While contemplating the strange mess that Hewlett-Packard's board of directors has gotten itself into, one of my FORTUNE colleagues asked last week, "Why do companies have boards of directors, anyway?"

Why indeed? I did a little checking (starting with the strenuous act of going to SSRN and searching on "corporate board history"), and discovered that legal scholars have been asking themselves the same question for a while now, and struggling to come up with a convincing answer.

The problem is that the historical purposes for which boards evolved no longer apply, and the task they are now expected to fill--that of vigilant monitors of CEO pay and performance--is one for which they are ill-suited. So why don't we abolish the institution? Perhaps because the two main alternatives just seem so weird. More on that later. But first, a little history.

In a fascinating if very long 2004 article (available as a pdf here) in the Hofstra Law Review, Franklin A. Gevurtz of the University of the Pacific's McGeorge School of Law locates modern boards' historical roots in medieval town councils and guild leaderships (whose memberships often overlapped). The guild boards were mainly occupied with resolving disputes between members--an entirely different function from that of today's boards. But as the guilds evolved into companies of independent merchants, the merchant companies into trading companies in which the merchant-investors pooled their money to outfit ships to sail to faraway ports, and the trading companies into modern corporations, the institution of the board came along for the ride.

Still, as recently as the early 1900s, the board had a pretty clear function. It was the perch from which big shareholders and creditors watched carefully over the men they had hired to manage their companies (as is true today at companies controlled by private equity firms). But the very success of some of these pioneers of industrial capitalism led to the undoing of this model. Corporations outlived their founding shareholders, outgrew the need to borrow money, and, as the stock market captured the public imagination in the 1920s, found their shares in the hands of thousands of small investors in no position to watch carefully over anything. Managers naturally took charge, and boards became appendages entirely beholden to them.

That's how things stayed through the 1960s, and on the whole you'd have to say it worked okay. CEOs decided what to do, boards rubber-stamped their decisions, and the U.S. mostly prospered. In the 1970s, though, the world changed dramatically. Corporate strategies that had worked well in a time of steady growth, low energy prices, and the absence of foreign competition began to founder. The country could probably have used a few strong boards unafraid to change course and throw out a CEO, but those didn't exist. What we got instead was the phenomenon of corporate raiders, junk-bond-funded upstarts who bought floundering companies out from under their managers. In many cases the raiders did a better job of running the companies, in others they drove them into bankruptcy. And in response, corporate executives successfully lobbied state legislatures to make hostile takeovers much harder.

In the meantime, stock ownership was becoming reconcentrated in the hands of huge pension funds and mutual funds whose managers were far less patient and far more able to make their impatience known than small individual investors were. For various reasons, though, they were not willing to sit on corporate boards. So what we got was the modern board, whose members are expected to hire and fire CEOs and monitor their performance on behalf of shareholders, but are themselves mostly part-timers with insignificant ownership stakes who remain more or less beholden to the CEO for their jobs. (The formal legal requirements of the modern board evolved a bit differently, as you can read here, but ended up in pretty much the same place.)

Under pressure from shareholders, some of these boards have thrown out CEOs, perhaps most notably at General Motors in 1993 and HP last year. But there have been far more cases of boards failing to rein in out-of-control chief executives (think Enron, Worldcom, HealthSouth), and a zillion Sarbanes-Oxley Acts won't change that. The incentives simply aren't there for board members to play the role of monitor and do it consistently well.

This is not to say there aren't lots of sage and diligent board members who add value by virtue of the advice and guidance they give. But a CEO who is willing to listen to the sage advice of a board member would presumably be just as willing to listen to the same wise person hired as a consultant. The institution of the board doesn't matter.

So what are the alternatives? Well, there's direct democracy--putting all major corporate decisions, including the choice of a new CEO, to a shareholder vote. And then there's absolute corporate monarchy--allowing management to make all the decisions without oversight. Law professor Gevurtz sees strengths in both. Shareholders "know as much as boards do," he says. On the other hand there are lots of successful investment vehicles (hedge funds, for example) that aren't governed by boards or subject to shareholder votes. Unhappy shareholders can simply vote with their feet.

However, even after observing in his article that boards have a lot in common with tonsils ("a largely useless, if mostly harmless, institution"), Gevurtz shies away from recommending their abolition. Why? Because direct shareholder democracy isn't going to happen (the nation's CEOs would be violently opposed) and corporate monarchy, while perhaps a more honest representation of the true power relationship in corporations, just wouldn't feel right. "If you had self-perpetuating management and no representation, people would question why that was," he says. "There's this feeling that managers shouldn't be in charge of these large aggregations of power, even though that's the way it is." So that's why companies have boards of directors: To make the rest of us accept their existence.

UPDATE: UCLA law professor Stephen Bainbridge has written a detailed analysis of why I'm full of baloney. I will offer up an explanation next week of why there's more to my argument than just processed meat product, but in the meantime check out Bainbridge's cool wine blog. My response is here.