here are hopeful glimmers that the long-anticipated yet elusive recovery in the information technology business may finally be coming into view. Besides the recent surge in technology shares on Wall Street are more substantive signs like the fact that orders for many computer products and semiconductors are stabilizing and even gradually gaining ground.

Yet many skeptics remain. And even if a recovery is indeed getting under way, industry analysts say, it promises to be a pretty sober affair — slow and uneven. Instead of leading the economy, as it did in the 1990's, the technology sector is being pulled along with it, behaving more like other industries and less like a business that operates by its own "new economy" rules.

And so it may be that Wall Street has gotten ahead of itself, acting as if an old-fashioned tech rebound were under way. Stock prices on the Nasdaq, where many technology companies are listed, are up roughly 30 percent since March 11. And the Standard & Poor's 500 Information Technology Index is up about 50 percent since hitting a recent low last October.

Perhaps the most significant reason to expect only a measured, plodding recovery is the changed attitudes of, and new pressures on, the corporate buyers of information technology — computer hardware, software and services — whose purchasing habits largely control the industry's fate.

Long gone is the irrational optimism of the 90's and the notion that technology alone can transform a business. Today, corporate executives regard technology as simply a tool — though a crucial one, if used wisely. But it is also a costly tool: Information technology accounts for nearly 60 percent of all business equipment investment. So there is plenty of incentive to restrain spending.

Technology is still a big corporate expense, but technology budgets have flattened out, if not fallen, at most companies. That is a very different world for senior corporate technology executives, often called chief information officers, or C.I.O.'s, who for years became accustomed to having their budgets rise by 10 to 20 percent annually.

"C.I.O.'s, after years of being asked to do more with more, are being asked in a serious way to do more with less," said Richard Hunter, a research fellow at Gartner Inc., a technology research firm.

When it surveyed more than 600 C.I.O.'s worldwide earlier this year, Gartner found that their top three goals, in order of importance, were cutting costs, shoring up information security and supporting innovation. "It shows you how much pressure C.I.O.'s are under, when cutting costs is No. 1 and supporting innovation is No. 3," Mr. Hunter observed.

Corporate spending plans remain crimped. According to a survey in May by CIO Magazine, corporate technology executives said on average that they expected their purchases of information technology to increase by just 3.3 percent over the next 12 months. Back in 2000, technology executives projected increases of 15 to 20 percent in the magazine's monthly surveys.

Even technology vendors that have thrived through the downturn, like Microsoft, acknowledge the shift in attitudes among corporate customers. In his yearly status report e-mailed to Microsoft's 54,000 employees earlier this month, Steven A. Ballmer, the chief executive, noted, "As I talk to business customers, there is less passion and enthusiasm for technology, and greater focus on doing more with less."

And at a time when the faith in information technology may be faltering, The Harvard Business Review has published an article provocatively titled, "IT Doesn't Matter." The essay, written by Nicholas G. Carr, the former editor at large of The Business Review, struck many industry executives as heresy when it appeared last month. It argued that as computer technology becomes more standardized, businesses will have a harder time gaining a competitive edge over their rivals through technology investments.

The article has been widely read and discussed, and The Business Review's mailbag has been filled with dissenting views (several are published online, www.hbr.org, and a few will be published in the July issue of the magazine).

But even the essay's critics scarcely make a case for a new technology boom. Paul A. Strassmann, 74, is an elder statesman of the C.I.O. fraternity, having held that title at General Foods, Kraft, Xerox and the Department of Defense, and he remains an executive adviser to NASA. Mr. Strassmann dismisses the Business Review article as "absolute poppycock" because of the "enormous unrealized potential of information technology."

The potential is unrealized, according to Mr. Strassmann, mainly because of two decades' worth of mindless technology spending through the 1980's and 1990's. The spending spree, combined with exponential improvements in processing speeds and data storage capacity, has meant companies are often overwhelmed by computing firepower they cannot use efficiently. "It was not about economics or productivity," he said. "It was a technology arms race."

Mr. Strassmann's prescription? "It's better to slow down and master the thing," he said. "That's the nub of the problem."

Anthony E. Scott, the chief of technology for information services at General Motors, also takes issue with Mr. Carr's article. "He's right that large parts of the industry are maturing," Mr. Scott said. "But there is still lots of room for innovation and to gain competitive advantage from information technology."

Mr. Scott is part of the technology team at General Motors that years ago began paring costs in ways that are now popular in corporate America, like reducing the number of hardware and software products used and consolidating data processing into fewer, larger computer centers. In 1996, G.M. spent $4 billion on information technology. Its budget for this year is just under $3 billion, and its plan calls for further cuts next year.

Still, G.M. is investing in software applications and systems to increase its use of the Internet and the Web to communicate more closely with suppliers, dealers, industry partners and customers. "Despite the dot-bomb and a difficult economy," Mr. Scott said, "that Web-based work to expand our external connections continues unabated."

Companies like Capital One, a large bank and credit card marketer that is still expanding its information technology spending, are the exceptions among technology buyers. Capital One is growing strongly and its technology spending will rise 10 to 20 percent this year, said Gregor Bailar, the company's chief information officer.

Capital One is renowned for its use of data mining and information analysis techniques to find and reach particularly profitable slices of the credit card market. The company, Mr. Bailar noted, has an "information-based strategy" in which technology and business considerations are deeply intertwined.

Like many C.I.O.'s, Mr. Bailar has read the recent Harvard Business Review article. And he said that Mr. Carr, by writing it, had performed "a great service" by stimulating debate about the role of information technology. But Mr. Bailar said the article's flaw was in discussing the maturity of information technology in terms of industrial technologies like railroads, steam and the telegraph. These, he noted, were single-purpose technologies, whereas the computer is a programmable, general-purpose technology with all but infinite possibilities.

The Business Review article, Mr. Bailar added, tended to treat computer hardware and software as if mere transport vehicles for bits — short for binary digits — the lingua franca of information technology. "The competitive advantage you gain from using information technology is not based on storing or moving bits around," he said, "but based on what you do with them."

Besides, technology executives say, the shift to industry-standard microprocessors and software building blocks does not mean a leveling of the opportunities for using technology to gain an edge over competitors.

Dell Computer, for example, has used industry-standard technology to run its business. "But Dell has used information technology to gain an extraordinary competitive advantage in the personal computer business, becoming superefficient because of its supply-chain system," said Charles Fitzgerald, a strategy executive at Microsoft.

Mr. Carr, the author of the Harvard Business Review article, acknowledged that technology, used cleverly, can still give companies an edge in the marketplace. "But my point is that the layer of technology that is customizable, and therefore can give a company a competitive advantage, is getting thinner and thinner," he said. "And technology-based competitive gains won't last as long in the future. I still think my conclusions are true."

Even a success story, like Dell, cuts both ways in the context of looking for a technology recovery. The company's strong growth has come largely through taking sales from its rivals — not because of solid growth in the personal computer industry in general.

The semiconductor industry, meanwhile, does appear to have begun a gradual recovery. Earlier this month, the chip industry's two largest contract manufacturers, Taiwan Semiconductor Manufacturing and United Microelectronics, reported that their monthly sales in May were the strongest in more than two years. And the Semiconductor Industry Association, in its most recent forecast, estimated that this year should be one of recovery, with sales up about 10 percent, compared with last year.

But the semiconductor trade group added that its forecast through 2006 reflected "the new realities of the semiconductor industry," with expectations of average yearly growth of 8 to 10 percent. That compares with an average annual rate of 17 percent going back more than two decades.

"The industry is maturing — that is the view that everyone is coming to share," said Kevin Krewell, a senior editor of Microprocessor Report.

The nature of that maturation is a subject of debate, both in the industry and on Wall Street. The long-term trend, going back to the 1960's, is that spending on information technology grows at two to three times the rate of growth of the economy.

Investors seem to assume that a rebound to that historical rate is certain and probably right around the corner, even though corporate buying remains weak. "Right now there is a disconnect between the run-up in tech stocks and the technology spending expectations of C.I.O.'s," said Edward Yardeni, chief investment strategist at Prudential Securities.

Technology stocks are valued as the corporate residents of a growth industry. The 100 companies in the Merrill Lynch technology index are priced at 29 times analysts' projected earnings for 2004, compared with 16 times for the Standard & Poor's 500 index.

Some analysts say the technology stock investors may be on the right track. "There's a tremendous divergence of opinion and many of my peers think this is a false start," said Richard Whittington, a semiconductor analyst at American Technology Research. "But I think this is going to broaden out, and we're seeing the start of a real recovery."

But Steven Milunovich, an analyst at Merrill Lynch, argued that the current rally might be short-lived without more solid evidence of stronger technology spending. Asked to explain the recent market rally, Mr. Milunovich observed, "There is still a bit of romance about technology."