Future Tense

The Very First Social Network

How a 1979 chain email about science fiction spawned the internet we know today.

Computer terminal with dancing figures wearing Star Trek outfits on the screen
Doris Liou

Adapted from LikeWar: The Weaponization of Social Media by P. W. Singer and Emerson T. Brooking. Out now from Houghton Mifflin Harcourt. More at likewarbook.com.

When computers formed the first network 49 years ago, the precursor to the modern internet was all about the science.

The brainchild of the Advanced Research Projects Agency—ARPA, the progenitor of today’s DARPA—ARPANET was envisioned as a way for military-funded scientists to share an incredibly rare and costly commodity back then: computing time and power. At the time, only a handful of universities had computers. The only way to use one (or to transfer a file from one device to another) was to travel to the facility where the computer was housed. IBM even had planes whose job it was to fly computer files back and forth. ARPANET promised to solve this, making it possible for scientists not just to share computing time but to direct multiple computers toward a single problem or acquire a mountain of data with a single, fast-as-light query.

In October 1969, the experimental network got its start when researchers linked a computer at UCLA to a computer hundreds of miles away at Stanford University. (Its first communication could only be considered a semisuccess: the network crashed during the typing of the very first word on the internet, and LOGIN became just LO.) Within weeks, a computer in Santa Barbara, California, and then another in Utah joined the party. By 1971, 15 university computer labs had been stitched together. By 1973, researchers made their first international connection by incorporating computers at the Norwegian Seismic Array, which tracked earthquakes and nuclear tests.

With the bold idea of computer communication now proven workable, more and more universities and labs also linked themselves together. But instead of joining ARPANET, many began forging their own mininetworks. There was an Atlantic Packet Satellite Network on the East Coast, the Packet Radio Network in the Bay Area, and ALOHAnet in—you guessed it—Hawaii (where it communicated via repurposed taxi radios managed by a computer named Menehune, Hawaiian for IMP: the ARPANET Interface Message Processor).

These various mininetworks presented an unexpected problem, however. Rather than forming one “Intergalactic Computer Network,” as the early memos envisioning ARPANET had described it, computer communications were becoming isolated into a bunch of little clusters. Each had their own set of infrastructure and governing rules, making it difficult to link them together.

Fortunately for all of us who are reading this article online today, this is when Vint Cerf entered the scene. As a young researcher in the nascent field of computer science, Cerf recognized this compatibility problem would keep computerized communication from ever scaling. So he and his friend Robert Kahn designed a single common protocol that could govern the transmission of data across an exponentially expanding “network of networks,” or “internet.” The adaptable framework, TCP/IP, allowed the original ARPANET to bind together all of these mininetworks that had been springing up around the world—and it remains the backbone of the internet this day.

Even with the expansion, however, this new mode of communication via computer remained limited to its original vision, helping a niche group of scientists conduct government-funded research.

That is, until 1979.

That fall, Cerf logged on to his workstation to find an unopened message from the MIT Artificial Intelligence Laboratory. It had been sent over the network using the recently developed “electronic mail” system. Because more than one person was using each computer on the network, the scientists had conceived of “e-mail” (now commonly styled email) so they could share information directly from one person to another, rather than just between computers. As with regular mail, they realized they needed a system of addresses to send and receive the messages. Thus the @ symbol was born: It served to separate the mailbox identifier from the serving host, and the single character saved typing time and scarce computer memory, an early version of what one might think of as a “hack.”

But the message sent to Cerf’s email wasn’t a technical request. And it hadn’t been sent just to him. Instead, an email with the subject line “SF-LOVERS” had been sent to Cerf and his colleagues scattered across the United States. The message asked all of them to respond with a list of their favorite science fiction authors. Because the message had gone out to the entire network, everybody’s answers could then be seen and responded to by everybody else. Users could also choose to send their replies to just one person or a subgroup, generating scores of smaller discussions that eventually fed back into the whole.

About 40 years later, Cerf still recalls this as the moment he realized that the internet would be something more than every other communications technology before it. “It was clear we had a social medium on our hands,” he said.

The thread was a hit. It also created what might be thought of as the first online social network. Though individuals had been connected via this internet before, this was the first time they were using it for social interactions and, importantly, building a larger community identity through these personal connections. After SF-LOVERS came YUMYUM, another chain email that debated the quality of restaurants in the new Silicon Valley. (In-house gourmet chefs were still decades away.) Then WINE-TASTERS appeared, its purpose self-evident. The socialization also inspired more science with HUMAN-NETS, a community for researchers to discuss the human factors of these proto-online communities.

Soon the forums began to be used to share something else: news. Here too the spark was science and science fiction, discussing things like the exciting rumors that the 1960s TV series Star Trek was on its way to a movie revival.

This new use of the computer network, however, created all sorts of new problems too. One was the fear that a user might share information that someone else didn’t want to see. This led to the very first online “[SPOILER ALERT],” which the author put above his message to warn readers before describing (spoiler) the death of a certain heroic Vulcan at the end of The Wrath of Khan.

More serious was the fact that U.S. military budgeters were less than excited about all the idle chatter this was producing on their expensive new network. They discussed how to banish it (perhaps the first debate over internet censorship) but relented when the engineers convinced them that the increased message traffic made a good stress test for ARPANET’s infrastructure.

Chain letters and freewheeling discussions soon proliferated across the network, and email began devouring two-thirds of available bandwidth. ARPANET was no longer simply facilitating remote computer use and file transfers from one database to another. Those who had access were now using it to create interactive communities, transforming what entire groups of people thought and knew by means of a networked computer connection. Soon enough, it would even change how they spoke to each other.

Perhaps no one —the engineers using it at the time included—understood how much they were laying the foundations of internet culture. In another first, at precisely 11:44 a.m. EST on Sept. 19, 1982, computer scientist Scott Fahlman changed history. In the midst of an argument over a joke made on email, he wrote:

I propose that the following character sequence for joke markers:


Read it sideways. Actually, it is probably more economical to mark things that are NOT jokes, given current trends. For this, use


And so the humble emoticon was born.

Despite all these early social networks, ARPANET was still not the internet as we know it now. It was a kingdom ruled by the U.S. government, populated mostly by Ph.D.s in a handful of technical fields (as the formal creation of the emoticon shows us). Even the early social platforms these computer scientists produced were just digital re-creations of old and familiar communications: the postal service, bulletin boards, and newspapers.

Yet it was growing fast. By 1980, 70 institutions and nearly 5,000 users had access to ARPANET, and the U.S. military came to believe that the computer network it was paying for had expanded far beyond its needs or interests. After an unsuccessful attempt to sell ARPANET to a commercial buyer (AT&T could have literally owned the internet, but said No thanks), the government split the system in two. They allowed ARPANET to continue as a chaotic and fast-growing research experiment. And they created a new MILNET for military use. For a time, the world of war and the world of the internet went their separate ways.

About four decades later, the story of ARPANET’s early social history may seem quaint. But it’s also informative. Today’s internet is now dominated by social networks and the companies that administer them. Facebook now has more members than any single nation’s population, and for many users in places like Thailand and Myanmar, Facebook literally is the internet. This evolution—from two linked university computers to a few dozen connections that supported nerdy discussion threads among a niche group of researchers to a globe-spanning, world-changing information network—shows how technologies built for one purpose are often redirected toward something else. And there’s a catch, because the story of the internet, in a way, is also starting to run in reverse.

SF-LOVERS saw what was originally a military tool transformed into a fun, increasingly informative online community. Today, however, these communities are being turned into battlefields, and the information in them weaponized to win everything from elections to wars. Social media platforms now play host to never-ending battles to drive ideas viral through algorithmically determinative “likes” and irresistible lies. The result can create and divide communities both online and in the real world. When we think of cyberwarfare today, the threat is not just the hacking of networks and network data but also the hacking of information, people, and communities on these networks. Such “likewars” have literally changed the world beyond the network. They shaped elections that changed American politics and the future of Europe. They influenced battlefield decisions in Iraq and Israel. And they also helped fuel a spate of gang shootings in Chicago and genocide in Myanmar. And they have happened in the very same place we continue to debate things like the best science fiction or the best restaurants, still using smiley faces and frowns. The first example of social media on ARPANET asked members what science fiction they liked, but now, “likewars” rage across its successor. And that may be the most sci-fi outcome of all.