Softwar: An Intimate Portrait of Larry Ellison and Oracle

Collaboration is very nearly the right word, as Ellison reviewed Symonds' manuscript before publication and, while he did not alter it, he did make a large number of comments, which appear in the book as footnotes. As Symonds is a good journalist who attributes most of his material, Ellison is able to take issue immediately with statements other people make about him and his company. The overall effect is hypertextual, and represents an important new biographical technique that other writers should imitate. Softwar succeeds because Ellison has a fantastically interesting life, tremendous experience, and carefully considered opinions, and because Symonds communicates them with clarity and style. --David Wall

Topics covered: The life, times, acquaintances, tastes, toys, and opinions of Larry Ellison, the database entrepreneur and CEO of Oracle Corporation.

Product Description
In a business where great risks, huge fortunes, and even bigger egos are common, Larry Ellison stands out as one of the most outspoken, driven, and daring leaders of the software industry. The company he cofounded and runs, Oracle, is the number one business software company. Perhaps even more than Microsoft's, Oracle's products are essential to today's networked world.

In Softwar, journalist Matthew Symonds gives readers exclusive and intimate insight into both Oracle and the man who made it and runs it. As well as relating the story of Oracle's often bumpy path to industry dominance, Symonds deals with the private side of Ellison's life. With unlimited insider access granted by Ellison himself, Symonds captures the intensity and, some would say, the recklessness that have made Ellison a legend.

With a new and expanded epilogue for the paperback edition that tells the story behind Oracle's epic struggle to win control of PeopleSoft, Softwar is the most complete portrait undertaken of the man and his empire -- a unique and gripping account of both the way the computing industry really works and an extraordinary life.

It appears that Larry Ellison was one of these early programmers, whose maturation is documented in this book. But as with any maturation, it includes the acquisition of blind spots.

For while I in general support Larry's goal of eliminating "islands" within organizations of isolated and contradictory data and code, I am more pessimistic than he as to whether it can be accomplished.

The well-known and by now well-worn theme of Derrida, that of the undecidable gap between writing and speech, means that the ultimate grand vision, of "one" data base, may never be attained.

Larry is right about the Internet: it is the Last Big Thing. This can be proven apriori. For given two or more networks, and given zero cost and high benefit in their connection, whether through a narrow gateway or broadband, then we can say that the two networks "want" to become one network and instantaneously, at warp speed, shall do so. In the late 1980s, several networks operated in academia, government and privately did just this because there is, absent security considerations, a seemingly irresistable craving on the part of networks to join other networks and indeed to become the Internet.

This is the synthetic apriori argument, for both the existence and unity of the Internet as a given. However, and as soon as it is constructed, the reverse, analytic argument against the Internet's usability by the corporation may be constructed, which will return us to Mr. Ellison: for I fail to see how the possibility, of constructing a single logical path to a single data base for the organization, means it can be actualized.

I fail to see this because this has long been an unmet promise of ultimate managerial control within organizations (the "executive dashboard" being one such foolish idea), a control which manages to dismiss the fact that an organization consists of the labor of intelligent beings all the way down...to the person who picks up the trash.

I fail to see this because as a form inescapably of writing, data systems imply their own multiplicity. The "scribe" in all societies develops his own agenda and there is no check on him available to power as such, because power as such relies on the self-interested "scribe" to transmit its will and an almost (but not quite) mathematical problem results in the self-reflexivity.

The crisis is in Mr. Ellison's genuine concern with the way in which data and human intelligence systems failed to predict September 11, a concern which I happen to share. Indeed, I believe that September 11 starkly fulfilled a dismal prophecy of the late hero computer scientist Edsger Dijkstra.

Unlike many highly-placed figures in the computer science establishment, hero computer scientist Dijkstra was concerned, all the way down, about the quality and even the basic correctness of the data systems being designed over his lifetime, and he said at one point that he feared that organizations would collapse under the cumulative overcomplexity of their unmastered data systems. The stark images of a collapsing center for symbol processing on September 11 may be the fulfillment of this prophecy.

One of the FBI field agents assigned to investigation of terrorism prior to September 11, Colleen Rowley, testified before Congress that she did not even have the capability to enter Boolean format queries in the FBI data base, for example of the form "terrorist association and attends flight school".

Of course, Oracle data bases of the sort Larry and his company provide, provide this capability in mass quantities. At the same time, their very complexity (which may be unavoidable) generates scribal bureaucracies which are in both Plato's and Derrida's sense pharmakon, poison and cure, and, in general, the hair of the dog.

It is clear that these sorts of scribal bureaucracies at the FBI felt that some sort of extension or hack to provide rapidly the needed capability at the FBI was a "hard" problem, because these scribal bureaucracies reproduce themselves by insisting that such problems are "hard", and that the CEO is too busy to involve himself with writing...in a stark, if completely unconscious, replication of Plato's account of writing.

The result today is that a great deal of social inequality, created in part by fortune-seeking by the scribal class, means that it's impossible to create a unified written "intelligence" for policy making, and the result is an out of control foreign policy which as I write is creating preconditions for further terrorism.

Symonds breathlessly notes that Larry and his wife are both big fans of Donald Rumsfeld. Bush, and Bush's war, have deep roots in the self-interest of the new, successful American elite.

This elite marched and protested its parent's war in Vietnam, and, Ellison was a supporter of Robert Kennedy's fatal bid for the 1968 presidential nomination. Rumsfeld, for that matter, was an anti-war Republican under Nixon.

However, it appears that Larry may be blind to realities in much the same way that middle-aged managers were blind to the downside of enormous mainframe computing in the early 1970s. He views the future as one of large corporations competing, especially in his own industry, for a diminishing pie.

However, large corporations are composed of intelligent agents, who act from a unique combination of self-interest and complete irrationality, and, just as Ellison's own generation constructed its own reality in the form of microcomputer and micro culture, the next generation may prove him wrong. Or, Dijsktra's prophecy may come true, in which case we'll be busy gathering firewood and not worrying about SQL.