Michael Nguyen: Fuzzy Logic in Patent Law

Michael T. Nguyen recently published a very interesting law journal note proposing the use of fuzzy logic in trials of certain kinds of patent cases. See "The Myth of 'Lucky' Patent Verdicts: Improving the Quality of Appellate Review by Incorporating Fuzzy Logic in Jury Verdicts," 59 Hastings L.J. 1257 (2008). Mr. Nguyen provides a nifty summary of how fuzzy logic can control one kind of process (footnotes omitted):

"Fuzzy logic" is reasoning with fuzzy sets. Bart Kosko refers to the "fuzzy principle" in stating that "everything is a matter of degree." Instead of using the crisp truth values "1" and "0," fuzzy logic uses truth values as fractions from 0 to 1. Thus, the statement "John is tall" can be 66% true, and John would have a membership value of 0.66 in the fuzzy set of tall people. When using these percentages, fuzzy logicians do not imply that probability or chance is involved. It would not make sense to say that John has a 66% chance of being tall or that my lawn has an 89% probability of being green.

To illustrate a fuzzy set further, let us look again at the green lawn example. Few lawns are 100% green. Often, a lawn contains a few brown or yellow patches. Thus, the word "green," in the context of lawns, stands for a fuzzy set of colors that constitute green. "We think in fuzzy sets and we each define our fuzzy boundaries in different ways and with different examples." While the definition of these boundaries may differ from person to person, "the very looseness of the fuzzy set enhances its expressiveness." So, while I may make the statement, "My lawn is green," in reality, my lawn might be 89% green, or may have a membership value of 0.89 in the fuzzy set of green lawns, because of a few yellow and brown spots. Most people round up to 100% as a matter of convenience.

Fuzzy reasoning requires the creation of fuzzy rules in the form of "if-then" statements. The fuzzy "if-then" rules express the relation between fuzzy sets. By combining fuzzy rules, we can create a fuzzy system that automatically converts inputs into outputs. Building a fuzzy system can be done in three steps: first, select the inputs and outputs of the system; second, pick the fuzzy sets; and third, choose the fuzzy rules.

My favorite illustration of a fuzzy system of fuzzy rules is the washing machine example. Suppose we want to construct a machine that ""knows' to wash dirtier clothes for a longer duration than clothes which are relatively clean." The "input is the degree of dirtiness and [the] output is the duration of the wash." The fuzzy inputs would be: almost completely clean, relatively clean, slightly dirty, dirty, and very dirty. The fuzzy outputs would be: rinse, wash lightly, wash, wash thoroughly, and wash vigorously. Finally, we choose the fuzzy rules: (1) if the clothes are almost completely clean, then only rinse them; (2) if the clothes are relatively clean, then they are lightly washed; (3) if the clothes are slightly dirty, then they are washed; (4) if the clothes are dirty, then they are washed thoroughly; (5) if the clothes are very dirty, then they are washed vigorously.

This fuzzy system can be "defuzzified" by attaching specific numbers to the vague concepts. Fuzzy concepts can be defuzzified by averaging or finding the centroid (i.e., center of mass) of the output numbers. Defining dirtiness as a range of particles of dirt from 10 to 100 and duration of the wash from 10 to 100 minutes, we can assign certain values to our fuzzy sets. Thus, the washing machine will literally think for itself and determine how long to wash laundry based on how dirty it is. Such products have been developed in Japan "to raise the machine IQ of camcorders and transmissions and vacuum sweepers and hundreds of other devices and systems."

the dynamic evidence page

consulting on investigation strategy and the law of evidence

Frederick Schauer's Defense of General Rules (Maxims?) about the Epistemic Worth of Categories of Evidence

There is an interesting new paper by Frederick Schauer: "In Defense of Rule-Based Evidence Law - And Rule-Based Epistemology Too".

The thesis is, in general, good: It is epistemologically possible and sound to have (some) general rules about the probative worth of (some) classes of evidence.

Schauer's general thesis is, thus, sensible. But more arguments in favor of his general thesis must be made. For example, one might consider how it would be possible to learn from experience if one could not extract (whether implicitly or explicitly) from experience any general principles about the workings of the world and the relationship of events in the world to phenomena that seem to serve as indicators or signs of events; complete "individuation" of judgments about probative value (a/k/a evidentiary value or force) would seem to bar the possibility of knowledge based on experience.

A separate (and important) question is whether the the particular generalizations that are or may be embedded in the American law of evidence about the probative value (or lack of probative value) of certain categories of evidence (e.g., hearsay) are warranted. The mere fact that there must be some generalizations does not mean that American law has identified the correct ones. Still, the argument made by Schauer is refreshing. It is the beginning of a sensible attack on the ludicrous (so I would say) hypothesis that the probative value of evidence depends entirely on individual circumstances and details.

the dynamic evidence page

consulting on investigation strategy and the law of evidence

How Youths and Drunkards Sometimes Die and How You Can Sometimes Prove It in New Jersey (and Perhaps Elsewhere)

The full text of the following essay is now available (gratis) via SSRN: P. Tillers, "The Death of a Youth and of a Drunkard: A Remarkable Story of Habit and Character in New Jersey."

To retrieve this essay (and story) go here.

the dynamic evidence page consulting on investigation strategy and the law of evidence

Ho Hock Lai's New Book

A new book on evidence and the law of evidence landed on my desk yesterday (courtesy of amazon.com). A scan of the book's contents suggests the book is worth a very careful read. See Ho Hock Lai, A Philosophy of Evidence Law (Oxford 2008). See the blurb (and more) here.

the dynamic evidence page

consulting on investigation strategy and the law of evidence

Visual Evidence

the dynamic evidence page

consulting on investigation strategy and the law of evidence

The Story of a Signature Crime

I am usually wary of claims that a crime committed with a particular m.o. is a "signature crime." Such claims usually strike me as urban myths or as the products of self-aggrandizing "experts." However, when the Litvinenko business first broke, it struck me (see, e.g., here) that the poisoning and eventual killing of Litvinenko with a highly toxic and barely detectable poison, Polonium 210, was indeed the signature -- the likely hallmark -- of a crime committed by the present-day successors to the KGB. Well, it turns out that my suspicions (and those of many others) were probably correct. See this NYTimes book review by Simon Sebag Montefiore. (How can one doubt the credibility of a person with such a magnificent name?) But, dyed in the wool academic that I am, the interesting part of this story for me is why my notions about how the KGB committed crimes turned out to be correct. Part of the explanation is that the newspaper stories that reported on the Litvinenko business also reported other cases in which the KGB and FSB had supposedly used a similar m.o. But even before I read those relatively recent newspaper accounts I had the impression that the Russian secret police were in the habit of doing their dirty work with furtively-administered poisons. So I was the captive of an urban myth. But it was a myth that turned out to be very probably true. Isn't that interesting?

N.B. Didn't Ian Fleming use furtive poisoning as one of the tactics of his SMERSH in at least one of his James Bond novels?

the dynamic evidence page consulting on investigation strategy and the law of evidence

Evidence Codification and Non-Codification in Connecticut

Evidence codification has an unusual status in Connecticut. In a case said [in Westlaw's text] to be decided on August 19, 2008 -- in State v. DeJesus --- A.2d ----, 2008 WL 3290164 (2008) -- the Supreme Court of Connecticut writes [will write?]:

"With respect to the defendant's appeal, we conclude that, despite the adoption of the code [of evidence] by the judges of the Superior Court, the appellate courts of this state retain the authority to develop and change the rules of evidence through case-by-case common-law adjudication."

&&&

"We agree with the defendant that the adoption of the code did not divest this court of its inherent common-law adjudicative authority to develop and change the rules of evidence on a case-by-case basis. We further agree with the defendant that, in light of our recent clarification of the nature and scope of the common scheme or plan exception in State v. Randolph, supra, 284 Conn. at 328, evidence of uncharged misconduct admitted under the liberal standard ordinarily does not reflect the existence of a genuine plan in the defendant's mind. Nonetheless, given the highly secretive, aberrant and frequently compulsive nature of sex crimes, we conclude that the admission of uncharged misconduct evidence under the liberal standard is warranted and, therefore, we adopt this standard as a limited exception to § 4-5(a) of the code, which prohibits the admission of "[e]vidence of other crimes, wrongs or acts of a person ... to prove the bad character or criminal tendencies of that person."

&&&

"The foregoing history [of the adoption of the code of evidence by the judges of Connecticut's Superior Courts] does not support the conclusion, however, that the code was intended to divest this court of its inherent authority to change and develop the law of evidence through case-by-case common-law adjudication."

&&&

"There was no discussion of the effect, if any, that adoption of the code would have upon this court's common-law adjudicative authority to change and develop evidentiary law on a case-by-case basis, an inherent authority that it has enjoyed since the seventeenth century.

...

"In the absence of any discussion at the meeting of the judges of the Superior Court concerning the impact that adoption of the code would have on the future development of evidentiary law, it is illogical to conclude that, by adopting the code for the purposes of ease and convenience, the judges intended to divest this court of its long-standing inherent common-law adjudicative authority over evidentiary law." &&&

In a footnote the Supreme Court dealt with the question whether Connecticut's code of evidence prevents Connecticut's trial judges from changing rules of evidence in case-by-case adjudication:

"FN24. Because the code merely restated the prevailing common-law evidentiary rules, which the judges of the Superior Court already were bound to apply, and was intended to expedite and streamline judicial proceedings by serving as a shorthand reference to those rules, the code clearly was intended to be binding authority in the Superior Court. Section 1-1(b) of the code specifically states that '[t]he [c]ode applies to all proceedings in the superior court in which facts in dispute are found, except as otherwise provided by the [c]ode, the General Statutes or the Practice Book.' The code therefore differs fundamentally from a treatise or handbook, which has persuasive value only. The question presented in this appeal, however, is not whether the code is binding authority in the Superior Court, but, rather, whether it is binding authority in this court such that we are precluded from reconsidering our own prior precedent codified in the code. For the reasons explained in the body of this opinion, we conclude that the judges of the Superior Court did not intend their adoption of the code to divest this court of its inherent authority to change and develop the law of evidence via case-by-case common-law adjudication."

the dynamic evidence page

consulting on investigation strategy and the law of evidence

Questions about the Bruce Ivins Anthrax Case

Dr. Gerry Andrews, a professor of microbiology, raises some interesting questions about the evidence against his former colleague, Dr. Bruce Ivins, the suspected "anthrax killer" who recently committed suicide. See Gerry Andrews, "Open Questions on a Closed Case" (Op-Ed), NYTimes Online (August 9, 2008). Of particular interest to me are several questions that Dr. Andrews raises about the scientific evidence in the case.

The US attorney supervising the investigation and other government agents spoke about the "match" between the anthrax that killed the five victims and the anthrax that had been in Ivins' possession. One government agent -- I believe it was the the US attorney who held the news conference about the impending closing of the case -- even said that the government had located the "murder weapon," the flask containing the lethal anthrax that, it was said, killed the five victims.

Although I have no reason to doubt that the federal government has (finally) fingered the right (dead) man, one must always be suspicious when there is talk about a "match." The word "match" implies evidence that uniquely identifies the thing (or, worse yet) person who left some trace at the crime scene. Dr. Andrews points out several circumstances and considerations that raise questions about the government hypothesis that the scientific evidence in the case points uniquely to Bruce Ivins as the culprit. He makes his point this way:

As a scientist, ... I feel compelled to comment on what should have been the Federal Bureau of Investigation’s strongest link between Dr. Ivins and the terrible crime — deadly anthrax spores. In the summary of its findings, the F.B.I. states that investigators used four different genetic techniques to match the anthrax-laced attack letters to a unique DNA footprint of a single anthrax spore preparation in one flask that had been in Dr. Ivins’s custody.

Sounds reasonable. Yet the investigators present no details on the scientific methods they used to make this match or how they employed them. That’s a problem, because without such detail it is hard to tell if they specifically ruled out a similar match between the anthrax in the letters and anthrax preparations with the same DNA footprint kept at a number of other labs around the country.

For good measure, Dr. Andrews points out how "non-scientific" questions about the custody of the "murder weapon" raise additional doubts about the strength of the scientific evidence against Ivins. Andrews begins by noting that "Dr. Ivins was an investigator in the case before he was a suspect. After the anthrax attack, Dr. Ivins himself worked directly with the evidence. The F.B.I. asked Dr. Ivins to help them with the forensics in the case by analyzing the contents of suspicious letters. And he did so for years, until the authorities began to suspect that the anthrax spores used in the mailings might have originated from his lab." Dr. Andrews argues that the lab at which Ivins worked did not have the capacity to produce the "refined weapons-grade anthrax" that had been used to kill the five victims. Dr. Andrews then adds:

But even leaving that aside, there are important questions left unanswered. First, isn’t it possible that the manipulation of the contents of the anthrax letters in Dr. Ivins’s laboratory might have contaminated the work environment enough to potentially jeopardize the integrity of subsequent samples taken from the lab? Might that perhaps explain why the anthrax powder used in the attacks was later found to have the same DNA footprint as the other anthrax preparations in Dr. Ivins’s lab? At the very least, wouldn’t this call his guilt into doubt?

These are the sorts of questions -- e.g., the degree to which the DNA of the physical sample found at the scene of the crime together with the DNA of a specimen found elsewhere is a distinctive identifier of the physical source of the crime scene specimen, the possibility of contamination of the physical evidence found either at the scene of the crime or elsewhere, the availability of the possible instrumentality of the crime (in this case the "murder weapon") to persons other than a specific suspect, -- these are the sorts of questions that have emerged in innumerable "humdrum" murder and rape cases. Despite the new scientific techniques that are said to have been developed during the investigation of the anthrax case, these sorts of issues are pertinent in the Ivins anthrax case as well in humdrum murder and rape cases. We should avoid being unduly dazzled by novel scientific methods or technology. We must always keep in mind what sorts of questions a powerful scientific method can answer and what sorts of questions it cannot answer.

the dynamic evidence page

legal consulting in the law of evidence, investigation strategy, and evidence planning

Truly Demonstrative Evidence

Martha Neil, "Gov’t. Puts Car in Court in Murder Trial," ABA Journal Online (August 7, 2008):

In an unusual effort to prove a murder case, prosecutors in Washington state have persuaded a judge to let them bring a car into the courtroom. The 2000 Ford Mustang in which victim Ronald Whitehead, 61, allegedly was shot to death "was hauled into the King County courtroom of Judge Steven Gonzalez in three bundles Wednesday night and assembled in time for when defendant Jon Ogden, the dead man's 20-year-old stepson, walked into trial Thursday morning," reports the Seattle Times. The government wants to show the jury how they believe Whitehead was shot four times by someone hiding in the vehicle.

the dynamic evidence page

legal consulting in the law of evidence, investigation strategy, and evidence planning

How Do Fingerprints Point?

Kenneth Chang, "Fingerprint Test Shows Not Only Who but What," NYTimes (August 7, 2008)

the dynamic evidence page

Peter Tillers on Evidence: legal consulting on evidence, investigation strategy, and evidence planning

When Will They Ever Learn -- about Circumstantial Evidence etc.

The byline in today's NYTimes reads "Anthrax Evidence Is Called Circumstantial." (The story is by Scott Shane.) The implication of the byline is that the evidence against the suspected (and now-dead) malefactor is therefore weak. The NYTimes should know better. There is weak circumstantial evidence. But there is also strong circumstantial evidence. The same holds true for "direct" evidence given by a putative eyewitness. The same holds true for scientific evidence. The same holds true for confessions.

"When will they ever learn?" Folk Song, the version by Pete Seeger and The Weavers still being the best known

N.B. The above post does not reflect -- and it is not meant to reflect -- my personal opinion (if I have one) about the guilt or innocence of Bruce E. Ivins.

the dynamic evidence page

A Merger of Subjective and Objective Factors in Factual Inference: Understanding a Bee's Vision

It is possible to make inferences about how bees are likely to view their world by carefully conducting experiments and inspecting the physiological and psychophysics data. We know that bees can see ultraviolet wavelengths of light. Many flowers have evolved flower patterns that suit the visual capabilities of bees, for example, some flowers have patterns that can only be seen with ultraviolet sensitive vision.

"Interview [of Dr. Adrian Dwyer]: Through the eyes of a bee," Alexander von Humboldt-Stiftung Newsletter 2/2008

the dynamic evidence page

Aleksandr Solzhenitsyn, Dead

See Aleksandr Solzhenitsyn Is Dead at 89," NY Times (Aug. 6, 2008).

This sad news has something to do with evidence. Before Solzhenitsyn's appearance many people in the West -- I am old enough to remember this -- were dismissive of claims that Stalinism led to the deaths of millions of people. But the evidence of death on such a massive scale was publicly available long before Solzhenitsyn burst on the international scene.

the dynamic evidence page

Narrative and Factual Inference

In recent years it has become fashionable -- almost de rigeur -- to talk about "narrative" when talking about factual inference in legal settings. This habit of mind has even percolated into the halls of the Supreme Court. (I assume that the Supreme Court building has halls.) See Old Chief v. United States, 519 U.S. 172 (1997) (Souter, J., for the Court). This shift in interest is, on the whole, a good thing, I think. But it is also worth asking why it this development is a good thing. Yes, (some) stories have rhetorical and persuasive force. Yes, (many) stories serve as powerful mnemonic devices. But if the question is whether we should be interested in stories if our ultimate interest is accurate (as well as entertaining or advantageous etc.) factual inference in settings such as trials, the answer is less obvious. We must begin, I think, by distinguishing between two types of "stories." One type of story is a scenario, which, in the parlance of social scientists and other such folks, is a causal hypothesis-a hypothesis about the connections between specified events over time. Another type of story is broader: this broader type of story is an account that has, not only causal explanatory force, but also ingredients such as dramatic appeal, emotional punch, human actors, and the like. A strong case can be made that either all factual issues or almost all factual issues are effectively scenarios -- this on the ground that all, almost all, or many factual hypotheses in legal settings are effectively hypotheses about some possible sequence of connected events in time. However, it may be less clear that epistemic considerations--i.e., truthseeking considerations--make stories in the sense of narrative necessary. But perhaps "necessity" is too strong a requirement here. Perhaps it is enough if we show (if we can) that stories with emotional wallop etc., generally promote [under some circumstances] the search for the truth, generally advance [under some circumstances] accurate fact finding. (This thesis is a variant of the sensible notion that practical epistemic maxims should take the human animal as we find him [or her].)

the dynamic evidence page

The Indeterminacy and Elasticity of Legal Language

I have come to detest the indeterminate word "indeterminate." But for present purposes no other word seems to do the trick I want to do.

For decades American legal theorists have talked about the indeterminacy of legal language. Probability theorists prefer to talk about the uncertainty of legal terms. But in a recent message to a discussion list Lotfi Zadeh once again noted that it is important to distinguish between uncertainty about the meaning of words (language) and the elasticity, or plasticity, of words (language). The distinction that Zadeh makes between uncertain meaning and elastic language is, think, very important for an understanding of the nature of legal reasoning and interpretation. Although elastic words (I would say) produce uncertainty, it is important to remember that words themselves are elastic, i.e., that words exhibit elastic "behavior." Note: it is possible, in principle, to know fairly precisely how elastic words behave under various circumstances. When we have such knowledge about a word, we are not really very uncertain about the meaning of the word but we still can say and must say that the meaning of the word in question varies, or stretches, depending (for example)on the context. (In such a situation there is only a very loose -- and possibly misleading -- sense in which it can be said that the meaning of the word is indeterminate.)

the dynamic evidence page

Potemkin Trials?

[B]ehind the judicial routine at the first trial for a Guantánamo detainee lies a parallel universe of law and lawyers. Secret evidence held in red folders is not revealed in open court. The gallery is mostly empty, because there are no members of the public. In what would be the jury box, every occupant wears a military uniform.

William Glaberson, "A U.S. Trial by Its Looks, but Only So," NYTimes (July 29, 2008)

Question: If some detainees are acquitted, will this prove that the Guantanamo somewhat-trial-like proceedings are just?

Answer: It may just prove that some military jurors are good-hearted -- or, alternatively, that some of the military actors are concerned about public relations or other such matters.

A system of adjudication can be distorted if it produces an unacceptable number false negatives as well as if it produces too many false positives. One thing seems reasonably sure about the proceedings at Guantánamo: much of the important evidence will not be subjected to adversarial testing. There is good reason to wonder if it will be subjected to an adequate degree and quality of nonadversarial evaluation.

the dynamic evidence page

As much as he is hated by some people, George W. Bush deserves credit ...

...for proposing and supporting expenditure of large sums of money to combat AIDS/HIV, malaria, and other diseases in Africa and in other parts of the world. See, e.g., the AP story "Bush eager to sign bill tripling AIDS assistance". Such health programs were not on the American agenda until Bush proposed them in his State of the Union address in 2003.

How much of a difference have such aid programs made? Consider just one piece of data from the story cited above:

The current $15 billion act, which expires at the end of September, has helped bring lifesaving anti-retroviral drugs to some 1.7 million people and supported care for nearly 7 million.

the dynamic evidence page

Trial by Ordeal Lives -- and Apparently Works

See the following YouTube clip of a National Geographic program.

Tillers' dynamic evidence page

Vagueness and Fuzziness

Probability theorists tend to think that everything that fuzzy logic can do it (probability theory) can do better. But probability theory is a procedure for dealing with uncertainty. Perhaps some things are vague -- legal language, for example -- without being uncertain. So perhaps the question of the legitimacy of fuzzy logic boils down to the question of the existence or non-existence of vague objects that are not necessarily uncertain.

But perhaps fuzzy logic also legitimately applies to reasoning about uncertain propositions -- because perhaps some or much reasoning about uncertain propositions involves vague objects (concepts). Be that as it may, perhaps it is true that fuzzy logic will gain a greater measure of respectability among standard probability theorists if the distinction between uncertainty and vagueness is solidified.

N.B. It may be true that vague concepts -- e.g., vague legal concepts -- work in ways that are uncertain to some degree. But does it follow that such (vague) legal concepts are "uncertain" to some degree? This is perhaps a nice test question for logicians and legal theorists.

the dynamic evidence page

Henceforth Criminal Defendants in Japan Will Be Allowed to Wear (Clip-On) Ties and (Fake) Shoes at Their Trials

Japan Times Online reports that criminal defendants will now be allowed to wear clip-on neckties and "and open-back slippers that look like leather shoes." These changes were precipitated by the advent of a new trial system in which lay judges as well as professional judges will preside in trials of serious criminal charges.

Thought is also being given to removing shackles and handcuffs from the accused before the accused enters the courtroom. In addition, under certain circumstances, the accused will be allowed to sit next to counsel, and not, as at present, between two police officers, with defense counsel sitting behind this trio. However, one of the police officers will plant a leg between the feet of the accused and defense counsel. In Japan justice may not be swift but it does have long and sturdy legs.

the dynamic evidence page