Pierre Costa-Marini doesn't think of himself as an especially courageous man. True, he did finally break away to set up an ambitious business venture of his own, and, true, this was a brave thing to do in France. Nevertheless, he recalls that he spent most of his career in the safe confines of a government-supported corporation, convinced that in the nation that gave us the word "entrepreneur," there was little hope of becoming one.

"I always wanted to build something from scratch, but the very thought of it made people around me uncomfortable," Costa-Marini recalled recently. He was sitting in a spartan office in a bleak suburb of Paris called Courbevoie. "This country just isn't friendly to the entrepreneur. You risk nothing here, or everything. There is little middle ground."

Yet even when he was working at Bull, now France's largest state-owned computer company, Costa-Marini's restless imagination set him apart from the staid, gray corps of his fellow executives. By 1968, it became clear that there was a growing market for computer services. Bull set up a new time-sharing division, and Costa-Marini was put in charge of it. Six years later, the division was reaping more than 25% pretax profits on some $25 million in annual sales.

In the United States, a success like that would probably have earned its author rapid promotion -- especially if the overall performance of the company was as lamentable as Bull's. In France, however, where Bull is the government's prize entry in the international computer show ring, Costa-Marini's bosses responded with something very like envy. He saw his managers recruited away to more attractive employment within the company and his salespeople restricted in scope.

The memory still galls him: "They sent me a list of people I should call," he remembers, "on the ground that they might be customers for Bull computers. It was like giving us a telephone directory and telling us to make sure not to call on the best customers."

By 1974, Bull's pettifoggery at last drove Costa-Marini to act. He left the company, and set up his own independently managed time-sharing division within a subsidiary of the then publicly owned (that is, not state-owned) Compagnie Generale d'Electricite. At the time the division was established, he and CGE agreed that once the venture became profitable, it would become an independent spin-off. By 1978, the new operation was clearly bound for success -- whereupon CGE was unwilling to comply with the original terms of the agreement. Costa-Marini threatened to sue, and the CGE subsidiary, rather than risk an embarrassing court battle, let him go. He took with him five employees and four customers. "I made a little blackmail on them," he said, "but they were probably happy to get rid of us."

The outcome of Costa-Marini's "little blackmail" was a company called Line Data. Today, six years later, having grown at an annual rate of 50%, Line Data now has 43 full-time employees and more than 40 customers (including such major banks as Barclays Bank, Chemical Bank, and Bank of America). Last year it posted sales of some $3 million.

Such entrepreneurial triumphs as Line Data's are extremely rare in France today. Much of the problem has to do with the fact that the independent-minded French businessperson usually operates within rather small horizons. "Attitudes are totally different there," observes Jean Deleage, a former French government civil servant and now a respected venture capitalist in San Francisco "The small businessman just wants to live well on his business with his family. He doesn't want to be diluted, only to pass on the business to his children."

Government policy has done nothing to change this cultural drift. On the contrary, while leaving individual Frenchmen to pursue their petites affaires, French governments have pursued centralization and cartelization, grandes affaires, which would redound to the greater glory of the nation.

It has mattered little whether the government in Paris was monarchical or republican, left, right, or center: The French state has almost always taken a very visible hand in fostering economic development. This is the policy known as dirigisme, and its principal lever has been finance. Since World War II, the proportion of bank assets under government control has risen from 40% to more than 80%. In the 1950s and '60s, the chief beneficiaries of this largesse were smokestack industries, such as steel and automobiles.

More recently, the emphasis has shifted to such high-technology fields as nuclear power, aerospace, and rail transportation. Conservative governments, from Charles de Gaulle's to Georges Pompidou's to Valery Giscard d'Estaing's, have been especially bemused by bigness. "Europeans were impressed with the IBMs, GMs, and ESSOs of America," one high-ranking civil servant recalls. "You drew the conclusions that you needed giants" to compete in world markets. Tax measures encouraged larger companies to acquire small companies. As the civil servant put it, "If no giant existed, we created one."

The results have been mixed. The steel industry is still suffering from overcapacity and inefficiency; Renault has gone through some hard times. On the other hand, the French nuclear-power and rail-transport industries are among the best in the world. Francois Mitterrand's Socialists saw these successes as a buttress to traditional socialist policies when they took power in 1981. Virtually all the remaining private banks were nationalized, and the state increased its share of direct control over industrial output from 18% to nearly one-third.

Whatever the merits of this policy in nuclear power and aerospace, which require long-term capital investments, it has proved utterly disastrous in the tumultuous marketplace of computer technology, in which the state now controls 70% of the French companies in the industry. France, a nation that once bowed to none in engineering and mathematics, is an infant in the field of microelectronics. In 1982, it was obliged to import 85% of its semiconductor circuits, and in 1983 its negative balance of trade in the field of electronics came to $1.5 billion.

Stephen Cohen, a California-based consultant to the last three French governments, thinks that the state of computer technology today is simply too disorderly to be grasped by the French administrative mentality. "The thing that's great about a place like Silicon Valley is the chaos of it all," says Cohen, co-director of the Berkeley Roundtable on the International Economy at the University of California at Berkeley. "The French administrative mind would hate Silicon Valley."

The enarchie (those trained at the elite Ecole Nationale d'Administration) also seems to have misunderstood the effects of government subsidies -- cheap credit and captive government contracts -- on its corporate beneficiaries, especially in high technology. Conservatives typically blame the 1981 nationalizations for France's weakness in this crucial field, but industry veterans like Costa-Marini believe that the problem began long before the Socialists came to power.

"The conservatives were also Socialist -- without the Marxism maybe, but they still believed in central direction," he says. "At Bull back in the 1960s, there was a sense of waiting for the government to do things. People get raised to feel that the government will help them if they're in trouble. It becomes part of their behavior. They become totally dependent."

Jean Auricoste witnessed with his own eyes the price that can be paid for that sort of dependence. Auricoste returned to France in 1955 with a master's degree in electrical engineering from Massachusetts Institute of Technology to help build a computer industry for his homeland. In the early 1960s, as an executive vice-president for engineering at Compagnie Internationale pour l'Informatique (CII), Auricoste helped develop a computer that was roughly equivalent to the new minicomputers that were being unveiled in America by Digital Equipment Corp. Commercially, the "mini" concept was very promising, and Auricoste and other top officials at CII (now part of the government-owned Bull) wanted to push it. But the government had quite different ideas.

In 1966, Washington turned down France's application to buy a large Control Data mainframe for its aggressive atomic energy program. De Gaulle reacted with indignation and immediately ordered CII to defend the nation's honor by developing a purely French-made mainframe. CII, spurred by government grants, shifted its emphasis to concentrate on the technically advanced but unprofitable mainframe. "We couldn't have ever been a European IBM," Auricoste reminisced recently, "but we could have tried to be a European DEC. Unfortunately, we had 'help' from the French government."

IBM Corp. looms large in the high-tech bureaus of the French "adminstrative mind," as large as it does in American high-tech circles. Big Blue's French subsidary is still the nation's dominant and most profitable computer company. But while America's computer industry has responded to the giant by persistently probing it for weaknesses, or by riding its back, French government and corporate leaders typically have responded by trying to beat IBM at its own game. They have had no interest in exploiting new technologies to find the sort of market niches where companies like Apple Computer Inc. and DEC have prospered. Costa-Marini recalls that IBM was almost an obsession at Bull. "If someone had a good idea, they'd ask, 'If it's such a good idea, why hasn't IBM done it?' And if IBM hadn't done it, then how could it be a good idea? I always figured that maybe it would be better to do what IBM wasn't doing, since once they decided to go into something, it was already too late."

Auricoste went on to found Eurosoft, one of the continent's leading software companies. That was his way around the enarchie. Costa-Marini's, of course, was service. Not only are both markets free of direct government direction, they also are far less capital-intensive than the hardware sector. This last is a vital consideration in all French entrepreneurship: Outside capital is virtually impossible to come by.

Costa-Marini's Line Data was started with only a few thousand dollars in capital. The company's greatest assets were its founder's knowledge of the field and his close relationship with a handful of large customers, developed during his long years with Bull and CGE. The company's strategy combines a traditionally French penchant for excellence in engineering with an American-style emphasis on marketing, service, and sales. Sales was the most novel feature of this mix, at least in the French context. "Being a salesman isn't a prestige activity here," he explains. "In the States, the salesmen make a lot of money and have influence, but in France everyone wants to be an engineer or a civil servant. The salesman is usually the person who can't do anything else, and he doesn't get the respect."

At Line Data, on the contrary, salesmen get not only respect but also sizable commissions, often as high as 30% of their salaries. High commissions, Costa-Marini believes, helps keep his sales force free of the nine-to-five reflexes of the larger companies. "Our people are very dedicated," Costa-Marini said not long ago, as he guided a visitor through his busy headquarters. "If a customer calls at 7 p.m, he knows that one of our people will be there."

Line Data's time-sharing service business still accounts for 65% of sales, but Costa-Marini expects that the company's growth area will be in sales of packaged software to the banking industry. As computers become more commonplace, with the increasing use of microcomputers, companies like his won't be able to rely on providing custom services to a few large companies. The challenge will be to transform Line Data into an "industrial" company capable of selling software packages to an expanding, diverse customer base.

Jean Auricoste would agree, with qualifications. He has been in the software field since 1970, when he founded Eurosoft with $50,000 of his own and friends' money. "Our software firms have been body shops so far," he says, "fixing customer's problems one by one. We have to develop our own techniques and products. Microelectronics is making it easier to manufacture hardware. You don't need the oscilloscopes and soldering irons if you have a good CAD/CAM system. With that, we can do dedicated pieces of hardware."

True to his own analysis, Auricoste has moved beyond software applications, which accounted for more than half of Eurosoft's last year's pretax revenues of roughly $15 million, into such new areas as robotics, microcomputer software, and dedicated micro systems.

At the moment, software sales and service are the one flourishing oasis in the desert of French microelectronics. France leads the field in Europe, and, last year, despite a recession that crippled many industries, companies like Line Data and Eurosoft showed strong growth. The future, however, presents a somewhat less agreeable prospect -- at least in the rapidly expanding market for personal computer software. Here, sooner or later, France will have to suffer for the profound weaknesses of its hardware industry. Philippe Sahut d'lzarn, formerly with the Ministry of Research and Industry in the Giscard administration and now an adviser on high technology to Banque Nationale de Paris, puts the problem this way: "Our software firms have been built without external capital, with a minimum of risk. They tend to rely on large clients who pay every month. With the coming of the microcomputer, however, there will be a much larger customer base, and more risk. Software is now becoming more of an industry than a service, and the French firms are slow to react to this. That's why most of the software for micros here will be American."

So, very likely, will the micros. Bull's follow-the-leader strategy has left it hopelessly behind not only the leader, IBM, but the smaller U.S. and Japanese companies as well. In the private entrepreneurial sector there is a company, Leanord of Lille, that manufactures a line of microcomputers. But the recent history of this company has not been encouraging. Unable to raise growth-capital from France's conservative banks and stock market, Leanord's president and founder Bernard Pronier took the desperate step, in 1976, of selling his company to Creusot-Loire, a major manufacturer of heavy industrial machinery and specialty steel. His intention, of course, was to tap the larger company's capital resources. The scheme backfired The computer company did just fine: Last year's $10 million in sales showed a substantial profit. The parent company, however, has been nearly pauperized by the recession, and it is now less of a resource than a liability for Leanord.

"They are restricting our growth, so I am trying to find 10 to 15 people to invest and let us go independent again," Pronier said recently in Lille, an industrial city near Belgium. "If this were America, I believe that with our product we would have stayed independent, gone public by now, and be exporting around the world. But we just can't get the money."

Far from despairing, however, Proniet has taken heart from recent moves by the Mitterrand government to encourage the entrepreneurial sector. The Ministry of Education, for example, has ordered nearly 4,000 Leanord computers -- a dramatic departure from the usual procedure of favoring government-owned companies like Bull. This must be personally gratifying to Pronier; but Socialist policy has aimed at structural changes, too. Bernard Attali, a leading Socialist intellectual and head of France's powerful regional development agency, DATAR, said not long ago: "We face a crisis now in France -- 9% unemployment and a lagging growth rate -- and we are very pragmatic to find solutions. For years we have lived in an environment in which all decisions were made in Paris. We put too much stress on large companies. But France is a diverse nation of thousands of smaller businesses, and we want to give them a responsibility."

In line with this aim, the government has taken remarkably bold steps to free a once dismally timid venture capital market. It has fostered a secondary market in over-the-counter stocks. It has gone so far as to grant three-year corporate tax exemptions for newly formed independent companies. Even the nationalizations have been useful here: With fewer grandes affaires available, investors have been obliged to look elsewhere for new growth companies on which to stake their money. lean Fonteneau, vice-president of international operations at Soffinova, France's oldest and biggest venture capital firm, has said, "I think this government is sincere in wanting to help the small companies." It seems that other investors would agree. The annual flow of venture capital has doubled over the past three years to $37.5 million.

But if the Socialists' intentions for the entrepreneurial sector are good, some of their practices could be decidedly counterproductive. This past summer Mitterrand promised to reduce taxes, individual and corporate, beginning in 1985, from 44.8% of the national income eventually to 42%. He has also begun to show signs of a "free marketer's resistence to bailing out some of the country's sick giants. Nevertheless, in recent years, marginal taxes for people in the upper-middle-income brackets have been increased from 60% to 65%, with a 5% or 8% surcharge (depending on the amount of taxes due) added on for good measure. There is also a stiff new wealth tax on people with assets exceeding $400,000. Many American entrepreneurs would argue that this is hardly the way to seduce Frenchmen to emulate the exploits of a William Gates of Microsoft Corp. or a Steven Jobs of Apple Computer.

Socialist intellectuals like Attali naturally take a different view. "I strongly believe that there are many kinds of motivation besides money," Attali says, rather hotly. "The sense of belonging to an enterprise, the challenge of success. There are. such things as pride and honor that motivate men."

In men like Pierre Costa-Marini, however, such appeals to la gloire are no more heart-stirring than offers of government help. "We don't even want the government to encourage us," the 55-year-old businessman said, as night fell over Courbevoie. "To me, getting money from the government is like getting money from the devil. Once you start begging, you're always begging. All we want is to be left alone. We want freedom -- it boils down to this -- to do our business."