that, in an organized system, not all events are equally likely.
    Dowling & Harwood (1986) refer to these expectations when they say that:
Perceptual learning throughout a listener's life, as well as during listening to a specific piece of music, leads to the formation of structural schemata (: a series of implicit and explicit rules) that embody expectancies which novel music violates. Those violations trigger autonomic nervous system arousal that activates further cognitive activity in a search for meaning. The meaning, when found, merges with the arousal in an experienced emotion.

According to Leonard Meyer we can have 3 types of musical meaning (related to the 3 types of expectation outlined above).
    a) Referentialism: Term referring to arbitrary association between musical features and meaning/emotion, established through temporal contiguity and/or reinforcement. (i.e. any national anthem and the relevant nationality; Beatles' "Good day sunshine" and cornflakes; Davis' notion of D.T.P.O.T.; major/minor tonalities and happy/sad connotations.)
    b) Formalism: Term referring to explicit knowledge of musical structure and context (i.e. history.) Many music theorists would argue that it is here that music's 'true' meaning lies.
    c) Expressionism: Term referring to music's embodied meaning: music's temporal organization of sound and silence. Meaning arising from patterns of tension and release within the music's unfolding in time.
Meyer's taxonomy of musical meaning owes to a model by Charles Peirce (1940's) that defined 'meaning' semiotically, in terms of signs.

 

The model introduced in this class by Roger Kendall borrows the concepts of 'referentialism' & 'expressionism' from Meyer and links them to Peirce's concept of 'icon'.  These concepts are understood not as discrete categories but as flexible 'stages' on a continuum of referentiality that has referentialism at one end and expressionism at the other. Music then, in its game with expectations, moves fluidly along this continuum. The three main 'stages' in this continuum are:

    a) Index: Related to referentialism. It denotes an arbitrary association between signifier and signified (between a symbol and its referent). [i.e. lexical units are largely indexes. There is no relationship between the word 'cat' (signifier) and the animal it points to (signified) other than arbitrary association / convention.] Index therefore denotes an arbitrary association between the musical message and an extramusical meaning. Examples include national anthems, the idea of leitmotif or idee fixe in the manner of Berlioz or Wagner, etc. Many examples of musical indexes can be found in film music. The key aspect of indexes is that the connection between the music and the external visual concept or idea is entirely arbitrary, a matter of learned association.
    b) Icon: (very important in the relationship between music & image in film) Partly referential. It denotes a connection across different frames of reference suggested by common patterns or forms (i.e. as in naming a tree 'weeping willow'.) In the case of music, sound patterns/forms (in terms of pitch, dynamics, tempo, timbre, etc.) can suggest connections to iconic features in other frames of reference. Unlike index, the connection is not arbitrary but is based on some form of resemblance in terms of form, shape, pattern, motion, transferred from one domain to another.
Iconic features can be related to S. Langer's physiognomic features. According to S. Langer's theory of emotion, music has physiognomic emotional qualities. In other words, music's patterns of tension and release (: signifier) suggest emotional patterns of tension and release (: signified).
Recognition of iconic features can be based on both explicit and implicit rules with the following 3 'iconic archetypes/prototypes' being part of implicit knowledge (at least in Western European culture): a) rising pattern, b) falling pattern, c) arch.
    c) Syntax (Purely a-referential) Related to expressionism. Denotes meaning that arises out of our efforts to temporally organize a series of events into a coherent whole by identifying boundaries, categories, structures. Syntactical meanings precede indexical and iconic ones. Before any arbitrary or formal association becomes possible, we must perceptually identify boundaries and uncover/impose some sort of organization to any incoming information. In audio/visual composites, the interaction between musical/visual accent structures (as found in some animation, dance sequences etc., where we essentially have to deal with patterns of contrast in the visual and musical domains), is the best expression of syntactical meaning. The accents do not suggest forms, shapes or motion as in icon, nor are arbitrary associations as in index. Syntactical meanings arise from pure juxtapositions of time-ordered elements.   
In this case the patterns of tension and release that constitute music's embodied meaning are ultimately determined by the listener.
Tension, release, expectation, all are closely related to what we know. It is this relationship between the music and the listener that allows this 'embodied' meaning to become meaningful to us.

  

Summary of the main points presented in the class

To summarize, the class has approached music as follows:
The physical properties of sound are important but constitute only one of the aspects involved in the making, communication, and appreciation of music. Higher level cognitive operations interpret physical stimuli based on physiology, formal training, and more importantly, experience, giving rise to schemata. Cognitive schemata have been presented as being at the basis of the interaction between implicit and explicit rules/knowledge, an interaction implied and assumed by the Gestalt theory as applied to music. Gestalt laws were approached as principles/rules for prediction rather than laws of behavior. Information theory was introduced as one link between predictability / uncertainty (redundancy / randomness; periodicity / chaos) and preference or interest. Periodicity - predictability gives rise to expectations. Works of art play with those expectations by temporarily inhibiting them, blocking them, or, more importantly, gradually changing them, with this inhibition/blocking/change being at the very root of our intellectual and affective response to art's embodied meaning (i.e. Dewey's conflict theory of emotion as applied to music by Meyer.) Music, with its powerful a-referential potential, occupies a special place in art's games with expectations. It can shape, rather than simply mirror expectations and, in terms of its embodied meaning (temporal organization of sound and silence - patterns of tension and release), the expectations it mirrors/shapes are related to our experience of time and our affective state or mood.

 

Those with time to kill can click here for something (not so) different.