5

FROM THE MINITEL

TO THE INTERNET
The Path to Digital Literacy
and Network Culture in
France (1980s–1990s)
Valérie Schafer and Benjamin G. Thierry

When France Telecom retired its Transpac data network in June 2012, it put a definitive
end to the Minitel. This “intermediate technology”, as historian Pascal Griset once put it
(Griset in Schafer and Thierry 2012: 7), was an early alternative to the public Internet, and
a vestige of French voluntarism—both state and industrial—as well as an important, but often
overlooked, tool for digital education that died in France.
At its birth in the early 1980s, the Minitel had been far from an isolated project for countries
in the Organisation for Economic Cooperation and Development. The convergence between
telecommunications and computers, in what was known as telematics, seemed to be promising.
With state support, the monopoly of the telephone lines, the packet-switching network
Transpac that opened in 1978, and an electronic telephone directory project, the powerful
French Telecommunications Administration made a bet on services. The Minitel was a small
terminal in typical 1980s design that supported Télétel traffic and met many challenges
successfully. First, there were technical challenges, such as the creation of a massive database
for the electronic telephone directory, the development of an affordable terminal, the
invention of original interfaces, and interactive devices. Second, it responded to economic
challenges, as with the creation of a business model, a telematics market, and profitability.
Third, it also dealt with political challenges, as when it faced initial hostility by the media,
which brought some members of parliament along with it. And, finally, it met the social
challenges of introducing screen-and-keyboard devices into the home, attracting new users.
Yet, far from being a success story, in the mid-1990s, the Minitel gradually went from a
symbol of modernity and industrial voluntarism to becoming synonymous with outdated,
centralized technology, a “dumb terminal” blamed for “France’s Internet delay”. This “little
French box” entered history as it exited French quotidian life. It is time to consider the original
innovation that it helped bring about, and identify and evaluate the intersections and
divergences that unite telematics and the Internet, to try to understand digital culture as one
of constant reinvention and surprising continuities.

77
 V. SCHAFER AND B.G. THIERRY

The Minitel: A French Exception

From Captain in Japan, to Bildschirmtext in the Federal Republic of Germany, to the British
Prestel, to the Canadian Telidon, numerous videotex experiments were conducted in the
late 1970s, but none enjoyed the success or longevity of France’s Minitel. The Minitel did
little to heed the wishes of the French Telecommunications Administration to export its
product. This failure, which was caused by several factors, from the battle for a common
European display standard, to the Directorate-General of Telecommunications’ (DGT) desire
to export a turnkey system, doesn’t diminish the originality of the device that was developed
in the 1980s and 1990s, and which remains the chief symbol of the successful implementation
of a public telematics project. These lasting choices would set the Minitel apart, as well as
the Télétel interactive system more generally.

After the Phone Catch-up

In the first half of the 1970s, before the idea of launching a large-scale telematics public project
was ever put forth, the DGT, under Presidents Georges Pompidou and Valéry Giscard
d’Estaing, made a last attempt at bringing the telephones up to date. Indeed, for nearly a
century, France had been cursed with a state of poor equipment and chronic breakdowns
linked to high costs, lack of planning, and mismanagements. This even became a laughing
matter, giving rise to the saying, “Half of France waits for the phone, the other half waits
for the tone”. A total of 442,000 applications for a phone line were awaiting treatment in
1966, in rural and urban areas.
To remedy the situation, the DGT, led by Gérard Théry from 1974, undertook Delta
LP, an unprecedented plan to upgrade the telephone lines and infrastructures. From 1970
to 1975, the telephone lines had increased from four to seven million. In parallel, the
DGT, which held the monopoly, was already brainstorming how to recoup its invest-
ment after the upgrade. It would be what wasn’t yet known as telematics, but was at this time
called “téléinformatique”—a contraction of the French words “telecommunications” and
“informatique”—that later generated the establishment of new services with high added value.
From the beginning of the decade, the National Centre for the Study of Telecommunica-
tions (CNET) and the Joint Research Centre for Broadcast and Telecommunication (CCETT)
worked on implementable services on the telephone and its network. To this end, the
CCETT’s Bernard Marti developed Antiope, a standard for videographic data production
and dissemination, and then Titan, a first draft for interactive data broadcasting. The first TV-
based service, the Antiope Exchange, was founded in 1977.
For its part, CNET worked on a phone-based desktop computing system in 1970 that
allowed for simple operations to be completed from long distance on a computer that delivered
results using voice, and, later, graphics. The arrival on the market of the first American and
Japanese pocket calculators disappointed any expectations of marketing that service, but the
same team soon developed the Tic-Tac system (Terminal intégré comportant un téléviseur avec
appel au clavier) (Thierry 2013), an integrated terminal composed of a monitor and a
keyboard—similar in principle, but not limited to arithmetic. The system went public in 1975,
allowing for databases to be queried and their responses displayed on the screen.
In the late 1970s, France’s Antiope and Britain’s Prestel developed similar, but competing,
“videotext” systems (a term denoting a telecommunications service allowing pages of text
and simple graphics to be sent in response to a user request). In September 1977, at Berlin’s
Funkaustellung, the French presented Antiope, and Gérard Théry, fully aware of his system’s

78
 FROM THE MINITEL TO THE INTERNET

Figure 5.1 An Antiope page (created around 1980 by Bernard Marti). Attribution: I,
Liagushka, CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/3.0/)

capabilities, but taking note of Britain’s heavy promotion of their own system, decided to
ask his government to support further development of the new services. On 30 November
1978, the Cabinet agreed to a “full-scale” test of telematics, which would take the form of
an experiment in the Parisian suburbs. In 1981, this experiment, which later became known
as Télétel 3V, connected 2,500 users to 200 content providers under the aegis of the
Telecommunications Administration.
The year 1978 was thus pivotal. The opening of the Transpac network gave birth to an
infrastructure allowing for data transfer. The government decided on experiments with new
telephone services and manifested its political will. The Nora-Minc Report, named for its
authors, two finance inspectors, provided the modernizing ideology and accompanying
discourse on the necessity of the “informatisation de la société” (Nora and Minc 1978), meaning
the necessity for society to adopt computers on a wide scale, towards which end DGT would
be the central actor. Through its monopoly, the Telecommunications Administration was
able to claim not only a special role in the data networks, but also in services, all the while
supporting innovative work within the parameters of traditional activities. In effect, the
development of public telematics came through a flagship service that lent legitimacy to the
DGT, which was inspired notably by an experiment led on a much smaller scale in the United
States: the replacement of the paper telephone directory—a costly service that was becoming
overwhelmed by the growing number of subscribers, while telephone information services
were inundated with requests—by an “electronic directory” made available online.

79
 V. SCHAFER AND B.G. THIERRY

The Path of Technological Humanism

While having one or more interactive screens in the home or the workplace is now common-
place, it was not so at the end of the 1970s. The widespread adoption of screen-and-keyboard
devices providing access to databases, and later to “interactive services”, seemed highly
uncertain at the time. Would the French accept this terminal into their homes? And, another
question, partly contained in the first: even if they accepted telematics, would they understand
how to use it?
To simultaneously ward off the risk of early rejection, and allow for a user-friendly
experience, CNET and CCETT did not limit themselves to technical concerns during the
system’s design phase, but also initiated a comprehensive human-centered development process,
beginning in 1978. The first concern was to define an interface that could facilitate user
interaction with the services. This initial interaction actually developed right within the digital
telephone directory.
In 1979, its specifications, drafted by Jean-Paul Maury’s team, placed average users at the
heart of the discussion. The dumb terminal model and videographic standards imposed a
display simplicity that would impede neither the legibility nor the immediate intelligibility
of function. Instead of a scoring system, the screen displayed fields for the user to fill in using
a keyboard whose function keys could access help, get clarifications, or submit queries. It
was a middle path between the logic inherited from computing and programming, in which
the user made requests using lines of code, and the telephone information service’s model,
which was based on alternating questions and answers.
Definitive choices were finally made in 1982, following several rounds of field testing in
Saint-Malo and Rennes (in the Western part of France), in the context of a concrete, and rather
unusual, collaborative construct: average users drove innovation from the design phase for-
ward. This logic of active user participation led to a “full-scale” test of the first services, which
were offered beginning in July 1981 to 2,500 “guinea pigs” in the Parisian suburbs. This experi-
ment, known as “Télétel 3V”, lasted until December 1982, and allowed for the exploration
of the terminal’s first uses: digital sales, synchronous and asynchronous messaging, and infor-
mation consulting. It also ensured the possibility of imagining early public adoption of telematics.
This test phase was a critical step in the development of French telematics. The human-
centered approach was particularly striking in the changes it brought to the keyboard and its
layout, to the way forms were formatted, and to the various services that were provided in
response to participant/user feedback. It also strengthened this characteristically French
approach to developing the telematics interface. Text-based, without a scoring system, and
based on the most “natural” user-system interaction possible, it diverged with US experiments
conducted in the same period that led to the appearance of graphics interfaces like Windows,
Icons, Menus, Pointing devices (WIMP), which had been popularized by the microcomputer.

A Winning Economic and Technical Bet

At the same time, an economic model was emerging. Very quickly, telecommunications
agreed that the terminal, whose cost ranged between 500 and 1,000 francs (from US$80 to
$160, approximately) when produced in bulk, should be provided to customers free of charge.
It was a risky bet. Tacked onto the cost of the network infrastructure, it transformed the
project into a real gamble. The evaluation of the overall financial requirements is complex,
but from 1984 to 1995 it required an investment of around 17 billion francs (approximately
US$3 billion).

80
 FROM THE MINITEL TO THE INTERNET

Figure 5.2 Minitel 1 (France, 1982). Attribution: I, Deep silence, CC-BY-SA-3.0

(http://creativecommons.org/licenses/by-sa/3.0/)

Nevertheless, the break-even point seems to have been reached shortly after the early 1990s
(Cats-Baril and Jelassi 1994). In 1993, 6.5 million Minitel terminals, and nearly half a million
modem cards allowing network access via computers, generated revenue to the tune of 6.7
billion francs, of which three billion went to service providers. This success was partly thanks
to a novel pricing model called the “Kiosque”. Developed in 1983, introduced in 1984, and
extended to the whole country in 1985, this system was based on paying for service use
according to duration as indexed across several different pricing levels. The amount was directly
applied to the customer’s phone bill, and 60 percent of it on average was paid to service
providers. For most services, the cost for one hour of use was between 50 and 70 francs
(approximately US$9–12).
This pricing led the entire ecosystem towards very fast growth: profitability was viewed
favorably by service providers, who were multiplying and increasing the variety of services
offered, which in turn increased the number of users, and so on. Between 1984 and 2000,
net revenue was about four billion francs, for a total cost of 60 billion francs, with a rate of
return oscillating between 11 and 15 percent, according to various interpretations. From an

81
 V. SCHAFER AND B.G. THIERRY

industrial standpoint alone, the national effort served its purpose, injecting 5.5 billion francs
into the components industry (Masset 1986). Content providers also reaped rapid and
significant rewards, and new market entrants stood out, like AGL, whose famous pink Ulla
messaging service earned US$16 million profit in 20 years.
The Telecommunications Administration could congratulate itself for having made a
successful economic bet, and for accomplishing a technical feat. The Transpac network
encountered some very occasional difficulties in supporting the Minitel traffic as it grew
exponentially in the mid-1980s, but it was a robust network, secure and capable of adapting
and evolving with its customers. The system was based on PAVI (videotex access points),
which used modems to established links between the telephone network and Transpac, as
well as data conversion between analogic and digital modes. These achievements were made
possible with industrial support such as, in this case, that of Alcatel. At the end of 1991, the
network included 120 PAVI with 61,000 access ports, 20,000 of them pointing to the
electronic directory.
From a technical point of view, the electronic telephone directory was not the least of its
challenges: it included an access point connecting to the network, a “query center”, and a
database benefiting from a distributed architecture. The main challenge was to support the
large number of simultaneous connections entering from the front end of the query center
and the database. Service development and implementation, as well as information storage,
were also key technical goals. For services, page creation moved gradually from a traditional
method to an industrial system—at times, it took as long as eight days to build the first pages,
working bit by bit. For accommodation, the Administration allowed service providers to
choose between operating using the dedicated informatics service provided by the
Administration, the CITV (Computer Centre in Velizy Télétel), or their own system, like a
service company or an internal computer system. Most providers were happy to use CITV,
although some large companies preferred the second option to maintain control over their
technical developments.
When it came to developing the electronic telephone directory, the main challenge was
scale, even if the numbers might inspire a chuckle by today’s standards: the database in use
at the time was the world’s largest, identifying the information of 23 million individuals,
making 40,000 queries per day, for a daily update requiring nearly 20 gigabytes. In 1987
alone, the electronic telephone directory generated ten million hours of queries and 294
million calls in 1987 alone.

The Emergence of a Digital Culture in France

The development of a viable infrastructure and a free terminal, together with the wide
availability of services, brought France into its digital childhood earlier than other developed
countries. Indeed, a real culture for and by the digital settled permanently within equipped
populations. Many of the users were young people (25–45-year-olds), employed in the service
sector, with a post-Baccalaureate education. While exceptions to this profile may apply, the
Minitel was a medium widely and mainly disseminated in the middle class.
Email communication was a salient feature of this gradual digital acclimatization. Following
research on email conducted at CCETT in 1977, several services were offered to Minitel
users like Velizy’s messaging system, M3V. Others paved the way for “chatting”, like Gretel,
launched by the newspaper Les Dernières Nouvelles d’Alsace. They were followed by the so-
called “pink” chat rooms in the middle of the decade. Both the most popular “pink” messaging
services were Aline and Cum. They allowed online dating and erotic live chat between two

82
 FROM THE MINITEL TO THE INTERNET

or more users, and they made many headlines by questioning the place of sex and the
regulation of trade in this new medium. Those who were attracted to direct communication
with other Minitel users were the biggest consumers, generating up to 50 percent of Kiosque
traffic. Some “amateurs” collected astronomical bills. The great freedom that prevailed in
these virtual discussion spaces does not alone explain their success. “Taxi Girls”—hostesses
who maintained the interest of Minitel users with enticing usernames and unambiguous
subjects of discussion—also generated higher and stable attendance. These professionals,
pretending to be users, often assumed several identities simultaneously to stimulate conver-
sation, keep men (the main users) online and increase the benefits of the service providers.
Moderate users, the vast majority of Minitel users, did not participate in these discussion boards,
and focused on more directly practical messaging services.
Orders for merchandise via digital correspondence, especially textile, played a major role
in generating interest in telematics. Not only did the sector’s big names quickly seize upon
this new tool, but the shortening of order and delivery time, in comparison to traditional
postal logistics circuits, also won over customers. In 1983, the famous French mail order
company Trois Suisses launched its service, generating 400,000 connections annually, which
represented 3.5 percent of the company’s total revenue in 1986, and over 10 percent in 1988.
These practices became a permanent part of the French consumer landscape in 1994, when
1.2 million households placed orders on a Minitel mail order site. In that same year in the
United States, only 800,000 homes engaged in similar online shopping on the Internet.
In terms of information search, the Minitel also found success. Despite the press criticism
when the project was announced at the beginning of the 1980s, by the end of 1983 the daily
newspaper Libération, soon followed by others, had created a digital information service.
Similarly, banking and travel services settled permanently into the French telematics land-
scape. The SNCF (railway company), which had been involved in Velizy’s early experiments,
provided access to ticket booking and viewing train schedules. The French bank Banque de
la Cité was the first institution to offer online accounts in 1985; five years later, this type of
service took up 11 percent of online traffic.
For companies, Minitel also became an everyday tool. From the end of 1978, a number
of infrastructure projects were in development. In the banking sector in 1980, following the
first experiments transmitting stock information, the Bank Message Switching Centre (CCMB)
(Centre de commutation de messageries bancaires) opened, offering a complement to SWIFT
international banking—the transmission of payments and wire transfers. In the field of
transportation, the Tourist Teleinformatics Service allowed travel agencies to interact with
the reservation systems of major transport companies, beginning in 1979. More generally,
supplies were increasingly traded via Minitel. In 1992, the telematics business generated more
than 12 million connection hours, while other services generated 65 million.
Nevertheless, a slowdown in the growth of the number of terminals began to be felt at
the beginning of the 1990s. Numbers fell in 1995, when the Internet and the World Wide
Web began to generate discussion. The era of competition had begun.

Rethinking the Minitel’s Role in the “French Internet Delay”

In 2000, the report of the Working Group on the Future Internet highlighted the French
delay in general Internet use. Indeed, as of November 1999, the country had only 5.7 million
Internet users in private or professional settings, as compared with 110.8 million in the United
States, 12.3 million in Germany, and 13.9 million in the United Kingdom. An exacerbating
factor was the low market penetration of computers into France: only 26 percent of

83
 V. SCHAFER AND B.G. THIERRY

households owned a microcomputer at the end of 1999, compared with about 50 percent
of American homes in December 1998.
In this context, many identified the Minitel as a slowing, or impeding, factor. In late 1999,
the Médiamétrie Institute stressed:

The Minitel is [. . .] a serious competitor with the Internet in the field of online
services, but by having encouraged French familiarization with these services, it may
be a future ally when the Internet offers an interface as simple as the Minitel interface,
access tools to effective information, and richer and more attractive services tools.

Was the Minitel an ally or an obstacle to the implementation of the Internet? It was noted
at the time that:

The French have paid a price by using a system that the rest of the world does not
support. France is caught in a paradoxical situation. The Minitel has led to France’s
being behind the rest of the world in technology, the same problem Minitel was
created to fix.
(Kerr 1999: 11)

The whole matter is certainly a little more complex.

Did France Miss the Internet in the 1970s?

In 2013, Louis Pouzin, a French pioneer of research on data networks and the head of Cyclades,
a packet-switching network developed at IRIA (Institut de Recherche en Informatique et
Automatique), now INRIA (the French Institute for Research in Computer Science and
Automation) in the 1970s, received the Queen Elizabeth Prize for Engineering alongside Vinton
Cerf, Robert Kahn, Tim Berners-Lee, and Marc Andreessen for his participation in the creation
of the Internet. The French discovered with amazement that France might have passed by
the invention of the Internet in the 1970s. While it is also tempting to hold the Minitel primarily
responsible for France’s peripheral role in Internet development during the 1990s and 2000s,
such a statement is in need of greater nuance and contextualization.
Above all, it is important to emphasize that Louis Pouzin was undoubtedly a figurehead
in data networks, and the French contributed to the history of the Internet protocol (including
the genesis of TCP/IP), but this missed opportunity concerned the pre-Web Internet, rather
than its mainstream development. It should also be underscored that any historian
uncomfortable with alternative or counterfactual histories should be careful in responding:
without the Minitel, would France have gained faster uptake of the Internet? This is a
possibility, but by no means certain.
Undeniably, in any case, we can highlight France’s contribution to the genesis of the
Internet, and the role of Louis Pouzin and the Cyclades team at IRIA from 1971 to 1979 in
developing what became TCP/IP. In an email to David Reed, Vinton Cerf himself called
Louis Pouzin the “guru of datagrams” (Net History 2014)—one of the foundations of the
TCP/IP protocol which circulates data packets according to adaptive routing. Vinton Cerf
also mentioned Hubert Zimmermann, another member of the French Cyclades team, who
was one of the main designers of the first version of OSI, the seven-layer open architecture
that was presented in the 1980s and early 1990s as a serious alternative to TCP/IP.

84
 FROM THE MINITEL TO THE INTERNET

Cyclades was thus a milestone in the history of the Internet. In effect, it maintained a real
technical and relational proximity with the achievements of the US teams who worked
on ARPANET. In addition to inventing datagrams, its members participated in discussions
within the International Network Working Group to work out open protocol specifications.
They were also present at the first official demonstrations of ARPANET in 1972, and Gérard
Le Lann was at Stanford in 1973 alongside Vinton Cerf, who defined TCP, with Robert
Kahn.
Yet, despite these important achievements, Cyclades bore the brunt of the competition
with Transpac, a network which, as mentioned above, later allowed the traffic generated by
Minitel services. Indeed, beginning in 1972, the Cyclades and Transpac teams, which both
developed packet-switching networks, attempted to collaborate, but were unable to agree
on a joint project. Disagreements focused on the packets routing. Cyclades preferred an
adaptive model, while telecommunications, which was sensitive to the service quality,
criticized this “best effort” model and its potential pricing difficulties. This view led tele-
communications to opt for virtual circuits, whereby all of a message’s packets followed together
to ensure a safe delivery. Louis Pouzin called this a “belt and suspenders” model. The conflict
led to open criticism on the international scene, and a race towards standardization that
telecommunications won by standardizing virtual circuits to CCITT in 1976 (Després 2010).
However, it also meant the extinction of Cyclades in 1979. This was partly due to the influence
of the Telecommunications Authority, with its monopoly on the phone lines, and the ability
to offer both a network and a suitable commercial supply to businesses. After Cyclades was
buried, the Internet undoubtedly lost the battle in France: telematics and its thousands of
services reached the majority of users, while the Internet remained inaccessible to the main-
stream public for many years. Noting this paradoxical situation at the beginning of the 1980s,
the historian can only conclude with a question that remains largely open: did getting it right
too early ultimately mean getting it wrong?

The Internet from Early Adopters to the General Public:

Birth of a Market
In France, telematics allowed some scientific communications in the 1980s (Thierry 2013),
but it was not really a suitable tool. A small community of early adopters came together to
explore their interest in the “network of networks” via personal and institutional links within
the North American Internet space. Indeed, from the 1980s on, some researchers at INRIA
who worked or had links with the United States lamented not having the opportunity to
receive emails or news from the Usenet community, particularly those based on UUCP and
the Unix operating system. They found technical and organizational solutions enabling access
to these services. In 1983, when Berkeley Unix 4.2 joined with the TCP/IP protocol, more
users moved toward the Internet. This was not only the case in the United States, but also
a few years later at INRIA with FNET. FNET was the French branch of EUnet, the European
Unix community. It benefited from its membership by having access to the newsgroups of
the UNIX community through an agreement with CWI, a Dutch computing center. FNET
converted from UUCP to TCP/IP in 1986.
Later, in 1988, INRIA’s Christian Huitema used his relationships with US researchers to
obtain a direct liaison with Wisconsin (Griset and Schafer 2012). And in 1993, INRIA was
one of the co-founders of RENATER, the network for research and higher education, which
provided a backbone connecting the regional university hubs. In 1995, INRIA became the
second host, after MIT, of the consortium W3C that Tim Berners-Lee, founder of the Web,

85
 V. SCHAFER AND B.G. THIERRY

created to manage specifications and protocols (Griset and Schafer 2011), since in the
meantime, the Internet had changed scale.
The world of research, especially computer science research, comprised early Internet
adopters, but it is from the angle of service providers that we have to look for the driving
force that allowed the Internet to mutate from a tool for the happy few to a genuine
communication market. Although France Telecom didn’t launch Wanadoo, its public Internet
access, until 1996, several Internet service providers (ISPs) had already begun to explore the
new market. EUnet (as we have seen, its French branch, FNET, was part of INRIA) was
the first, claiming 53 percent of connected companies and laboratories; RENATER was the
second (37 percent of French Internet), followed by Oléane, a division of the IT company
Apysoft and member of the European commercial network Pipex (4 percent). Internet-Way,
created in 1994, should also be mentioned. It targeted professionals and was connected to
the European network Ebone. Other ISPs like Calvanet, Calvacom’s Internet branch that
was born in the Apple universe, as well as Francenet and Worldnet, offered Internet access
for the consumer market. And we can’t leave out the myriad of small regional ISPs born
before the Minister of Telecommunications wished that every French citizen could connect
to an ISP for the price of a local phone call (Rebillard 2012). This favored the establishment
of a mass national Internet market concentrated around a few large ISPs. And, indeed, the
Telecommunications Administration took the technological turn at that moment, after
hesitating about the Minitel’s possibility for evolution, as well as with offering proprietary
services like CompuServe and AOL.
Télécom Multimédia was a structure led by Gérard Eymery and associated with France
Telecom Group, at a time when two teams merged. The first was an “Innovation” team led
by Daniel Sainthorant at the Ad Agency, who had helped develop the marketing campaign
for the electronic phone directory during the 1980s, and who now carried out a fairly typical
reapplication of his telematics expertise to the Internet. His team worked on a “mall” project.
The second team was led by Jean-Jacques Damlamian from France Telecom’s Sales
department. It acquired skills by participating in the construction of RENATER. The
ambitious goals of this new entity, which was baptized France Telecom Interactive under
the chairmanship of Roger Courtois in early 1996, included launching Internet access for
the general public, marketing complete Internet offerings (electronic phone directory and
online services), and evolving the Minitel. In fact, the page did not turn quite so easily, and
for users it was still a time of caution, no less than for the service providers at first.
For most service providers familiar with telematics’ Kiosque and profitability, the transition
to the Internet, without a clear payment system or business model, seemed like a leap into
the unknown. They complained of poor service performance, network vulnerability, and
subscriber turnover, then estimated at 40 percent. In comparison with the multiple problems
highlighted (Internet saturation, crashing, hacking, unstructured supply), online services that
were crossed between Télétel and the Web (Prodigy, CompuServe, and AOL) seemed
reassuring. With the Web:

The user is “spoiled with choices” in the negative sense of the term: the first Internet
users even speak of “surfing the Net”, that expression [which] aptly characterises
such random navigation.
(Grellier 1995: 67–68)

Offerings were still dominated by the idea of a portal guiding navigation, which “stream-
lined” and thus maintained an influence, however small, on the consumption of services. On

86
 FROM THE MINITEL TO THE INTERNET

the supply end, the delay seems therefore to be explained by Minitel’s ability to create value,
while the Internet of the day had a less clear economic model. While Minitel users paid both
a base fee and an additional tariff for the length of time they spent consuming services, the
Internet promoted a certain amount of free use and services (paid through advertising or an
extra cost). As Henri de Maublanc, President of the Association for Online Trade and Services,
explained in the newspaper Le Monde on 30 May 2001: “Today, when all the publishers on
the Internet are searching for a way to get paid by Internet users for content, the Minitel’s
surcharged line model now passes for a stroke of genius.” However, at that time, the general
French consumption of telematic services had begun to decline, and the Web was able to
find its place in homes and businesses.

Continuities of Usage and Updating Through Practices

Some service providers joined the bandwagon, displaying remarkable adaptability. Such was
the case of the iconic owner of Free, Xavier Niel, who continues to make headlines today.
His first steps in pink chat rooms allowed him to build up a “war chest” that he reinvested
in the acquisition of a 50 percent share of Fermic Multimédia in 1990, which soon became
Iliad. Niel then embarked on a transition phase with his gateways to the Internet beginning
in 1994, before obtaining the right to operate his own network, and founding Free in 1999.
There were others who transitioned successfully from the Minitel to the Internet (for
example, Denys Chalumeau’s Seloger.com and Jean-David Blanc’s Allociné.com). A true
digital entrepreneur community had made telematics its classroom, and would invest in the
Internet market in the 1990s when it was still embryonic.
For users, the transition also came in the second half of the 1990s. While the cost of using
the Minitel was high, averaging around 60 francs per hour, the terminal itself was free. This
was not the case for Internet access, which not only was sold on a per-minute basis until
“unlimited” packages became available in 1999, but also required an initial investment of
several thousand francs to purchase equipment, specifically a computer that could connect
through its modem. From a logistical standpoint as well, the use of the Minitel had nothing
in common with the difficulties of dealing with a micro-computer whose emerging graphic
interfaces had not yet eliminated bugs and complexities (Thierry 2012).
During the 2000s, the catching up proceeded more quickly. In 2004, Europe’s average
monthly residential connection rate was 47 percent, with France located in the upper half at
49 percent. By the last three months of 2007, 72 percent of the French population was
connected at home and 40 percent at their workplace. By 2009, 19.8 million people, out of
a total population of 64 million, were connecting to the Internet every day, and daily
connection time approached one hour and twenty minutes.
Even more than these overall connection figures, usage continuities were updated, from
telematics habits to practices reinvested in the Internet. In 2008, one in five people in France
purchased movies, music, books, periodicals, or software online, while 10 percent of movie,
concert, theater, opera, dance, and sporting event tickets were purchased online (Berret 2008).
France led Europe in terms of connections leading to the purchase of a cultural product,
while it was tied for third place with Iceland, behind Germany and the United Kingdom
(which each claimed 8 percent) for purchasing movies and music delivered or updated online
(it claimed 7 percent). It is difficult not to see here the beneficial effects of online ordering
that had been popularized in the 1980s by the Minitel. The Minitel was an indisputable
pathway for French digital development, and it maintains multiple and complex relationships
with the Internet that are resistant to single factor analysis.

87
 V. SCHAFER AND B.G. THIERRY

As the outcome of political will and a pioneering vision of modernity enabled by

teleinformatics communication networks, telematics was an alternative to the world of
networking driven by the United States. The standards used, the “dumb terminal” model,
and the technological humanism that governed its introduction to all users, borrowed from
a professional culture specific to French telecommunications. The early success of services,
enabled both by the bet made on the free terminal and the pricing system that favored the
emergence of a rich and vivid ecosystem, makes France a pioneering nation in enabling the
immense possibilities of data networks available to the greatest number of people. As President
Jacques Chirac enjoyed recalling in 1997, the Aubervilliers baker could check the balance of
his bank account from home, while his equivalent in New York could not. This early start
should generate reflection about a society’s pathways to innovation. Although one emphasizes
the bottom-up model of Internet development, the model of entry into society from below
is symbolized by the Minitel, a “dumb terminal” whose ease of use and operation was part
of its initial success, but inhibited later its ability to evolve. On the contrary, at its beginning,
the Internet was a model of entry from above. Born in labs, designed as a tool for its own
designers, and progressively adapted for the consumer market, its plasticity guarantees its
continuity.
Today, while the Transpac network has definitively ended, the study of telematics is
revealing the winding and multicentered nature of innovation trajectories. The Minitel’s
history is, too often, briefly summarized or overlooked as an “intermediate innovation” in
relation to the technological winners of the day (Campbell-Kelly and Garcia-Swartz 2013).
Telematics also examines long-term continuities, beyond the boundaries of the object alone.
We can thus investigate the perpetuation of billing models initiated by telematics, which have
re-emerged today in mobile telephony and software sales—for example, with Apple’s App
Store, which is in some ways reminiscent of the DGT’s Kiosque. In an irony of history’s
turnarounds, isn’t today’s Web in the midst of a “Minitelization” of sorts, with its quasi-
monopolistic actors and its increasingly closed service offerings? A model widely denounced
by the various technical clergy of enlightened amateurs who argue for an open Internet while
denouncing the gilded cage that, nevertheless, the great majority of users call for through
their purchases? Time will tell whether this Internet Minitelization is only a fleeting moment
in the evolution of digital economies, or a lasting model initiated by the Minitel and
brilliantly promoted by Apple some years after. For today’s historian, these multiple reversals
are a reminder not to indulge presentist analysis, but to question the foundations of acquired
positions, and never to mistake the winners of the day for those of yesterday, and still less
those of tomorrow.

Archives
This chapter is based on archives that we consulted at INRIA and CCETT, on the French
National Archives and France Telecom Archives, on audiovisual and press archives. We also
had an interview campaign with around 30 protagonists of this history.

Further Reading
Ponjaert, M., Georgiades, P., and Magnier, A. (1983) “Communiquer par Télétel. Les Acquis de l’Expérience de
Télétel 3V et de l’Annuaire Électronique en Ille-et-Vilaine” (Communicating through Télétel: Skills Acquired
with Télétel 3V and the Electronic Telephone Directory in Ille-et-Vilaine), France Télécom Historical Archives,
99026/03 (a detailed study of Teletel experiments).

88
 FROM THE MINITEL TO THE INTERNET

References
Berret, P. (2008) “Diffusion et Utilisation des Tic en France et en Europe” (Distribution and Use of ICTs in France
and Europe), Culture Chiffres, 2: 1–15.
Campbell-Kelly, M., and Garcia-Swartz, D. (2013) “The History of the Internet: The Missing Narratives”, Journal
of Information Technology, 28: 18–33.
Cats-Baril, W., and Jelassi, T. (1994) “The French Videotex System Minitel: A Successful Implementation of a
National Information Technology Infrastructure”, MIS Quaterly, 18(1): 1–20.
Després, R. (2010) “X.25 Virtual Circuits: Transpac in France—Pre-Internet Data Networking”, IEEE
Communications Magazine, 48(11): 40–46.
Grellier, C. (1995) “Internet et ses Concurrents” (The Internet and its Competitors), Communication et Langage, 105:
67–68.
Griset, P., and Schafer, V. (2011) “Hosting the World Wide Web Consortium for Europe: From CERN to INRIA”,
History and Technology, 27(3): 353–370.
Griset, P., and Schafer, V. (2012) “Make the Pig Fly! l’Inria, ses Chercheurs et Internet des Années 1970 aux Années
1990” (Make the Pig Fly! Inria, its Researchers and Internet from the Seventies to the Nineties), Le Temps des
Médias, 18: 41–53.
Kerr, R. (1999) “Minitel versus the Internet in France”, Senior Research Projects, Paper 86, available at: http://
knowledge.e.southern.edu/senior_research/86
Masset, M. (1986) “La Télématique en Plein Boom” (Telematics in Full Swing), La Revue du Minitel, 5: 24–29.
Net History (2014) On the Design of TCP/IP, available at: www.nethistory.info/Archives/tcpiptalk.html (accessed
14 April 2014).
Nora, S., and Minc, A. (1978) L’Informatisation de la Société (The Computerization of Society), Paris: Seuil.
Rebillard, F. (2012) “La Genèse de l’Offre Commerciale Grand Public en France (1995–1996): Entre Fourniture
d’Accès à l’Internet et Services en Ligne Propriétaires” (The Genesis of the Consumer Commercial Offering in
France (1995–1996): Between Internet Access Provision and Online Service Owners), Le Temps des Médias, 18:
65–75.
Schafer, V., and Thierry, B. (2012) Le Minitel: L’Enfance Numérique de la France (The Minitel: The French Digital
Childhood), Paris: Nuvis.
Thierry, B. (2012) “‘Révolution 0.1’: Utilisateurs et Communautés d’Utilisateurs au Premier Âge de l’Informatique
Personnelle et des Réseaux Grand Public (1978–1990)” (‘Revolution 0.1’: Users and User Communities in the
First Age of Personal Computing and Public Networks (1978–1990)), Le Temps des Médias, 18: 54–64.
Thierry, B. (2013) “De Tic-Tac au Minitel: La Télématique Grand Public, Une Réussite Française” (From Tic-
Tac to the Minitel: Mainstream Telematics, a French Success), in Actes du Colloque Les Ingénieurs des
Télécommunications Dans la France Contemporaine. Réseaux, Innovation et Territoires (XIXe–XXe Siècles), Paris: IGPDE,
pp. 185–205.

89