Forgive me for not joining in the general huzzahs, but having lived through the ill-fated ““financial supermarket’’ and ““get as big as you can’’ manias of the 1980s, I have a funny feeling that the new Morgan Stanley, Dean Witter, Discover & Co. isn’t going to work out as planned. The name alone is enough to give you pause–imagine having to say all seven words in one breath when you answer the phone. The joke going around Wall Street is that the new firm, which combines snooty Morgan Stanley with Middle American Dean Witter, should really be called ““White Shoes & White Socks.’’ This is a play on Dean Witter’s adventure of 15 years ago, when Sears, Roebuck bought it and set up Dean Witter offices in Sears stores. Wall Street wags called this idea ““stocks and socks,’’ and it was a resounding flopperoo.

The folks running these companies say, in effect, that was then and this is now. Their rationale for combining Morgan Stanley and Dean Witter sounds great–but then again, it always does. If it didn’t sound great, they wouldn’t be doing it. Morgan Stanley, an investment-banking company that does business all over the world, is supposed to be a natural fit with Dean Witter, whose business is almost entirely U.S.-based. Morgan’s main businesses–trading securities, advising corporations, selling new issues of stocks and bonds–are supposed to complement Dean Witter, which owns the Discover credit card and the nation’s third largest retail stock brokerage. Morgan Stanley will help Dean Witter and Discover go global. Combining the firms will give Morgan Stanley outlets for its underwriting products and give Dean Witter brokers sexy new securities to sell. In the one part that makes clear sense, the firms would combine their mutual-fund businesses, creating one big mediocre fund company from the two existing mediocre fund companies. You can run one big fund business with far fewer people than it takes to run two smaller ones. Downsizing city, here we come.

My prediction of trouble ahead isn’t because of ““clashing corporate cultures,’’ which is a polite way of saying that Morgan Stanley’s high-paid investment-banker types aren’t likely to become bosom buddies with Dean Witter’s lower-paid employees. The culture problem is merely a symptom of the underlying problem. To wit: why do these businesses need to be combined? Dean Witter’s 9,300 stock brokers aren’t going to blindly peddle securities that Morgan Stanley is underwriting, and Morgan Stanley isn’t going to put deals together just to create securities for Dean Witter’s brokers to sell. If a Morgan Stanley deal is good, firms other than Dean Witter will happily take the securities. If the deal’s not good, no one will take it, not even Dean Witter brokers.

History is a problem, too. Except for Merrill Lynch, which has managed to create a formidable investment-banking business largely from scratch, no firm that I know of has ever successfully combined investment-banking and retail-brokerage businesses for the long term. Certainly, no investment bank or brokerage house has successfully bought its way into the other business. And even Merrill hasn’t been tested in a market downturn.

In a world in which information is becoming increasingly available and decentralized, talking about financial supermarkets and one-stop shopping is lots easier than creating and running them. Ditto for going global. Will the Morgan Stanley guys who do deals in Asia succeed in hooking Koreans and Japanese on the Discover card? Will they even care to try? Yes, Morgan Stanley’s mergers-and-acquisitions fees are subject to sharp ups and downs. But when M&A hits a down cycle, the stock market–and thus retail brokerage–often hits a downturn, too.

I’ll be the first to say that I may be wrong here. I tried to talk to the top executives of the companies, but they never made themselves available. I’m going purely on instinct, and everything these two smart, disciplined firms say about their deal sounds wonderful. But beware. If praying mantises could talk, they would probably make a good case for their behavior, too.