Week 1

172A. Cognitive Psychology of Music (Introduction)
Undergraduate, non-major course
Week 1

Summary of lectures

i s s u e s c o v e r e d

Philosophy of science.
Cognitive Psychology of Music and its area of study.

Cartesian thought versus Humean thought.

A working definition of music.
Musical examples.

Philosophy of science
Music as a product of human behavior

According to Galileo the world can be divided into two areas:

   i) The first area refers to 'Res Cogitans' (the thinking things), and its study deals with issues of perception, thinking, understanding, learning, etc.
   ii) The second area refers to 'Res Extensu' (the external things), and its study deals with the external world; the material world that is 'outside' ourselves.
   In terms of Galileo's world view, this class is going to be dealing with 'Res Cogitans'.

Basic Questions:

   i) How do we come to perceive music?
   ii) How does music become meaningful?
   iii) What is the relationship between our perception of music and music's meaningfulness? (i.e. between music perception and music cognition)

Cognitive approach to these questions:

    Cognitive (from cogitans) psychology of music is a behavioral epistemology (i.e. a science of knowing based on human behavior) and utilizes the approaches of empirical science, even for areas such as aesthetics. For example, experimental aesthetics address feeling as the outcome of cognitive processes). The class will not deal with clinical or therapeutic psychology approaches to music.
We will examine:
    i) What aspects of music are common with other areas of knowing, and what aspects are unique to it?
    ii) What are, if any, the fundamental musical properties that have universal, cross-cultural significance?
    iii) What is the relationship of such properties to culture-specific musical symbols?

Basic Theses:

    i) Musical activity is the inevitable result of humanness and not arbitrary or a random 'extra'. It grows out of fundamental cognitive processes, as does language, and they are both to a large extend essential to survival.
    ii) There are aspects of human experience and understanding that are best articulated through music.

Areas of inquiry:

    i) Philosophy. How do we know what we know? - Can behavior be objectified and studied scientifically? - What are the problems and limits of trying to understand a system which we are part of? - Can the 'mind' ever fully understand itself without 'escaping' itself so to speak? - Does science escape those problems? (Reflective frame of reference.) In essence, all thought on 'reality' is reflective thought. The concept of science is an invention that allow us to model reality.

    ii) Psycho-Acoustics (communicating music via vibration - vibrational frame of reference.)

    iii) Psycho-Physiology (human body reaction to sound vibrations - biological/physiological frame of reference.)

    iv) Musicology/music theory (notational frame of reference: in many respects a special case of semiotics.)

    v) Semiotics: examination of the relationship between symbols -notational and otherwise- and meaning in music (symbolic frame of reference.)

    iv) Psychology per se: linking some or all of the above frames of reference with the perceptual frame of reference (some sort of observation) in an attempt to answer musical questions.

*Frame of reference: A collection of concepts and definitions accepted by a particular field or bounded area of discourse. It represents a particular world view.

Psychology (of Music):
Process of Empirical Investigation

*(Experimental) Psychology: Science of observable behavior.

The key term in the above definition is "observable". In the empirical investigation of music we deal with things we can observe, i.e. behavioral changes.
As a science it involves the following process of empirical investigation.

*From Kendall, R. A. "Empirical Approaches to Musical Meaning." In R. A. Kendall and R. W. H. Savage, Perspectives in Systematic Musicology, Selected Reports in Ethnomusicology, Vol. 12, 2005.

Example:

    Real world of music: In our culture, major tonalities are generally associated with happiness and minor tonalities with sadness.

    Question/Problem: Do children recognize the difference between major and minor before they have learned to use language? (An actual such experiment was conducted by Crowder at Yale University.)

    Scholarly literature: A review of existing studies relevant to the above question.
i.e. One of the first relevant studies was conducted by Kate Hevner. A research design would be preceded by serious consideration of this and other related studies.

    Problem analysis: This stage involves the selection of variables and their operational definition.

        *Variables: Constructs or concepts, invented by us, that can take different values and allow the breaking down of a question into measurable related parameters.
i.e. major/minor tonalities - representations of happy/sad states.
        There are two types of variables depending on their relationship to the investigator:
        *Independent Variable(s): Variable(s) manipulated by the experimenter. These are usually musical or acoustical concepts.
In our example the independent variable is music's tonality with two levels: major and minor.
       *Dependent Variable(s): Variable(s) that represent the responses of subjects (that 'measure' behavior), presumed to change as a result of the manipulation of the independent variable(s).
In our example the dependent variable is the state communicated to the subjects by the music, and also has two levels: happy and sad.

        *Operational definitions: Definitions based on explicit procedures / rules that approach a variable in terms of its criterial attributes (*criterial attributes: attributes that are both necessary and sufficient to define a concept or category.) According to Jerome Brunner (1960's), concepts or categories arise from the activity of identifying criterial attributes. Such definitions are not fixed and do not have any claim to absolute truth. They represent interpretive decisions made by an experimenter that allow him to control the independent variables and measure the dependent variables. These decisions are very important since they are partly responsible for the feasibility and relevance of a study.
In our example:
i) the independent variable (major versus minor) may be defined in terms of chord triads (to be discussed in more detail later) representing the said tonalities.
ii) the dependent variable (happy versus sad state) may be defined in terms of some representation (other than written words, since the children are not

supposed to be language-competent yet) that children can reliably recognize as communicating those states. One way to do this is through face drawings such as those on the right


    Problem Statement: Based on the selected variables and their operational definitions, the initial question/problem can be restated in the way it is going to be addressed in the specific experimental design.
In our example: Do young children (with no language competence) point consistently to different face figures when they respond to major versus minor chord triads?
Notice that the above question does not refer to major/minor, happy/sad in general but to their operational definitions.

    *Theory: A set of ordered, non-contradictory statements of relationships among variables, directed towards explanation.
A theory is not a guess. It is strongly connected to the idea of a model and signifies a broad collection of relationships among variables that make testable predictions. It is a constructed model that links our observations of the world in general (or of human behavior in particular for the case of cognitive psychology) into a network of interactions / relationships.
In our example: Children are able at an early age to discriminate between musical pieces based on major and minor tonalities as well as generally associate major tonality with happiness and minor tonality with sadness.

    *Hypothesis: A statement of expected relations between variables, as they have been operationally defined within the context of a specific experimental design.
In our example: Pre-school children will consistently point to the happy face when they hear a major chord-triad, and to the sad face when they hear a minor chord-triad.
Every hypothesis is a specific statement supported by a broader theory.

    Data collection/analysis: Methods of collecting subject responses and manipulating them numerically. To do that, the operationally defined variables must be in some way measured.
    *Measurement: Assigning by rule a value to a variable.

    Evaluation of results: Interpretation of the data-analysis is based on explicit rules and will lead to the confirmation or disconfirmation of the hypothesis. The power and efficacy of a scientific investigation depends strongly on the experimenter's healthy skepticism (i.e. creating experimental situations that will challenge his/her hypothesis; questioning results that appear to confirm the hypothesis/theory) and his/her willingness to accept the disconfirmation of the hypothesis. Both, confirmation and disconfirmation, contribute equally to knowledge.

    Reformulation of hypothesis/theory: If the hypothesis is confirmed, another facet of the supporting theory is taken up for examination. If the hypothesis is disconfirmed, it is reformulated (sometimes along with the theory that supported it) and reexamined. Either way the cycle of the process of empirical investigation is continued.

Possible Dangers
Reification of concepts/theory: The belief that our invented variables (concepts) and the modeled relationships between them (theory) are the world. The belief that the world is as fixed, as simple, and as singular as a theory.
Since, however, we can approach, understand, and experience the world only in terms of models/theories, realizing that models and theories are not the world does not mean abandoning them. It rather means understanding them as tools that help us:
    a) organize our experience into meaningful worlds
    b) translate uncertainty into specific questions, relevant to those worlds
    c) reduce uncertainty by offering answers to those questions.

Two philosophical approaches to science

Cartesian thought/Cartesianism

_Named after Rene Descartes. His idea that observations/reality can be replaced by symbols and examined through symbol manipulation represents an important intellectual leap.

_Science is a topological map of reality, and the process of empirical investigation reflects the world as it really is. Reality is therefore one, fixed, and knowable.
Its map can be constructed through observation and expressed by symbol sets.
Symbols are defining aspects of objectivity and frames of reference expressed by symbol sets (mathematics, music theory) are privileged. (In the Middle Ages music was studied as part of mathematics.)
There is an isomorphic relationship between theory and reality (truth-fact).
Approach related to Galileo's division of the world.

_*Isomorphism: A state where a one-to-one relationship exists between variables in different sets.

_Machine: A system of relationships that is governed by laws.

_ The world is objective and our perception subjective.
(If this is the case, is it possible to have a science that deals with res cogitans rather than res extensu?)

_ The world is a machine obeying physical laws (that are external to the observer.) We are part of the world machine so we also must obey laws.
Behaviorism: psychological approach based on behavioral laws (on the idea of humans as machines). i.e. : Pavlov's study of conditioned reflexes (1927).

_ According to cartesianism, as scientists, we discover the laws of nature and of behavior.

Humean thought

_Named after David Hume, Scottish philosopher of the 18th century. His ideas represent a radical paradigm shift, often referred to as 'Hume's wrecking ball.'

_Reality (our understanding of it) is the product of the observer.

_No dichotomy is recognized between objective and subjective because the observer is always part of the system being observed.

_Science is not a topological map of reality. Rather, it invents reality.
Laws of science are inventions that model reality, describing in a non-contradictory way predictable relations between variables. Therefore, there can be more than one possible realities.
(If this is the case, how can we distinguish between good/bad, right/wrong inventions?
What are the evaluation criteria?)
Facts and truths are not related isomorphically. A fact/truth is not something fixed but a reliable event; something that is repeatable.

_Science, therefore, does not deal with fixed truths but with facts understood probabilistically.

_Question of Reliability: Does the science prove what it purports to relative to a given context?

_Question of Validity: What is the relevance of the reliable world, created/constructed for scientific investigation, to the world of our experience (musical or otherwise)?

i.e. Question:
If a tree falls in the woods with no one around, is there a sound?

a) Cartesian Answer:
Yes. The physical world defines reality.
No distinction is made between the physical and perceptual frames of reference. Sound is vibration (or there is, at least, a lawful connection between vibration and sound.)

b) Humean Answer:
No. Perception defines reality. Sound is a concept belonging to the perceptual frame of reference, distinct from the concept of vibration. In order to say that there is a sound, vibration alone is not sufficient. An observer is also required.

*VALIDITY: Relation of a study to a real world problem.
Are the variables and the relationships hypothesized between them connected to a "real world" problem (i.e. a problem that we are likely to encounter and which we find significant)?
In our earlier example, for instance, how valid is the representation of major/minor with chord triads, and of happiness/sadness with face drawings? How close is such a context to a possible musical experience? Will answering this question tell me something useful regarding the relationship between the affective potential of a piece of music and its overall tonality?

*RELIABILITY: Repeatability/predictability of observed relationships.
How repeatable, predictable, and therefore factual (in the Humean sense) are the relationships demonstrated by a study?
In our earlier example, if we find that the hypothesis is confirmed (children point to the happy face for major chords and to the sad face for minor chords), how certain are we that the changes in response were a result of changes in the chords? Are there parameters, other than chords and drawings, influencing the responses? If everything other than those two variables is changed, are we still going to get the same results?

Roger Kendall's Axiom
All research is a struggle between validity and reliability. As validity increases (i.e. as experimental contexts approach music-experience contexts), the complexity of the model increases and the reliability (i.e. degree of predictiveness/replicability of results, and therefore their claim to demonstrate scientific facts) decreases. Conversely, as the complexity of the model decreases, reliability increases but validity decreases.

(Psychology of) Music

*Music is temporally organized sound and silence, a-referentially communicative within a context.

Emphasis is being put on the temporal and communicative aspects of music. We are interested in music as it relates to listeners and on whether and how manipulating music affects "knowing". Music is therefore seen as a link between performer and listener. In listening to music there is always some sort of response, some kind of behavioral change indicating that we "received" something. What is received is an intention communicated in terms of patterning/configuring/organizing (the 'prosaic' aspect of music is also very important in communicating musical expression.)
The importance of context is a direct consequence of music as communication. In order for any communication to take place some sort of shared knowing is required and this is what is going to be broadly referred to here as context.
The potentially a-referential/self-referential nature of music is what distinguishes it from other forms of communication (i.e. language). As an analogy music may be seen as a "noun-less" language, made just from verbs (: potential, motion, action, narrative time.) Reference to the temporal aspects of music (music as articulating human experience of time) attests to music's significance (Suzan Langer: Music as virtual time.)

Musical examples:

a) Music from the film Koyaanisqatsi (1983, directed by Godfrey Reggio). Music by Philip Glass. Demonstrates the configuration (structuring) of time through the repetitious musical patterning. It also uses the contrast of major vs.
minor to signal good (nature) vs. bad (industry).

b & c) Music from the film 2001: A Space Odyssey (1968, directed by Stanley Kubrick). Monolith scene. Music by G. Ligeti, "Requiem." Music from the film Forbidden Planet (1956, in stereo, directed by Fred M.Wilcox). "Electronic Tonalities" by Louis & Bebe Baron. (Very unconventional score. Intent: "otherworldliness.") These are examples where unexpected elements such as unusual instruments and lack of periodic (repeating) accents through time demonstrate that music needs neither melody nor rhythm as conventionally conceived to be accepted and perceived as musical.

d) Music from the trailer 75 Years of MGM. This demonstrates congruence of musical accents and visual accents (more on this to follow in later weeks).

e) Music from the film Psycho (1960, directed by Alfred Hitchcock.) Original Music by Bernard Herrmann. This demonstrates how something non-melodic, brief, and powerfully suggestive can enter the culture's iconic lexicon for associations such as pain, danger, and fear.