READ IN French
Computer Science / Book Review

Vol. 4, NO. 2 / October 2018

An Antique Land

Benjamin Thierry

Letters to the Editors

In response to “An Antique Land

Minitel: Welcome to the Internet
by Julien Mailland and Kevin Driscoll
The MIT Press, 240 pp., $35.

With their book, Minitel: Welcome to the Internet, Julien Mailland and Kevin Driscoll have sought to fill a historical gap. Outside France, accounts of the ongoing digitalization of daily life have largely ignored the importance of telematics as a step leading to the emergence of the web. There is, in particular, the development of the Minitel network in France. “As the first computer network to reach mass-scale participation,” Mailland and Driscoll write, “Minitel presaged the near-universal adoption of Internet access in many major metropolitan regions today.”1

In Minitel, the authors provide a detailed account of France’s entry into the age of telematics, which spanned the early 1980s until around the turn of the century. The development and implementation of Minitel is examined in detail, as well as the services that were developed, and the social impact of the network. This is a story, it should be noted, for which primary sources are scarce. The messaging archives are long gone and there is no software left to evaluate. These shortages notwithstanding, their book is enlivened by Mailland and Driscoll’s mastery of the surviving archival material, and by their efforts to incorporate evidence drawn from interviews. The book is an essential reference.

As meticulously researched, richly detailed, and well informed as it is, Minitel is more than a historical recapitulation. The book defends a twofold thesis: first, that Minitel, despite protestations to the contrary, was not a complete failure; and second, that a careful analysis of its operations and governance can yield some useful lessons.

Launched in the late 1970s, the goal of the Minitel project was to equip every French telephone subscriber with a free terminal that provided access to a wide variety of online services. The initiative was, in part, intended to address the dire state of the country’s telecommunications infrastructure. “At the time,” the authors note, “the French telephone network was one of the worst in the industrial world.”2 Revenue generated from voice services alone was insufficient to fund such an ambitious nationwide program. This led to an unprecedented industrial and commercial challenge. The Minitel terminals, sourced by the government from private-sector manufacturers, would be issued with no upfront fees. Customers paid only for the time they spent connected to the system, the fee added to their monthly phone bills. Play now, pay later, as the authors put it.3 The terminals were distributed by the same agencies that had previously been responsible for the circulation of phone books; a free, searchable online telephone directory, l’annuaire electronique, was one of the first services available on the new network.

Twenty years after its launch, the French government’s gamble appeared to have paid off. In 1991, when the system reached peak adoption, a fifth of French telephone subscribers had a Minitel terminal. Citing a 1991 report by France Telecom, the authors note that adoption rates “ranged from 23.9 percent in Paris and 22.4 percent in Lille to 16.2 percent in Clermont-Ferrand and 15.3 percent in Ajaccio on the island of Corsica.”4 In 1993, the annual connection time for the system amounted to 90 million hours. It was not to last. The system began an inexorable decline. By 2005, annual connection time had shrunk to 20 million hours. Two years later, that figure had halved again.

The network was shut down in June 2012.

Among the many services available on the Minitel platform, it was adult services and, in particular, the messageries roses (pink messages, chat rooms) that proved to be one of the most popular aspects of the system. “Adult-oriented services,” the authors note, “were not cordoned off or marginalized; Minitel was infused with sexuality and innuendo.”5 Users interacted with each other in a digital masquerade. The extent of the novelty was matched only by its success. A 1986 study found that 70% of the 1.7 million hours of connection time during the months of May and June that year were spent on messageries (chat rooms).6 From these figures, the authors calculate that the annual revenue from the messageries industry as a whole would have amounted to 879.6 million francs.7 The early online adult services, dubbed Minitel rose, also proved controversial. There were numerous efforts to ban or curtail the services on moral grounds. Amongst other things, concerns were expressed that the messageries roses were being used by prostitutes for solicitation.

The controversies associated with Minitel rose have, to some extent, overshadowed the wider success of the system. Mailland and Driscoll emphasize that there was much more to Minitel. From its public launch in 1984, Minitel users were able to shop online, read newspapers, book train tickets, and check their bank balances. Students were able to find resources to help with their homework, and businesses could manage remote locations and access databases using the services.

In the mid-1980s, computer geeks traveling to Paris were confronted with glimpses of a telematic future. By 1986, advertisements for Minitel services were ubiquitous on the highways and high streets of France. All along the dark, gloomy taxi ride from the airport to the city, colorful billboards implored drivers to tapez, or “type,” inscrutable alphanumeric sequences into their home terminals.8

These services, both in terms of their variety and effectiveness, prefigured most of what we do online today.

  • Shopping
    3615 TMK (Tele-Market) was an online same-day delivery grocery shopping service with an inventory of several thousand items. In 1987, there were “four different services focused on delivering to the Paris area, and twenty-three total in France, enabling one to order from large stores, specialized wine retailers, or straight from local farms.”9
  • Science
    Rather than searching for extraterrestrial life, as was the idea behind SETI@home, thanks to the Cosmos Art Initiative, “in 1999, Minitelists were able to chat with aliens … Messages typed on the Minitel screen were beamed into space through the Nançay radio telescope.”10
  • Natural Language Searching and AI
    Along with searchable phone directories, “[t]ruly automated personal assistant services with natural-language interfaces began to appear around 1987, such as 3615 AK, a public-facing database of health information similar to WebMD.”11
  • Ticketing and Events
    The Billetel service allowed customers to purchase tickets to events from their homes, which would then be delivered by mail, or could be collected from an access point.
  • The Minitel of Things
    Interest in the so-called “internet of things” and the notion of the smart house has increased sharply in recent years. Domotique was the Minitel of things: “devices from the 1980s included thermostats, VCRs, security systems, lights, yard irrigation, and even kitchen appliances.”12
  • Financial Services
    Banking was always central to the Minitel platform: “Services ranged from checking balances and making appointments with bank personnel, to ordering checkbooks and transferring money.”13 Same-day online trading and portfolio management services also became available from the late 1980s.
  • Point of Sale
    Using the LECAM system, secure payments could be made to “a merchant from home or a point-of-sale system, in real time.”14 A smart card reader was developed, meaning that “Minitel became an affordable point-of-sale system enabling businesses to accept payment cards.”15

Rather than simply trying to prove that all the online services we use today had already been invented long before the advent of the web, the authors argue that Minitel shows how telematics was the basis for the development of original services that met the needs of both businesses and consumers. Mailland and Driscoll’s most important contribution is not the gallery of carefully selected examples, but the reasons they summon to explain their success. In the United States, in particular, Minitel is often presented as

the quintessential public works project gone awry: centrally planned, controlled, censored, and laden with bureaucracy. Economist Eli Noam called it a “technologically backward system,” and as media historian Fred Turner observed, Minitel is considered a “joke” in Silicon Valley, and an example of “what not to do.” … [V]enture capitalist Jacques Vallée described Minitel as the “exact opposite” of the Internet: “a closed system with no ability to grow organically.”16

The authors argue that the economic, political, and administrative conditions from which Minitel emerged were, in fact, a subtle mixture of both state interventionism and private initiatives.17 Minitel was not simply a “gated community or walled garden,” but “a computational and economic platform that displayed many features of openness.” In gated communities, it is the service provider that has the final say regarding publication and access to content.18 Organizations such as Facebook or Apple exercise this discretionary power freely. In the case of telematics, public authorities are both access providers and moderators. If a content provider wishes to contest a ban, or a user objects to blocked content, a recourse to arbitration by the telecommunications authority, and ultimately to the law, is possible. No such recourse is available in an environment managed by private entities.19 It is for this reason, Mailland and Driscoll argue, that Minitel should not be dismissed out of hand as just another closed centralized system.

Consider the response from Minitel’s administrators to the moral panics that accompanied the development of Minitel rose during the 1980s and 1990s. Repeated requests for controls on content were rejected by the authorities in order to maintain the neutrality of the network and to preserve the confidentiality of personal communications. By way of a comparison, in February 2011, Facebook unilaterally began censoring images of the painting L’Origine du monde by Gustave Courbet. Controversy ensued. L’Origine du monde, it should be noted, was originally painted in 1866. It now hangs in the Musée d’Orsay in Paris.

A platform can be understood as “any computing system with an interface ‘to which software applications can be written.’”20 A telematics platform should be considered semiopen or semiclosed. It is also infrastructure neutral. Antiope, the video display standard used by Minitel, was entirely separate from the networking protocols used by Transpac, the French public data network. The same was true of le kiosque (Kiosk), the Minitel billing system. A centralized approach to financial management proved to be extremely effective, not only for administration purposes, but also in encouraging further adoption and usage. This, in turn, allowed a wealth of service providers to emerge—more than 25,000 by the mid-1990s. As the authors note, “the payment structure of commercially successful platforms such as the Apple App Store more closely resembles Minitel and Kiosk than the historically decentralized Internet.”21

In French terms, the development of Minitel must be considered not only more important than the arrival of the web, which it prefigured, but in a wider sense as a notable innovation in its own right—a point long emphasized by French sociologists.

Translated and adapted from the French by the editors.


  1. Julien Mailland and Kevin Driscoll, Minitel: Welcome to the Internet (Cambridge, MA: MIT Press, 2017), 21. 
  2. Ibid., 4. 
  3. Ibid., 5. 
  4. Ibid., 10. 
  5. Ibid., 98. 
  6. Ibid., 99. 
  7. Ibid., 99. 
  8. Ibid., 95. 
  9. Ibid., 120. 
  10. Ibid., 121. 
  11. Ibid., 121. 
  12. Ibid., 122. 
  13. Ibid., 124. 
  14. Ibid., 125. 
  15. Ibid., 125. 
  16. Ibid., 14. 
  17. Ibid., 3. 
  18. Ibid., 88. 
  19. Ibid., 90–91. 
  20. Ibid., 16. 
  21. Ibid., 61–62. 

Benjamin Thierry is Associate Professor of Contemporary History at the Université Paris-Sorbonne.

More on Computer Science


Copyright © Inference 2024

ISSN #2576–4403