The future of music

I've seen the future and it will be...

A battle between the Cathedral and the Bazaar. Before I explain what I mean, I will be the first to admit that all this is rather pretentious. Attempting to predict the future of a complex dynamic system is, by definition, impossible. However, I believe that we can influence the direction the system evolves and I hope that what I write becomes more of a self-fulfilling prophecy rather than anything else. With that disclaimer, let's move on:

A battle between the Cathedral and the Bazaar

A couple of years ago, Eric S. Raymond wrote an article titled The Cathedral and the Bazaar. It was not about music, but about software; how the "Bazaar" model (i.e., thousands of developers spread over the Internet writing software because of inherent factors, such as love for solving problems and/or coolness) is superior to a "Cathedral" model (where a small group of developers control the development of a piece of software). Raymond likens the development model of Linux, the one computer operating system to emerge as a credible threat to Microsoft's monopoly, to a Bazaar, and the Free Software Foundation's major efforts (and other commercial developments, including Microsoft) to building Cathedrals. I think this analogy is erroneous because the Free Software Foundation is as integral a part of the bazaar as the Linux development "team". The crux of Raymond's points lies not in whether the Cathedral or the Bazaar model is the better choice, but how the Bazaar model is in accordance with the theory of complex adaptive systems, and how, this approach, in general, is a long-term model for tremendous progress in terms of innovation and creativity.

There exists a similar analogy in spirit with regards to music, but different in practice because the "goal" of music is not utility. I have chosen to use Raymond's terms but have modified their definitions slightly to adjust for the differences in the two endeavours. The trackers, the home recorders, and the MP3ers are all part of the Bazaar. The major distributors and the distribution mechanisms comprise the Cathedral, siphoning the creative worth of musicians for monetary profit while remaining distant and unreachable from the creative and consumer bases. Today, like with software, thousands of musicians are creating and distributing music over the Internet, primarily because of inherent reasons, such as a love for music or creative ego, rather than any intention of making profit. As a result, a lot of this music is freely copied and distributed, and forms a key component of the Bazaar model. Creativity in the Bazaar occurs in a bottom-up environment (there are no restrictions; it doesn't even have to "work") as opposed to a top-down environment in the Cathedral (the major labels impose "rules" such as "has to sell well" on any creative output).

Chaos and complexity

So why should this bottom-up creation and distribution model work? It works because the individual agents in the system (artists, listeners, distributors) are not constrained by top-down rules, i.e., they have freedom. The freedoms are diverse and not necessarily explicit, including the freedom from commercial interests and the freedoms to modify, stand on the shoulders of giants, and improve. This diverse set of freedoms enables a work under scrutiny to evolve, following a non-deterministic exponential trajectory, i.e., in a chaotic manner. This results in an immense amount of creativity: not only is a given work built upon which it is built upon which it is built ... but this development also occurs in parallel and each time the output is different (it deviates from another trajectory or path exponentially)!

In other words, if someone distributes a song to many people, then it is likely that more than one person will use it as a starting point for a new-derivative work. Each of the people who create these works will do something very different (given the subjective nature of music) and distribute these modified works. Now more people are going to have access to this work which they will use as a starting base and the cycle will go on. The time evolution of the work follows a "non-linear" path or trajectory, and the differences (however you measure it) between any two paths grows exponentially over time. The non-linearity in the system results in non-determinism: each time this song creation process is repeated, what will happen will be extremely different from what happened before (thus the increased amount of total creativity). It is almost as if the work has a life of its own (the term for this is "emergence"). This evolution of this emergence has tremendous implications for issues regarding creativity, intellectual property, and censorship in music, and, ultimately, the future of the Cathedral.

The Bazaar vs. the Cathedral and are two examples of sites that enable distribution of Free Music (Free Music Archives or FMAs). In terms of volume of copies distributed, normalising for the relative obscurity of most artists, these sites are doing phenomenally compared to the distribution achieved via normal channels. I am an artist who distributes music through both these FMAs and my own site (, without any intellectual property restrictions whatsoever. I receive e-mail from total strangers who've incorporated music or sounds I've created. I also use ideas, notes, chords, and sounds that arise from people I know who use what arise from my mind and so on. It has gotten to a point where I've stopped keeping track of who's doing what, but just go on creating. Cogently, I realise that I am creating by standing on the shoulders of giants who are standing on the shoulders of giants--it's like the mythical idea of an infinite number of tortoises, one below the other, supporting the earth. The bottom line here is that there's a complex web of unconstrained creativity based around our creations which is spreading in an exponential manner as a result of Free Music.

Compare what happens in the major label Cathedral to what happens in a FMA Bazaar. The Cathedral essentially dictates what gets played on the marketplace. Mainstream publicity channels such as MTV and commercial radio are part of the Cathedral. Every creative work that comes out bearing the major label mark has gone through a bureaucratic approval process in the Cathedral, having been deemed to fulfill certain requirements for propriety, profitability, and controllability. It's debatable whether the artistic content of these works are even an issue.

I assert that the Cathedral is dying, and the reasons it will die are as follows: (i) Top-down systems cannot adapt as quickly as bottom-up systems. In the digital world, composition and compression technology changes rapidly. If a new format emerges tomorrow, is more suited to become before a company like Warner Bros. even realises what's happening. (ii) Niche markets in music will be the norm as opposed to being the exception. A large monolithic major label structure is not flexible enough to afford to exploit the small niche markets, whereas the FMAs can cater to every individual's taste. This isn't just because of the Internet, but because of the worry-free manner music is available on an FMA. (iii) Creativity is exponentially and non-deterministically enriched in the Bazaar. (More on this below.) (iv) Intellectual property controls are in direct conflict with points (i)-(iii). This has been argued cogently, in spirit at least, in John Perry Barlow's The Economy of Ideas and Raymond's The Cathedral and the Bazaar. The salient points are that worrying about intellectual property slows down reaction time as one is questioning how controllable a new technological format is, as the RIAA is currently doing with its actions (i); the potential profit margins in niche markets does not make it worth the worry price, generally what you pay the lawyers (ii); and the Bazaar, which enriches creativity in a radical manner, does so only at a risk of not maximising profit (iii).

The Cathedral will lose in competition to the Bazaar for all the cases in (i)-(iv) and Darwinian selection will see to it that the majors go the way of the dinosaurs. Again, I will stress that the bottom-up work model works only so long as the Bazaar remains a bazaar, i.e., the freedoms I speak of above exist (in practice at least). Also, it's possible that Cathedrals today will also adopt the Bazaar model and find other means of generating revenue besides basing it on the control of the spread of creativity.

Free Music: what it means

One can argue that sites like or are attractive because you're getting something without paying for it. But what you also get when you download a song is total freedom to do with the song as you please (at least for personal use, or until no one finds out), even though this isn't necessarily what the intended effect might be. Generally people who endorse a paradigm like the Free Music Philosophy understand the distinction between getting music for free and being able to copy music freely, i.e., without intellectual property restrictions.

But what is the fundamental reality? Why are the RIAA and the major distributors afraid of the prolific spread of digital audio? Sure, it threatens their oligopoly, but not just because the Internet has somehow made music more accessible to the masses, but rather because the distribution of music on the Internet follows a non-linear exponential trajectory without control. The fact of the matter is that more and more people are completely apathetic to the intellectual property concerns when downloading or even distributing music online. And they can't afford to be anything but. This scares most people because they cannot control what has happened with their creative output, and it's even more terrifying when the control of creative output is fundamentally linked to economic input. Lack of control on the part of the recording industry means more freedom for the musicians and the listeners. My main thesis here is that it is this freedom that is the key to making the Bazaar model work.

The freedoms I speak of above has the greatest impact on people who create or wish to create music. A few years ago, when I started getting into making recordings, musicians were looking at analog 4-tracks and 8-tracks and saying, hey, we can now produce great sounding albums without spending a lot of money. Today, with digital recording in place, musicians can now produce recordings that sound as good or better than a major label record in their basements. And the digital world has become the great equaliser: musicians can also distribute this music widely, probably better than what a major label would do for the average musician.

This ease of distribution is what makes music free, not how much it costs for the download (it costs someone somewhere something to download an MP3). Music is free because you can let your friend hear it, copy it, play it to their friends, and so on. In a more extreme situation, music is completely free when another musician can use your creation as a starting point for their own creation. This is when Free Music is at its most valuable. And without this freedom, human creativity will not likely be seen at its most awesome potential.

The future: a world of musicians

From a creative perspective, where I see all this going is that music becomes like language. Just as no one owns the English language, no one will own what music will become. The reason no one can own the English language is because of the number of people that have contributed to it, molded it, and made it grow and adapt--it's either owned by everyone or no one at all. It's a continuous and complex dynamic system, evolving in a non-linear manner, and growing from the previous changes created by feedback loops. The same will apply to music in the future: a song you write may involve so many contributions and meta-contributions that to claim exclusive rights to it would be a joke. Like a language, music will be a collage of ideas, notes, chords, and sounds from many many different creative minds. The term "collage music" already exists to describe such a phenomenon, pioneered, in part, by the views of artists like Negativeland and John Oswald and embraced by genres like techno. Music will be the communication that begins where conventional language ends.

There is a large proliferation of hard disk multitrack recorders in the music scene today. Consider a scenario where you can not only make your songs available mixed down in MP3 format, but also each of the tracks in MP3 format such that software and hardware-based MP3 players can handle data track by track. Imagine the possibilities: Don't like a guitar solo in the middle of the track? Edit it out, or record your own solo! Want to change the drum kit in the drum track? Given the sound to MIDI converters, this will be doable in real-time, so you can assign drum patches to a real drummer. Even the smallest tweak in the mix may result in a new song for the person listening it. Of course, most audiences will listen to what they're fed, but I argue a greater number of people (the audiophiles at least; people who don't play instruments but are picky about sound) will start "fine tuning" songs to suit their own tastes, much the way people adjust brightness and contrast and colour on TV sets. You can have multiple options for a given track and have it randomly played so the song changes everytime you listen to it!

We're only a few steps away from this becoming a reality: multitrack MP3 manipulators are technologically and economically feasible. MIDI already permits this sort of manipulation at the composition level, but unfortunately, there is no format that merges MIDI and sound elegantly. While I believe in compartmentalisation and think the different protocols are suited for the different things they do, it's not inconceivable to imagine a end-user software or hardware machine that takes any combination of MIDI/MP3 (or any arbitrary composition and compression format) and permit track by track manipulation. In fact, even though the process isn't entirely straightforward, many musicians (including myself) have collaborated with others over the Internet by exchanging multitrack soundfiles, tapes and MIDI files, and even interactively, without ever meeting.

The Bazaar model will enable creative endeavours between musicians who have the time and the inclination to pursue a full-time career and those who do not wish to dedicate their life to music exclusively. The latter is generally a requirement for working as part of the Cathedral. Some musicians want to create and then not have to tour, some might not want to promote as actively, and some may just prefer to remain anonymous. All of these musicians will have an excellent opportunity to be heard.

Another method by which I think creative cross-fertilisation will occur is by coupling appreciation of musicians (i.e., payment) with creativity. For example, in one of the FMAs, if the artists get a percentage of advertising revenues based on song downloads, then rather than just having the option of receiving actual cash, they may also receive hard copies of music by other bands. This way, an incestuous relationship between the artists will be developed. Given a large population of musicians, which will grow if the above multitrack models are implemented, this will result in a self-sustaining complex system with unimaginable creative dynamic. We're all musicians as well as listeners. The potential for breeding creativity is even greater if other creative ventures such as software, visual art, and literary art are coupled with music.

Economics in the Bazaar

I have dealt with the questions of "how will musicians make money" and "won't musicians starve", if they cannot control copying of creative works, extensively in the Free Music Philosophy, the primer on the ethics of "intellectual property", and other missives related to Free Music. Simply put, since a vast majority of musicians don't make a living from music anyway, the Bazaar model isn't going to make things worse and may make things better (because of the increased exposure), relative to what can be done with a major label. I quote from the Free Music Philosophy:

Musicians currently make money through a variety of sources: sales of records, merchandise and concert tickets, and royalties from commercial airplay. Freeing music will certainly not be detrimental to the sales of merchandise and concert tickets, nor will it affect compulsory or performance royalties. If anything, it will improve sales since people will continue supporting artists they like by going to their concerts and buying their merchandise. Profits from record sales will also not be affected because people will be encouraged to buy directly from the artist for the added bonuses of liner notes, lyrics sheets, and packaging. Thus Free Music can be used as a marketing tool to ensure that musicians do not starve. An approach where people send the artist a "donation", if they found value in the music they copied, is another way to make money in a direct fashion. This could become an ingrained practice in society, like tipping, where even though there is no enforced requirement to tip for various services, people do anyway.

As the Bazaar model follows its complex trajectory, the economic solutions will automatically arise in a complementary manner. Commerce abhors a vacuum.

The Bazaar is thriving and the Cathedral is dying

Imagine a complex adaptive web, where a musician records a song and distributes it with all the tracks. A listener adds reverb and echo to parts of certain tracks which is further distributed to other musicians and they sample or use parts of the modified track. Perhaps the original musician is fed back these modifications and creates a new variation which is further distributed. And so on. Imagine the richness of music that will result. That is the future. It's already happening.

Credits and further reading

This missive was inspired after conversations with many many musicians, listeners, and other visionaries. I thank them all for their intellectual input.

You might be interested in reading other documents related to Free Music.

Free Music || Ram Samudrala ||