(FORTUNE Magazine) – MANY OF THEM didn't know Jerry Junkins. But when the memo went out announcing his promotion, workers at the Dallas headquarters of Texas Instruments privately applauded. To them his elevation to president and chief executive signaled the end of an abrasive management style that many felt had contributed to a fat catalogue of the company's misfortunes. Junkins, 47, seems determined to change that style and is spending much of his time, he says, ''just listening'' to employees. He might get an earful. Texas Instruments is the world's largest producer of semiconductors, which brought in nearly half the company's $5.7 billion in sales and an estimated 80% of its $316 million in profits in calendar 1984. But TI concentrates on commodity chips that are particularly vulnerable to weakness in the market. Today's market isn't just weak -- it's anorectic. New semiconductor orders are running at only 72% of shipments, one of the lowest levels since the industry began keeping records in the mid-1970s. Texas Instruments reported an 89% drop in earnings for the first quarter, to $9.1 million. For the full year, many analysts expect the company to earn only about one- tenth last year's profits. The downturn in semiconductors has been accompanied by other miseries aplenty. Sales at TI's second-biggest group, military electronics, have temporarily flattened, and the business in which TI got its start over 50 years ago -- seismic equipment and services for oil exploration -- is stuck in the oil recession. Consumer electronics has produced some notable failures in digital watches and home computers. The company's foray into personal computers has missed the mark, and last year, according to E.F. Hutton security analyst E. C. White Jr., TI lost about $15 million. The appointment of Jerry R. Junkins may be the best news to come out of Texas Instruments in a long while. An engineer from Iowa who has spent 27 years at TI, Junkins is widely regarded as a sound, likable manager. Paul McCracken, an economist and professor at the University of Michigan, and a TI director, says Junkins was appointed partly because he was a ''people person.'' Junkins says: ''I believe I'm a good listener and I enjoy associating with people. It's a lot of fun, and that's where the work gets done.'' His record as a manager is strong: from 1975 to 1981 he headed TI's military contracting operations, which grew more than 200% under his management. He took over the data systems division four years ago -- in time to kill off TI's flagging home computer. Although the division only broke even last year, Junkins did improve marketing. THE BIG QUESTION is whether Junkins will have authority to implement major changes. Industry analysts speculate that the man in charge remains Mark Shepherd Jr., 62, chairman since 1976. About a year ago Shepherd relinquished the chief executive's title, giving it to then president J. Fred Bucy, whom Junkins replaced. But at the time, Shepherd took the title chief corporate officer and retained control over strategic planning, finance, and communications. Although he handed over those operations to Junkins in May, Adam Cuhney, a security analyst at the investment banking firm of Kidder Peabody, wonders if he will reassert control. ''With Shepherd there,'' says Cuhney, ''all you can do is wait.'' Many former TI employees describe Shepherd and Bucy as meddlesome and hot- tempered. (Both declined to be interviewed for this article.) ''We were working on a computer project, and Fred took some materials home one weekend,'' recalls a former employee. ''When he came back, he wanted to change the name of the project. It was silly. He couldn't see the big picture.'' Bucy immersed himself in the minutiae of operations. One former employee says that Bucy often took home so many interoffice memos that he used a dolly to carry his briefcases, and that every day the company jet ferried memos to him while he was on vacation in South Texas. (A Texas Instruments spokesman denies both incidents.) Shepherd has flown into rages that paralyzed underlings. According to Ron Green, a former TI executive who is now senior vice president of a bank technology consulting firm in Dallas, Shepherd once got so angry that he threw a loose-leaf binder against the boardroom wall. ''Everything fell apart,'' says Green. ''We had orders to find new binders that could survive hitting a wall. I remember seeing people throwing binders at a wall until we found one that wouldn't come apart.'' Former employees say Bucy and Shepherd drove off some of Texas Instruments' best talent. ''Junkins has to put middle management back in TI,'' says a corporate planner who recently quit the company. ''The two years I was there the company had 11% and 14% turnover. Most of that was people with five to 15 years' experience.'' Texas Instruments says its turnover has been about 10%, which it describes as low for the electronics industry. The TI culture has tended to stifle risk- taking. Though Texas Instruments accelerated the computer age by perfecting mass production of transistors 31 years ago, many of its products in the past 15 years have been improvements on old inventions rather than breakthroughs. ''They aren't on the leading edge anymore,'' says a vice president who quit two years ago. TI takes what one former executive calls a ''brute force'' approach to making semiconductors. It spends hundreds of millions of dollars on big new plants and works feverishly to operate them more efficiently than those of competitors, rather than concentrating on complex microprocessors that could give it a proprietary advantage. TI churns out many chips that sell on price alone. That's fine when demand is high, but not when prices crater; chips that went for $4 last year are selling for 75 cents. Junkins concedes that the company was slow in adopting technology called CMOS (complementary metal oxide semiconductors), which uses less power and produces less heat than older technologies. But he says a ''major acceleration'' is under way and notes that Texas Instruments now sells 500 varieties of CMOS chips. Other chipmakers remain unimpressed. ''They're lost,'' says Michael Bolan, director of marketing at Dallas Semiconductor, a small company that makes custom chips. ''They're working on a million-bit memory, but so are five big Japanese companies, all spending hundreds of millions of dollars. The potential for making money is extremely limited.'' Junkins acknowledges that marketing has also been one of TI's notable shortcomings. ''We have to listen to what the customer wants,'' he says. ''As the market evolves, a lot of other people will have good technology. It's important we make a balanced shift to a market-driven customer relationship.'' That's boilerplate stuff at most companies but an important change for Texas Instruments, where engineers, not salesmen, have been ascendant. Selling chips on price engendered a ''take it or leave it'' attitude that infuriated customers. ''Customers hate TI's arrogance,'' says Glen Madland, a management consultant in the semiconductor business. ''One chip buyer was pleading with me to find a company that would replace Texas Instruments for him.'' The home computer operation was a marketing disaster, and shutting it down led to a $145-million loss for the company in 1983. The publicity gave potential buyers of desktop computers, which TI still makes, the impression that the company had abandoned computers altogether. That didn't do much for sales of TI's Professional series of small computers, already sluggish because the machines are not fully compatible with the IBM Personal Computer. One sign that TI's marketers may be gaining influence is the impending debut of the Business-Pro, which is compatible with the popular IBM PC AT and runs software written for older TI machines. But even though it has gotten good reviews, the Business-Pro faces a tough battle for retail shelf space in computer stores. Junkins admits his expectations for it are ''modest.'' The brightest spot for Texas Instruments nowadays is Junkins's old bailiwick of military contracting. The company has particularly high hopes for its HARM missile, which uses hundreds of TI's most sophisticated electronic components to home in on enemy radar installations. TI won a $522- million contract with the U.S. Navy to build 1,571 missiles in the next three years, and the project could bring in $5 billion in revenues over the next five to eight years. HARM is TI's first major venture as a prime military contractor, but the company is likely to push for more government work. Junkins says that military contracts have ''always provided stable growth.'' Defense work is more predictable than commodity semiconductors. TI also seems more comfortable selling to a customer that awards contracts on the basis of technical analyses rather than to fickle consumers. TEXAS INSTRUMENTS has won praise for its work in artificial intelligence, where it has spent hundreds of millions of dollars developing chips and software that will help computers draw conclusions from a constantly updated base of knowledge. ''TI is really first class in artificial intelligence,'' says Egil Juliussen, chairman of Future Computing, a Dallas firm that tracks the industry. ''It's a window that won't open for another year, but they have a shot at a very important market.'' Again, the challenge won't be developing the technology so much as marketing it. Early indications suggest the company doesn't have a clear idea of who its customers are. Joe Harmon, vice president of CompuShop, a Dallas-based chain of retail stores, says TI spent two hours with him discussing artificial intelligence machines. ''When I asked how much one would cost,'' he added, ''they said $50,000. I asked myself why they were talking to me.'' The price vastly exceeds what customers expect to pay at a retail outlet. One of Junkins's biggest sales jobs will be on Wall Street. Investors have grown weary of Texas Instruments' booms and busts. From a high last year of $149.50 a share, TI's stock has dropped to a recent price of $90.50. BOX: INVESTOR'S SNAPSHOT TEXAS INSTRUMENTS SALES (LATEST FOUR QUARTERS) $5.7 BILLION CHANGE FROM YEAR EARLIER UP 20% NET PROFIT $245.3 MILLION CHANGE LOSS YEAR EARLIER RETURN ON COMMON STOCKHOLDERS' EQUITY 16% FIVE-YEAR AVERAGE 9% RECENT SHARE PRICE $93.50 PRICE/EARNINGS MULTIPLE 9 TOTAL RETURN TO INVESTORS (12 MONTHS TO 6/7) -28% PRINCIPAL MARKET NYSE Explanatory notes: page 154