Facebook's expert witness Dr. Saul Greenberg claimed not to know the purpose of basic web programming functions; flip-flopped from "wild guess" to learned opinion
Trial courts have a duty to disqualify unreliable science to prevent jury confusion
Unreliable science was presented by Facebook’s expert witness in Leader Technologies, Inc. v. Facebook, Inc., 08-cv-862-JJF/LPS (D.Del. 2008). This bad science appears to have confused the jury. Indeed, it would confuse experts since it needs a Venn diagram to figure out that the opinions are ambivalent. The court failed to catch it, allowing Facebook's "dark arts" to turn the case toward their early commercial activity claim—a claim added by Facebook only after the close of discovery. The untimely-filed claim prevented Leader from gathering evidence to defend against it. Leader Opening Appeal Brief, p. 9; also available at Leader. With no hard evidence itself of early commercial acitivity, Facebook created more jury confusion with an attack on the credibility of the inventor; conduct better suited to an episode of Law & Order.
A well-known "dark arts" tactic in patent infringement trials is for the infringer to find an expert witness willing to mislead the jury. Since a lay jury cannot assess the reliability of the science presented, it is the duty of the trial court to disqualify unreliable expert testimony. Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 US 579 (Supreme Court 1993) at 595-597 (the trial judge must ensure the reliability of scientific testimony).
"Dark arts" tactics include lying, misleading, bad science, motion practice (delay), twisting words, bad lighting, deception, sandbagging, procedural tricks, fishing expeditions, trick questions, slander, misdirection, innuendo, pontificating, feigning mistakes, fawning, intentional bumbling, fabrications, legalese, obfuscation, dishonesty, manipulation, condescention, destruction of evidence, insipid flattery, making contradictory claims, intentional confusion, pandering, doctored evidence, planting evidence, threats, witness intimidation, trial theater, juror tampering, political manipulation, and late claims.
On the subject of the provisional patent, Facebook's expert witness was Dr. Saul Greenberg, professor of computer science at the University of Calgary. Leader's expert was Dr. James Herbsleb, professor of collaborative science at Carnegie Mellon University. Both men are prolific technology writers and both are recognized in the field of science addressed by Leader's U.S. Patent No. 7,139,761.
Facebook wanted to have its cake and eat it too
Facebook asserted contradictory positions that would certainly confuse a jury. The jury opened the door to Facebook's "on sale bar and public disclosure" claim only after it decided not to recognize Leader's earlier provisional patent application date of Dec. 11, 2002.
Facebook's assertions are illogical. They claimed ambivalently that: (1) the provisional patent did not disclose an invention "ready for patenting" (consistent with its false marking claim) and (2) the technology did contain the invention, was "ready for patenting," had been offered for sale, and had been disclosed in element-by-element detail to alleged offerees prematurely (antithetical to its false marking claim). Tellingly, no offeree testimony, no source code, no technical documents, no expert witness testimony was presented. See No evidence. No problem. Fabricate it.
Put in layman's terms, Facebook accused Leader of disclosing all the elements of the invention to sales prospects, while simultaneously not disclosing all those same elements to the Patent Office where its rights could be protected. Illogical positions are always confusing until the aim function becomes clear.
Department of the "Dark Arts"
Facebook's "dark arts" department apparently figured it could not lose by asserting contradictory positions on the theory that the jury would be confused and likely let one of them through. In other words, Facebook wanted to have its cake and eat it too, and the jury obliged.
Even so, Leader argues that Facebook failed to present any evidence of this alleged early commercial activity. Nevertheless, alleged early commercial activity is not the focus of this blog posting. This posting focuses on Facebook's allegation that the provisional patent did not disclose a "tracking component." Dr. Herbsleb said that it did and gave multiple examples, including a blind-study write-up by one of his students showing that he understood that a tracking component was indeed disclosed by the provisional application.
Web browsing cannot work without at least basic user tracking. When one goes to a website, a special program tracks one's movement. Otherwise, one could not navigate from web page to web page. Programmers call your visit to a website a "session" and your location at any given time while on that site a "state." Every beginning web programmer learns this. Dr. Herbsleb Testimony, Trial Tr. 11326:7-8 ("It is well-known that this is the purpose of session state libraries").
First answer: "Wild guess"
On the other hand, Dr. Greenberg said he would be making a "wild guess" as to the purpose of the Java "import statements" in the provisional application. Trial Tr. 10903:10; See Leader Provisional Patent, p. 11. (Java is a popular web programming language developed by Sun Microsystems.)
Second answer: Dr. Greenberg described the contents of the "wild guess" box that he had just said was empty
However, just a few minutes later he contradicted that statement by "using my knowledge of programming" to describe the detailed actions of one of those import statements. In other words, Dr. Greenberg could later understand what he could not earlier understand. Trial Tr. 10904:8-10905:15.
For example, Dr. Greenberg provided detailed opinion about a generally understood element of the "actions" import statement, namely "action.addactionlistener," after having just finished saying he had no idea what could be in any of the import statements. This is not reliable scientific opinion. Tellingly, he was not similarly able to explain the generally understood user tracking purpose of a "sessionstate" import statement, and that code in the provisional application like "requestState.getCurrentUser( ).getId() );" would necessarily be tracking user movement. In fact, he went out of his way to say a flat "No" when asked if the provisional patent discussed user tracking. Id., Tr. 10905:16-24. Even an ordinary layman can read "CurrentUser" and "getId" and is about a user. And, an ordinary Java programmer knows that "State" tracks a user's movement from web page to web page. Id. 10903:14-10904:15; Leader Provisional Patent, p. 11-19.
At best Dr. Greenberg's answer is ambiguous since one cannot say credibly that one doesn't know what is inside a box, then proceed to describe its contents. Worse, Dr. Greenberg is obfuscating; counting on the jury being confused by the technical jargon. At worst, he is lying. None of these possibilities serves the purpose of an expert witness to assist the finders of fact in determining the truth.
The Federal Rules of Evidence 702, Testimony by Experts, states:
"If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case.""The sole justification and purpose of expert testimony is to assist the trier of fact to find a solid path through an unfamiliar and esoteric field." Thompson v. Merrill Dow Pharmaceutical at 241.
Dr. Greenberg failed to use reliable principles, and having failed there, did not apply reliable principles to the facts, and therefore failed to assist the trier of fact.
Exclude expert testimony that is not credible
The Federal Circuit Court of Appeals requires a trial court to exclude expert testimony that is not credible. Christophersen v. Allied-Signal Corp. at 1127 ("If the record establishes a critical fact contrary to the expert's testimony, or if a court may take judicial notice of a fact that fatally contradicts the assumptions of an expert, then his or her testimony ought to be excluded").
Mathematically speaking, Dr. Greenberg violated the principle of bivalence— stating two contradictory truth values when only one can be true. Dr. Greenberg is not permitted to first say a box is empty, and then proceed to describe its contents. That is like saying the lake is empty, but my swim in it was refreshing. A reasonable person knows that one cannot swim in an empty lake.
An opinion founded upon a wild guess is not credible, or reliable; Dr. Greenberg's testimony confused the jury
Dr. Greenberg's expert testimony confused the jury, who as a result invalidated Leader's earlier provisional priority date. Had that date been affirmed, then the entire early commercial activity claim would have been moot.
No requirement that a provisional application must "work"
Dr. Greenberg's testimony contains numerous misconstructions of the law with regard to the disclosure requirements of a provisional patent. For example:
DR. GREENBERG: "First, I have to say I don’t know if the code exists. I can't tell if this code is working code. Is it actually code that they've actual compiled to run? I don’t know." Trial Tr. 10902:13-16.The law makes no requirement that software provisional patents contain "working code." In fact, the law asks the patentee to exclude those elements that are well known in the art. Therefore, by the very nature of the U.S. Patent Office ("USPTO") instructions, no disclosure will actually "work" as Dr. Greenberg implies that it should. This would certainly confuse the jury. It is not Dr. Greenberg's role to interpret law to the jury. The trial court erred in not correcting this legal error. See MPEP 2164.01 ("A patent need not teach, and preferably omits, what is well known in the art").
Therefore, contrary to Dr. Greenberg's erroneous implications about the legal requirement, Leader had no requirement in law to disclose the general nature of tracking user movement on a website, and the court failed to tell this to the jury. Leader's disclosure duty was to disclose the innovative element(s) of what is being tracked, which Leader did disclose, namely "webs" and associated multiple "contexts," among other things. See Leader Provisional Patent; See also full trial record on infingement where there was a robust "battle of experts."
USPTO instructs inventors not to disclose menial tools
DR. GREENBERG: "I can’t tell is whether these files com.leader.util or debug [or "actions" or "sessionstate"], whether they exist or not. I have no idea whether these are just place holders or if they have stuff there." Id., Tr. 10902:44-10903:4.Since the USPTO asks inventors not to submit elements of an invention that are well known in the art, programming elements like "debug," "util," "actions" and "sessionstate" should not be included in the disclosure of a web application. See MPEP 2164.01 above; See also Matter of Application of Sherwood at 817 817 ("If these bridge-gapping tools are disclosed, there would seem to be no cogent reason to require disclosure of the menial tools known to all who practice this art").
DR. GREENBERG: "the actual stuff in these things designated by the import isn’t there. They did not deliver that. I've read other patent applications, other things, before and sometimes they come with a floppy or CD that says, here’s our stuff. For one, this is all I have to work with. I would be guessing." Id., Tr. 10903:14-22.
DR. GREENBERG: "Com.leader.util, maybe that means there's a utility program in it, but there’s another one called asp.facebook.util [sic], so I don’t know what's in it. I just make a wild guess." Id., Tr. 10903: 6-10.Programmers of ordinary skill know the general purpose of web-based utility import statements. Therefore, this statement is not credible. Dr. Greenberg's public writings reveal that he knows that "session state" tracks a user. Indeed, the entire computer industry knows this.
Dr. Greenberg knew the purpose of "sessionstate"
Dr. Greenberg's public writings show that he is being disingenuous and in fact knows the purpose of common import statements. For example, he wrote about "sessions" and "tracking over pages" in 1999 in a National Institute of Standards paper titled "Getting Back to Back: Alternative Behaviors for a Web Browser Back Button." He also recommended a Java program named ArgoUML (See also ArgoUML JavaDocs) in a Western Computer Graphics Symposium paper in 2000 that specifically describes a "buildTransition" web session program as "Build a transition between a source state and a target state" (emphasis added). One is hard-pressed not to understand that this describes user movement from a source web page to a target web page.
The computer industry knows the purpose of "session state"
The prestigious SANS Institute published a presentation in 2002 titled “State maintenance in Web applications” which says "State maintenance (also known as session maintenance, or session state) is a familiar issue for Web programmers" (slide 3) (emphasis added). Sun Microsystems, the developer of Java, stated in "Maintaining Client State" in 2002 that "the applications are responsible for maintaining such state, called a session" (emphasis added). The Microsoft Computer Dictionary defines "session state" saying "In the Java programming language… a session bean. . . can. . . maintain conversational state across methods and transactions. . . [and] manages this state."
One cannot have knowledge of something for which one has previously abjured knowledge; Dr. Greenberg transformed a "wild guess" into a detailed explanation "using my knowledge of programming"
Dr. Greenberg should not have been permitted to confuse the jury with bad science by first saying he couldn't give any opinion about the Java import statements in the provisional patent, then proceed to give an opinion on the parts of the disclosed code that would support his storyline.
MS. KEEFE: "I'd like to point your attention to the middle of the page where it says action.addactionlistener. Do you see that code?"The "dark arts" smoking gun "No"
DR. GREENBERG: "I do."
MS. KEEFE: "What does that code do?"
DR. GREENBERG: ". . . I believe in looking at this and using my knowledge of programming that what this essentially does is really [sic] the user interface part for somebody to manually set the relationship of one board to another. [Dr. Greenberg then proceeds to give quite a detailed assessment of what the code does]." Id., Tr.10904:8-10905:15.
MS. KEEFE: "Does anything in this disclosure [disclose] tracking a user’s movement from one board to another board?"This Dr. Greenberg testimony was the only testimony at trial that alleged that the Leader provisional patent did not disclosure a tracking component.
DR. GREENBERG: "Neither is it in this code and nowhere else in the code."
MS. KEEFE: "Does anything in this code disclose tracking a user’s movement from one context to a separate context?"
DR. GREENBERG: "No." Id., Tr. 10905:16-24
Trial court error—Dr. Greenberg’s testimony should have been disqualified as a matter of law—"Dark Arts" does not equal good science
The moment that Dr. Greenberg provided bad science the trial court should have disqualified his testimony. In effect, Dr. Greenberg proffered a detailed opinion about a "wild guess." This is not credible science. Juries must be able to rely upon the veracity of experts so that they can weigh the facts based upon credible scientific guidance from the expert witnesses.
Facebook's "Dark Arts" Perfected—MS. KEEFE: "Does anything in this disclosure [disclose] tracking a user’s movement;" DR. GREENBERG "No"
That was Dr. Greenberg's testimony. However, Dr. Herbsleb provided substantial testimony about what one of ordinary skill knew about the tracking component from the provisional patent disclosure. The evidence confirms Dr. Herbsleb's testimony, for example:
- Tracking a user's movement is endemic to the disclosure:
"com.leader.osapplications.sessionstate Provisional Patent Application, p. 11.
- Tracking a user’s movement is endemic to the disclosure:
"requestState.getCurrentUser( ).getId() );" Id., p. 13, 15, 16.
- Tracking is implicit in the statement:
"[a]s users create and change their contexts, the files and applications automatically follow, dynamically capturing those shifts of context." Id., p. 6, 9.
- A user tracking component is endemic to workflow automation:
"A > B > C is the workflow process we want to automate. We assign 3 different people to each item, Therefore A(l ,2,3) > B( 4,5,6) > C(7,8,9)." Id., p. 10; Lockwood v. American Airlines at 1572 (“the exact terms need not be used”).
- Tracking users is implicit to the very name of the application “DYNAMIC ASSOCIATION OF ELECTRONICALLY STORED INFORMATION WITH ITERATIVE WORKFLOW CHANGES." Workflow disclosures occur 27 times in the provisional. Id., starting at p. 1; See also U.S. Patent Office (even the USPTO assumes user tracking as endemic to workflow in its publications).
The record discredits Facebook’s expert
Without Dr. Greenberg's testimony, the jury's decision to invalidate the earlier provisional patent application date of Dec. 11, 2002 should not stand as a matter of law. However, Facebook attorneys know that within their "dark arts" bag of tricks is the knowledge that dueling experts create a circumstance that the courts are hesitant to touch: dueling expert testimony. Once provided, it is difficult to argue that the jury got it wrong. So what do infringers do to introduce some evidence, any evidence, however dubious? They make it up. Then, once made up, it gets placed on the scales of justice along with the real evidence. Some evidence thus perfected, the Department of Dark Arts then goes to work to fool the jury. This is one of the dirty little secrets of patent law that has robbed many a patent holder out of his or her just reward.
However, in this case the record of Facebook's "dark arts" proves that: (a) the trial court erred in not disqualifying Dr. Greenberg's testimony the moment he used bad science, and (b) the prima facie facts show that one of ordinary skill knew that a tracking component was disclosed.
Dr. Greenberg rewriting law on provisional patent disclosure?
If the jury's decision to invalidate the earlier provisional patent date stands, it will establish a troubling precedent in which anything short of providing complete source code in a provisional patent application will be insufficient. The direct implication is that all provisional patents will have to be reduced to practice before they are filed.
In addition, a decision to affirm Dr. Greenberg's expert testimony could set in motion a requirement that provisional patents must actually "work" to meet the disclosure requirement.
Dr. Greenberg's uncorrected statement "I have no idea whether these are just place holders or if they have stuff there" implied that the Leader provisional patent application needed to have worked to be an adequate disclosure. This assertion is contradicted by In re Hayes Microcomputer Products, Inc. at 1534:
"an inventor is not required to describe every detail of his invention. An applicant's disclosure obligation varies according to the art to which the invention pertains. . . [and] is sufficient to satisfy the requirement of section 112, first paragraph, when one skilled in the relevant art would understand what is intended and know how to carry it out."Was Dr. Greenberg's testimony "clear and convincing," a "mere scintilla," or null and void?
The Supreme Court stated in Anderson that it isn't good enough to have just a little evidence; stated in Munson that it is certainly not good enough to have no evidence; and re-affirmed the clear and convincing standard of proof in Microsoft v. i4i.
"The mere existence of a scintilla of evidence in support of the plaintiff's position will be insufficient." Anderson v. Liberty Lobby, Inc. at 225.
"When a prayer for instruction is presented to the court and there is no evidence in the case to support such a theory it ought always to be denied, and if it is given, under such circumstances, it is error; for the tendency may be and often is to mislead the jury by withdrawing their attention from the legitimate points of inquiry involved in the issue." Schuylkill v. Munson at 448.Dr. Greenberg's bad science disqualified his testimony as a matter of law. Therefore, since all the other evidence supported Leader's provisional patent priority date, the jury had no factual basis for invalidating the earlier date. A reasonable person can only conclude that the jury became confused by Dr. Greenberg's bad science and, per Munson, the court also confused the jury by diverting their attention from legitimate points of inquiry.
Leader states in its opening appeal brief that "if requiring clear and convincing evidence of invalidity is to be more than an empty semantic exercise, it must mean that the jury's verdict here cannot stand." Leader Opening Appeal Brief, p. 37; also available at Leader.
* * *