Playing in the band

Distributed cognition in musical production

By Jeremy Kashnow

School of Information Management & System

University of California, Berkeley

May 15^th, 2003

Introduction

Playing improvisational music involves distributing cognitive activities among bandmates and tools. Traditional cognitive models view cognition as an activity that occurs only within a single human brain. Some tasks require a broader view of cognition to more accurately understand the intricate flow of information between people and artifacts. The playing of improvisational music with other musicians requires each person to work both independently and interactively to achieve a common goal. The distributed cognition framework developed by Hutchins enables this shift in analysis which will make visible forms of communication that have traditionally been overlooked, misunderstood, or unknown.

What is distributed cognition?

Distributed cognition is an analytical framework developed by Hutchins “that takes as a unit of analysis a culturally constituted functional group rather than an individual mind.”[i] Traditionally, this framework has been applied to work settings in an effort to understand information flow between people and artifacts. Limited work has been done using this framework to focus on information flow between actors in artistic endeavors.

Distributed cognition is different from traditional cognitive models in two important ways:

It defines the boundaries of units of cognition in a broader sense. Rather than focusing on cognition within an individual alone, distributed cognition looks at multiple individuals and the artifacts they interact with. Furthermore, it states “a cognitive process is delimited by the functional relationships among the elements that participate in it, rather than by the spatial collocation of the elements.” [ii]

Rather than limiting cognitive analysis to representations within the brain, Distributed Cognition looks at external resources, as well as internal cognitive elements.

Using this framework, certain “kinds of distributed cognitive properties become apparent:

Cognitive processes may be distributed across the members of a social group.
Cognitive processes may involve coordination between internal and external (material or environmental) structure.
Processes may be distributed through time in such a way that the products of earlier events can transform the nature of later events.”[iii]

The collective improvisational nature of certain musical genres and the musicians’ reliance on external cognitive tools makes distributed cognition an ideal framework to help make visible the seemingly unnoticed or scantily understood mechanisms of communication between cognitive groups producing (and listening to) music.

Genres of music

Rather than analyzing musical production in general, we will use this framework to analyze a specific musical genre known as “jamming.” The unique properties of this style, notably the deep commitment to the application of improvisation towards more traditional rock, funk, and blues styles, makes it an appropriate candidate for analysis using distributed cognition. In order to study this phenomenon, we will examine how members of a “jamband” interact with each other and their environment to produce a group-developed work of art. Due to the nature of this type of interaction, focus will be applied towards the production of live musical events and not overdubbed recordings.

What is a “jamband?”

The jamband classification is nebulous. As the word becomes trendy, more groups call themselves jambands. As interest wanes, or the space becomes overcrowded, people choose new terms to describe themselves. Semantics aside, the jamband moniker could be applied to any group that consistently takes extended improvisational risks while creating music. Those that do it well have yielded incredible results… sometimes. The willingness to take chances means that success will not always yield. Sometimes the results are downright abhorrent. But when all of the pieces line up together, both musicians and fans feel as if they have witnessed something mystical. What they are witnessing is distributed cognition at its finest. A “group mind” emerges from the broth of musicians, tools, fans, and setting. A jamband is a group of musicians willing to take the risks often enough to achieve this blissful state.

In may now be appropriate to define the term “jam.” “To ‘jam’ is ‘to play collectively improvised music,’ according to jazz clarinetist Mezz Mezzrow in Really the Blues, published in 1946. The word became popular among ‘hot jazz’ players of Louis Armstrong’s generation of the 1930s, who called a group of musicians playing together – especially at a party – a ‘jam session.’”[iv].

John Bell of Widespread Panic describes jamming: - “’Jamming isn't just everybody mindlessly going 'blah.' It's action and reaction and a musical conversation where you're listening to your fellow musicians and they're listening to you,’ Bell told AP Radio. ‘That's the only way it can work, or else you'll just interrupt each other.’ According to him, communication is what separates a good jam band from a bad one. ‘The beauty of it comes when the spotlight is on the whole (band) — not just one person ... and a bunch of other guys waiting their turn.’”[v]

Can this framework be applied to artistic endeavors?

Before we begin our analysis, we must first decide whether or not this framework can be applied to artistic endeavors and if so, discuss what benefits it has over traditional analytic methods.

Levels of cognitive abstraction

A significant attribute of distributed cognition is its ability to analyze cognitive units at different levels of abstraction. This property is particularly appropriate for analyzing cognition within an improvisational group and among the people and artifacts they communicate with.

The measurement of success or failure at these levels may be less appropriate. Hutchins & Klausen use the work done in an airline cockpit as one level of abstraction. They contend “the question of interest to you as a passenger should not be whether a particular pilot is performing well, but whether or not the system that is comprised of the pilots and the technology of the cockpit environment is performing well. It is the performance of that system, not the skills of any individual pilot, that determines whether you live or die.”[vi]

The performance of a musical group is gauged much more subjectively (and less dangerously) than the performance of flying a plane. What may interest someone attending a concert may not be how well the bass player performs, but how well the ensemble works together, while another fan may appreciate the intricate guitar work exhibited in spite of a bass player who can’t keep time.

In another sense, success hinges on how well certain necessary units work together. The question of whether or not a fan is enjoying the music may not be affected by how well the bass player performs if the fan can’t hear the music. The communication between band and audience using the public address systems must not fail. The implication is that there exists a minimum set of units that must perform well together in order for the system to work at all.

Interaction with media

Musicians’ reliance on technical artifacts plays an important role in the cognitive properties of the system. Hutchins contends “the properties of this computational system area as much determined by the nature of the representational media and the pattern of interconnection among representations as they are by the cognitive properties of the individual actors.” (Hutchins, 1992, p.2.)[vii] As we will discuss, typical jambands make use of developing technology to enhance their abilities to perform as a group and further their improvisational interaction.

How do we apply this to music?

The analysis of cognition among a group of improvisational musicians requires a multifaceted, or multi-model approach. No single framework will reveal all forms of communication or cognition. The distributed cognition framework can be used as a first approach to segmenting the cognitive work done by a group into a set of functional relationships. These delineations can then be used as a vehicle to apply other analytic tools to distinct cognitive groups.

Defining functional relationships

Defining the unit of cognitive measurement in multiple ways will allow us to examine the unique informational interactions between different social groups and artifacts. Conforming to the distributed cognition model, we will be focusing on functional relationships, rather than spatial collocation.[viii] This is a fundamental principle of distributed cognition and marks a significant departure from previous cognitive models. The functional relationships we will examine surrounding musical production are as follows:

Learning to play a piece of music
Playing music
Performing a concert

Learning to play a piece of music

Learning to play a piece of music involves first learning how to play a particular style of music, and then learning how to play the specific notes, keys, rhythm, and timing of a certain piece of music.

Learning to play a particular style of music

Lave and Wenger describe the process of becoming a member of a community of practice as legitimate peripheral participation.

“Learning viewed as situated activity has its central defining characteristic a process that we call legitimate peripheral participation. By this we mean to draw attention to the point that learners inevitably participate in communities of practitioners and that the mastery of knowledge and skill requires newcomers to move toward full participation in the sociocultural practices of a community. ‘Legitimate peripheral participation’ provides a way to speak about the relations between newcomers and old-timers, and about activities, identities, artifacts, and communities of knowledge and practice. It concerns the process by which newcomers become part of a community of practice. A person’s intentions to learn are engaged and the meaning of learning is configured through the process of becoming a full participant in sociocultural practice. This social process includes, indeed it subsumes, the learning of knowledgeable skills.”[ix] – Lave & Wenger – situated learning, p.29

This implies that what may be more important than “what one knows,” is “who one knows” and “how one knows what one knows.” These factors are evident in the community of “jamband” musicians.

Apprenticeship is common among this community. Certain core groups have formed splinter groups that propagate knowledge. Oldtimers distribute themselves among newcomers. The act of playing with other musicians is as much a form of teaching and learning as it is conversing and performing. The cross-pollination of group members is extremely common. This is made convenient by the aggregation of large numbers of groups at music festivals. Music festivals are held year long across the United States and are always punctuated by guest appearances of members of one group performing with others. One of the largest festivals is the annual New Orleans Jazz and Heritage Festival, held for two weeks in New Orleans, LA. By celebrating the heritage and cultural history of New Orleans, the festival fuels the fusion of styles by creating a forum for members of different ensembles to perform together.

One “tenet of the theory of distributed cognition is that the study of cognition is not separable from the study of culture, because agents live in complex cultural environments. This means, on the one hand, that culture emerges out of the activity of human agents in their historical contexts, as mental, material and social structures interact, and on the other hand, that culture in the form of a history of material artifacts and social practices, shape cognitive processes, particularly cognitive processes that are distributed over agents, artifacts, and environments.”[x]

In the same way that culture and cognition are intertwined in a symbiotic relationship, so are the acts of honoring traditional styles and developing new ones, as well as learning and participating. The act of playing music is the main form of teaching in this community. As newcomers and oldtimers spend time “jamming,” the newcomers learn style from the oldtimers and the oldtimers learn new interpretations of the music they already know. Whenever a new member joins a group, they alter both themselves and the group. This is true at all levels of abstraction: within the musical community in general, within certain genres of music, and especially within musical groups. Musicians’ desires to constantly learn provide an incentive to ask members from other bands to join them for a song or two. This willingness to both teach and learn enables the community to sustain and thrive.

Communication between a composer and a musician

How does a musician know which notes to play, for how long, and when? Most musicians playing together in an improvisational setting don’t read from sheet music as they play. They have most likely learned how to play a piece and are referencing the musical framework from memory, or they possess a deep understanding of the musical style that allows them to participate. In either case, there must be some initial communication between the composer and the musician (or from musician to musician) that teaches the student what to play.

Sheet music as an immutable mobile

How is musical information conveyed from a composer to a group of musicians?

Once a musical composition has been created, that information must be transmitted to the musicians who actually play the music. The development of a standardized communication medium has allowed this information to be distributed and widely understood. The traditional communication tool for this application has been sheet music. Sheet music is a relatively simple visual representation of complex auditory information.

Sheet music has many qualities in common with what Bruno Latour defines as an “immutable mobile.”[xi] It is a form of data storage that is both stable and portable. Sheet music is optically consistent, meaning that the same musical tones and durations are recorded the same way across all compositions. This consistency allows musicians to learn to read a standard set of symbols that represent musical tones and timing. Once this language has been learned, a musician can theoretically “decode” any piece of music stored in this form.

While sheet music is a very compact and concise, portable method of data storage, it achieves this high compression rate in a “lossy” way; Information is lost in the translation. Certain technical details are present to help musicians recreate the piece, but the nuances associated with actually playing the piece are not present. This information can be conveyed using complementary tools such as audio recordings, written notes, discussion with other musicians familiar with the piece, or even direct conversations with the composer.

It is interesting to see how the nature of the representational media affects the information being transmitted. Sheet music provides a blueprint for recreating the music, but not an example. Audio recordings provide an example to follow, and show what sounds to make, but not necessarily how to make them. Accompanying text or a conversation may provide further details about the ideas and motivations behind the composition, but not the technical details of how to play it or an example of what it should sound like. Through each mode of communication, some information is added and some is lost. These tools must often complement each other for musicians to effectively learn a piece of music.

Measuring success at this level of cognitive abstraction is still highly subjective and dependent upon the expectations of both the composer and musicians. If a composer hears a rendition of his composition and is dissatisfied with it, does that constitute a breakdown in communication? In some cases interpretation is not only expected, but also encouraged. It is certainly desirable for musicians to sound like “themselves” rather than simply imitating the person who wrote (or usually plays) the music.

Sometimes information is even hidden elsewhere. The Robert Hunter/Phil Lesh composition “The Eleven” is a framework for deep improvisation. What makes the song so unique (and difficult to play) is the time signature: 11. Non-musicians have often glossed over this fact, making up their own reasons why the song is called “The Eleven.” Again, this may be part of the authors’ intention: confuse some and amuse others.

Playing music

Playing improvisational music involves learning how to listen, how to think, and how to respond. Musicians must interact with each other and their tools in order to excel at this level of distributed artistic creation.

Communication between band members

An improvisational “jam” is actually a complex conversation between musicians. Although the overall structure of the song is usually set, the specific implementation is highly subjective. Song structure is akin to an outline. This previously agreed upon framework is not used as a steadfast blueprint as much as a guideline detailing how and when members should interact with each other.

Roles

Each musician or set of musicians has a predefined role (or set of roles). Sometimes these roles shift among members and sometimes more than one individual fulfills the role. A bass player in a “jamband”

“emphasized the need to strike a balance between taking chances through adventurous improvisations and supplying the steady rhythms and fundamental harmonic structure that propelled and oriented the excursions of the other band members. Other People cautioned themselves not to ‘all try to solo at the same time.’ While such simultaneous improvisation often works splendidly, it always increases the risk of failure. To avoid this, players consciously directed themselves to provide basic rhythms and harmonies when these elements seemed absent from the performance. In essence, they tacitly created a ‘basic player’ role that shifts fluidly between musicians. When a player recognized the need for simplicity, he reined himself in momentarily. Through this action, he both communicated that the ‘basic player’ role was filled and invited his bandmates to participate in the newly created groove.”

“The balance between musical responsibility and improvisational freedom was most apparent in the playing of the bassist and drummer because their roles are most tightly defined by rock convention. Randy loved to improvise fast lead lines, yet recognized his responsibility for playing simple and direct lines that kept the music together. At times, his bandmates directed him to simplify his playing so that subtleties of others’ parts could be heard.” [xii]

Within a band, certain sections function as single cognitive elements. The rhythm section is typically comprised of the drummer(s), bass player and rhythm guitar player. Even within these groups, cognitive elements may be distributed among multiple members. Bill Kreutzman of the Grateful Dead has said that when he and drummer Mickey Hart play together, “We’re not trying to be two drummers, we’re trying to be one drummer with eight arms.”[xiii] Why does he say eight arms, and not four? He may be simply making reference to the eight extremities (arms and legs) of the two drummers, but he may be pointing out that the cognitive whole is greater than the sum of the parts. In either case, he suggests that the traditional cognitive boundaries of individuals have been transcended within this cognitive unit.

There are, of course, ethnographic factors to consider as well. In addition to the technical information about the song, musicians bring their entire history of experiences to their rendition. Their unique experiences make each rendition distinctive.

Auditory and visual cues

Band members use both auditory and visual cues to communicate with each other when they play. Because all of the musicians share a common knowledge of music, there is redundancy of information spread across members. Error correction techniques utilize this redundancy. If a drummer loses time, the bass player can help him get back to where he needs to be. When a band member violates a norm, this is sure to be recognized by his bandmates. Often a simple glance at the right time can convey a wealth of information.

When one bass player decided to indulge himself in a solo at an inappropriate time, he recalls, “Such attempts often evoked glares from bandmates who felt he was stepping out of his role.”[xiv] Visual cues play an important role in group communication. The “leader” of the band may simply look at an individual to tell him it’s his turn to solo. With such a primitive mode of communication, errors abound. Simply looking at the wrong musician could have disastrous results. There must be other ways in which musicians communicate while playing.

Because the nature of communication is interactive as well as improvisational, we must realize that a “conversation” is present. Musicians can use the music itself as a communications medium.

The role of the drummer and bass player as timekeepers allow them to use the music they are playing to help bandmates understand “where they are” in a given piece. Their roles as timekeepers provide global cues fro other musicians to follow. Flor & Maglio[xv] argue that vocal cues are another form of global cueing. Reliance on these auditory global cues actually reduce musicians’ cognitive workload. If they are listening carefully to one another, they don’t necessarily need to know exactly where they are in a piece. They can listen for the end of a vocalist’s lyric or a specific drum-fill to cue them to change chords.

Communication between band members and artifacts

Instruments

The most apparent interaction between musicians and artifacts is their relationship with the musical instruments they play. A musician typically interacts physically with an instrument to produce sound. This may take the form of opening valves on a flute to create different sound energy, or plucking the string of a guitar and amplifying the resonance. Information is propagating across media states, moving from mechanical energy to auditory information. (When recorded, this auditory information is transferred to a durable medium and stored for later propagation back to acoustic sounds.)

Digital technology affords musicians a limitless range of interaction with which to create sounds. For example, pressing a key on a synthesizer may communicate information to a microprocessor to recreate the sound of a trumpet, or a group of trumpets, rather than a piano. The development of MIDI has created a standard to allow digital artifacts to communicate.

MIDI (Musical Instrument Digital Interface)

MIDI is a standard developed to allow musical instruments to communicate with one another. The notes played on an instrument are converted to data that can be used to create new sounds. In reference to the Grateful Dead: “Adopted by the band in the late ‘80s, MIDI capability enables Garcia to trigger the sounds of a flute from the fretboard of his guitar, Welnick to play violin or percussion from his keyboard, or Lesh to halo the tones of his bass with a chorus of human voices. The introduction of MIDI to the Dead’s setup has had a more profound effect than any other development since the Warlocks went electric, adding to the bands vocabulary virtually any sound imaginable.”[xvi]

The use of MIDI supports arguments that Law makes regarding the blurring of lines between people and machines. “The dividing line between people and machines (and for that matter animals) is subject to negotiations and changes. Thus it is easily shows that machines (and animals) gain and lose attributes such as independence, intelligence and personal responsibility. And conversely, that people take on and lose the attributes of machines and animals.”[xvii]

The use of technology not only delegates agency to machines, but also allows for a new type of group cognition to emerge. Bob Bralove, MIDI expert for the Grateful Dead comments, “I see much of what I do as designing virtual instruments. If you build a guitar, you can process what it does to make it sound an infinite number of ways. There’s room enough for infinity in any instrument. But in building a virtual instrument, you create how you produce the sound as well – whether you’re going to pluck it or hit it, what any action is going to do. If you then look at that virtual instrument as a central brain that can be divided or controlled by six or seven people simultaneously, you’ve opened up a whole new conceptualization of what might happen. I’m trying to build instruments for the Grateful Dead that have built-in infinity for everybody, and at the same time, maintain everybody’s power over their own universe, because that’s the nature of this band.”[xviii]

This new type of group cognition would not be attainable without the use of technology. Furthermore, it would not be attainable without the knowledge relating to how to use the technology. The Grateful Dead required Bob Bralove, an expert in MIDI technology, to create new virtual instruments for them to play, teach them how to use these new instruments, and sometimes even manipulate the properties of the instruments in real-time, making himself a virtual member of the band.

Performing a concert

If we now examine the functional relationship among members performing a concert, we have added new cognitive groups to the level of abstraction we are analyzing. In addition to the musicians (and the composers they’ve interacted with in the past), we now have an audience and venue to interact with.

Communication between band and audience

The bidirectional nature of communication between the band and an audience can’t be emphasized enough. Traditional views of musical performances view the musicians as performers and audience as passive recipients of entertainment, rather than as active participants. “Jambands” dismiss this archaic model and embrace the active role of the audience. This philosophy stems from the lineage of this genre that had its birth in an arena that focused on the lack of distinction between performer and participant. These early experiments were called “Acid Tests” and put on by an amalgam of artists and fans made up primarily of Ken Kesey’s Merry Pranksters and The Grateful Dead.

“Admission was a dollar, even for the musicians, and ‘everyone was involved,’ explains Dick Latvala, the Dead’s tape archivist. ‘It wasn’t audience and performer – those distinctions were deliberately blurred. You are it, you are the experience you’re witnessing – that was what it was all about.’”[xix]

Charging musicians and fans alike was a tool designed to encourage fans to take on more participatory roles.

Audio recordings

The nature of communication between band and audience transcends time and place. It begins with exposure to the music of the band, usually in the form of an audio recording. Even these seemingly direct methods of communication have facets worth exploring. The medium that the music is recorded on contains valuable metadata. For example, if someone gives you a DAT, you would expect the tape to contain live music. An LP would have connotations about the timeframe the music was written in. An .mp3 file might make you consider the copyright statues that apply. Album and CD covers also portray representations about a musical experience. The relationship between band and fan begins well before a person attends a concert.

Tickets

Communication regarding a specific performance begins with the purchase of a concert ticket. Just as Korzybski points out that “the map is not the territory,” or as Watts puts it, “the menu is not the meal,”[xx] similarly, the ticket is not the show. However, important metadata about the show is communicated through this medium. Logistical information regarding time and place, as well as behavioral guidelines (no cameras) can be communicated. Sometimes the size and appearance of tickets affect expectations about the performance. The Grateful Dead, and more recently, bands like Phish or The String Cheese Incident have been known to produce elaborate oversized tickets for special events such as New Year’s Eve concerts. The grandiose nature of these tickets undoubtedly affects everyone’s preconceived notions about a performance.

Venue

The acoustic and technical properties of a venue obviously affect the communication between the band and the audience. More importantly, however, the social construction surrounding certain venues affect the way the musicians interact with each other and how the band interacts with the audience. It is the sociohistorical nature of venues that shapes people’s cognitive frames.

Take, for example, the Fillmore Auditorium in San Francisco. In existence for less than three years (1966 – 1968), the hall hosted some of the most influential artists of the time. The people who played there are still legends and their performances are legendary. That energy was present in future shows. ”There was very much a whole kind of Fillmore energy coming off the audience that combined with the band,’ recalls Eric Clapton, who played the Fillmore with his primordial power trio Cream. “I was encouraged to get outside the format. I was encouraged to experiment.” The history of that venue has a direct positive effect on both fans and musicians.

Sometimes the right mix of people, places, and times conspire to make an indelible mark on a venue. Sometimes the setting of a venue has a dramatic effect on a performance. Legendary amphitheaters like Red Rocks in Colorado inspire both fans and musicians with natural wonders. These types of distributed improvisational events are extremely sensitive to both “set” (mindset) and “setting” (location). In this case, the venue is the “setting,” which has a direct affect on one’s “set.”

Auditory and visual cues

Dancing and singing provide tremendous feedback to a band. This is the real-time feedback mechanism for musicians while they are playing. Collective singing, dancing, and screaming, are all seen as positive feedback to the band. The bands feed on this energy and push it right back into the music. A lackluster show can move from mediocrity to brilliance when even just one member of the audience begins yelling and screaming. The band processes that information, perhaps not even conscoiusly, but it affects the mood in which they play. The option for fans to participate makes these events unique. The Dead may play Red Rocks again, but not with the crowd that was there in 1987, and not with everyone in the same seats feeling the same way they were then. That set of conditions is impossible to recreate.

Negative feedback can also be present, but it usually takes a different form. You have to remember that in general, these two sociological groups have deep respect for each other. Negative feedback is held in reserve. Even when one piece of the cognitive unit of the band fails, the audience finds “appreciation” in that mishap. It’s not uncommon for a lead singer to forget his lines and hear the crowd laughing with him.

Summary

We set out to apply Hutchins’s Distributed Cognition framework towards the work done in an improvisational music group. This framework helped us view cognitive activities distributed among musicians and their artifacts. We have made “visible” some of the communication that often goes unnoticed or is not well understood. For example:

Improvisational compositions are actually complex conversations between people and tools.
Communication among a Band and an audience begins well before the performance starts.
Artifacts such as Compact Discs and concert tickets convey metadata about the performance.
A simple glance can communicate a large volume of data, given a specific context.
Sheet music contains only a fraction of the information necessary to truly recreate a composition.

The Distributed Cognition framework is an appropriate tool to analyze the work done in a collaborative artistic endeavor.