When Composers Recordings Inc. decided to shut down in January, there were obituaries, lamentations and the usual muttering about the malaise of the recording industry. There was also a palpable sense of battle fatigue from those who have watched helplessly as the classical music recording industry tanked over the past decade: Major labels cut classical and new music releases, artists were summarily dumped from their recording contracts, and American orchestras (which had once enjoyed lucrative and prestigious recording gigs) were priced out of the market and mostly into oblivion by increasingly bottom-line-oriented corporations.

And yet, with the loss of CRI, which after almost 50 years had become a beloved and venerable fixture of the new music world, there was also a just barely audible sense of something new in the air. When things can't get any worse, then perhaps they may begin to get a little bit better. The story of CRI's closure supports all these readings, gloom and optimism alike.

"I've written a number of obituaries in these pages, but none so sad as this," wrote Kyle Gann of the Village Voice. Gann's obituary was, in part, a personal reminiscence of the music he discovered on CRI. In the music business, anything -- a record label, a singer, a string quartet -- that lasts 50 years is cherished as an inviolable part of one's personal musical landscape. Hence the mix of both rancor and sadness when that landscape inevitably changes. CRI, for many listeners, was not just an entree into new music but appealed to an anarchic way of listening: adventurously, without expectations, and individually, as an explorer of sound unfettered by what authorities (critics, professors, pompous friends) dictate.

Susan Sontag once called for replacing the old habits of interpreting, classifying and systematizing art with something she called "an erotics of art," a love of the real and tangible surface. CRI made possible an erotics of American music, because it was wildly eclectic and it subscribed to a policy of keeping its large catalogue in print. Young listeners, tired of whatever music they were weaned on, could find music on CRI that was, by virtue of being the forgotten avant-garde of 20 years before, far more foreign and fascinating than the newest of the new.

The label went through various cycles of history during its five decades of recording -- it put out a lot of academic music in the 1970s, and then took a turn toward the more raucously spirited "downtown" sound in the 1990s -- and it kept it all there, jangling together with stylistic discord in the same archive.

"Major labels delete titles," says Joe Dalton, the head of CRI during the 1990s. "There's a certain half-life for a new release, and then it's gone. The composers who founded CRI in 1954 had already had that problem, so they instituted the In Print policy."

This meant CRI not only made available the music of American composers but also strived to keep it permanently available. Composers who recorded with CRI were not likely to enjoy the marketing and distribution advantages that came with a large label, but they could rest assured that they wouldn't see their recordings disappear into the corporate vaults once the large label had moved on to the latest flavor of the month. And unlike the myriad small labels founded over the years to promote new music, CRI was a reliable brand name.

"When composers talked to me, I'd say one reason to go with CRI is that we're still here, while the entrepreneur down the road is going to own your master tapes when he decides to get out of the business in five years," says Dalton. The death of a brand name is about more than nostalgia; good brands take with them a legacy of trust and affection.

In 1976 CRI reorganized as a nonprofit record label. It raised funds to cover costs and expected composers to bring money to the table as well. It was never a vanity label, and its board was filled with composers, which helped give credibility to the release schedule. As the market changed, however, being a small nonprofit record label began to be a double liability. Because it was small, it was hard-squeezed by changes in the retail business, especially Tower Records' policy of paying labels only on a yearly basis for product sold. And larger commercial labels started asking composers to front money to fund projects as well -- reducing the pool of funds available for making difficult-to-sell projects for all involved.

"For-profit companies are now playing the nonprofit game," says Paul Marotta, managing director of New World Records, another small nonprofit American label (specializing in the entire history of American music back to the colonial era).

Unfortunately, CRI's policy of maintaining its recordings in print was not just the intellectual substance of the company but also a drag on its resources.

"I had approximately 15,000 CDs in a warehouse in Texas," says John Schultz, the most recent and now former executive director of CRI. "We would do a run of 2,000 CDs. Some would sell 200 copies and some would sell 400. Over 20 years, those unsold CDs accumulate, and the costs to house them. That was over $1,000 a month."

Schultz and his board of directors have attempted to close the company gracefully. CRI will liquidate its stock, with composers having first crack at their own CDs. It is also attempting to get legal permission to transfer its invaluable master tapes to Marotta's New World label, which has promised to keep them available in some fashion. Marotta says his company is looking at the possibility of purchasing equipment to produce individual disks on demand. It may also create an archive that can be digitally downloaded. One idea is a database service that would be made available to libraries for a fee.

At a Library of Congress news conference in January, talk of preserving the library's vast archive of recorded music included discussion of a new frontier in which sound is preserved not on any particular physical medium -- LP or compact disc -- but as computer files that act as a master source as new media and new ways of accessing music come and go. CRI died at a moment when the music business is trying to sort out its future, and at a time when consumers are facing a future in which the idea of physically owning a musical recording may be outdated.

"I believe that CRI may have outlived its usefulness," says Schultz. "Technology catches up, composers are either selling their own CDs on their Web sites or going to major labels. There is no shortage of market out there."

The test of this benign assessment of CRI's death will be the future of its music in the ethereal digitized new age. New World can keep it available somewhere on a computer, but without a presence in retail stores, consumers may not know what's available. The future of CRI's treasure house of American sound requires not just new technology and new forms of access but also a newly aggressive curiosity about music on the part of listeners. Purchasing a physical recording used to be a way of committing oneself to the act of listening; in the future, consumers may theoretically have more access to this music than they did in the waning days of the retail economy, but only if they develop a new "erotics" of listening.