In 1830, a publisher named Lynde Walter launched a Boston paper called The Boston Evening Transcript. Transcript’s most important feature wasn’t its content or format, but its business model. A subscription cost only $4 a year, barely more than a penny a day. Walter could sell so cheaply because industrial production and middle-class consumption created a newly robust advertising market; slashing the price of the paper let him increase his audience so dramatically, he could more than recoup in ad revenue what he gave up in subscription fees.

Imagine giving Walter a tour of The Boston Globe today. He would not recognize the computers on reporters’ desks, or their phones. He would not recognize cameras, or delivery trucks, or even light bulbs. He would, however, recognize his business model, still at work. Transcript helped usher in modern news economics: The publisher gives the audience access to the news. The audience gives the publisher access to the advertisers. The advertisers pay the publisher for access to the audience. The publisher gets to keep providing the news.

The magazines and newspapers built mainly on advertising subsidy instead of reader support became collectively known as the penny press. The broadcasting industries of the 20th century—radio and then TV—followed this logic as well, with ad revenues providing virtually all the income. Though parts the news ecosystem don’t use this model—NPR, Voice of America, Ms. magazine—the subsidy of news via the triangle trade between publishers, audiences, and advertisers has been at the center of the American news industry for most of the last two centuries.

Every formerly stable leg of that triangle is buckling.

The problem for American journalism isn’t just that revenues are collapsing; the entire context in which traditional institutions operated is being altered. This leaves three options for American newspapers today (and for magazines and broadcast news in the near future): They can try to preserve their existing structure while shrinking their operations; they can restructure, changing not just size but organizational pattern; or they can collapse, simply extracting the revenues they can get before they vanish.

The most talked-about change in the old triangle is the relationship between newspapers and their audiences. The proliferation of the Web means that every publicly available source is now available to every member of the public. Even if digital distribution changed nothing else, that thousand-fold increase in competition would forever alter the news ecosystem.

Digital media also erodes audience habits. Publishers and salespeople often sold advertisers on the loyalty of their readers, but we readers have never actually been loyal. We’re just lazy. Prior to the Web, when options were scarcer, lazy meant continuity: We got the same paper we got yesterday; we read another article in that paper rather than searching out alternatives. Today, lazy means serendipity. We read what our friends send us, from wherever they got it. (Their friends, probably.)

We don’t select publications anymore, we select links. Even as the Web grew, publishers assured one another that the need for a trusted news source would preserve newspapers’ relevance, but it turned out that the trust we have in our friends is, for most of us, an adequate substitute for deciding what to read, watch, or listen to.

The relationship between advertisers and customers has also exploded. Some of this has been driven by new publishers—Gawker and HuffPo, Talking Points and BoingBoing—but far more radical is the rise of advertising as a stand-alone service, no editorial trappings anywhere in sight: Amazon, Google, Craigslist, Monster, Match, Backpages, Groupon, Freecycle, and on and on.

The last leg of the triangle, between advertisers and news outlets, hasn’t changed in any fundamental way, but it has changed enormously in practical ones. News organizations used to be able to overcharge and under-deliver in their deals with advertisers; the pizza place and the car dealership had nowhere else to go, and no one knew how many people saw, or acted on, a given ad anyway.