The Importance(?) of Understanding the Mechanics and Logic of Perception

Law journals devote quite a bit of attention to studies of the reliability and unreliability of eyewitness identification. But could the legal process produce better assessments of eyewitness reports if trial lawyers and judges knew more about the technology, or physiology, of perception and the logic that informs such perception?

Caveat: It does not necessarily follow that human knowledge of perception is presently good enough to be used in the courtroom -- and, even if such knowledge is useful "in principle" for forensic purposes, it does not necessarily follow that lawyers, judges, and jurors have the training or intelligence to make effective use of contemporary knowledge of human perception.

Counter-caveat: It is not prudent to underestimate the intellectual prowess of jurors; and some lawyers and judges have a pleasing degree of scientific literacy.

The question I pose here is not trivial -- for it is an iteration of the question of the extent to which human beings understand their world without understanding it, viz., of the extent to which human beings are capable of drawing inferences about the world without understanding the mechanics that make it work as does. Conversely stated, the question posed here implicates the question whether knowledge of causes improves inference even if it is true that some inference is possible without (much) knowledge of causes.

Counterpoint: The hypothesis that perception (truly) is (pretty good) inference suggests that human beings -- by virtue of their heredity, physiology, etc. -- know much more than they can put in words.

But the question remains: Can explicit knowledge of causes improve inference?

The answer to this question would seem to have to be "yes": It is very hard to deny that some explicitly-formulated knowledge of nature's mechanics -- e.g., gravity -- enables human beings to make better inferences and predictions (predictions are merely a special form of inference) in some situations.

A final word of caution: Even a worm knows how to burrow into the soil. (Indeed, a worm probably knows how to do that better than you do.) But (as far as I know) worms have not produced treatises on soil mechanics.

Perception as Inference (again)

E.T. Jaynes, PROBABILITY OF THEORY: THE LOGIC OF SCIENCE Section 5.4 at 133 (2003):

Seeing is not a direct apprehension of reality, as we often like to pretend. Quite the contrary: seeing is inference from incomplete information, no different in nature from the inference that we are studying here. The information that reaches us through our eyes is grossly inadequate to determine what is "really there" before us.

N.B. The discussion here does not suggest that Jaynes was intimately familiar with recent research on the logic of perception. But he was prescient in suggesting that researchers should investigate whether Bayesian logic informs perception.

Support for the Proposition that Values Depend on (Perceptions of) Facts

Some years ago I argued that there is evidence in law, that the values embedded in law (even in legislation) are in part a function of beliefs about factual propositions, including factual inferences that rest on evidence. See P. Tillers, The Value of Evidence in Law, 39 Northern Ireland Law Quarterly 167 (1988). Perhaps the following statement by Jaynes (amusing footnote omitted) offers some support for my view:

We consider it an important aspect of "objectivity" in inference -- almost a principle of morality -- that we should not allow our opinions to be swayed by our desires; what we believe should be independent of what we want. But the converse need not be true; on introspection, we would probably agree that what we want depends very much on what we know, and we do not feel guilty of any inconsistency or irrationality on that account.

E.T. Jaynes, PROBABILITY OF THEORY: THE LOGIC OF SCIENCE Section 13.12.5 at 424 (2003).

Inference Is Better-Grounded than Choice; and Analysis of Evidence Is More Secure than Economic Analysis -- Is It So?

E.T. Haynes, PROBABILITY OF THEORY: THE LOGIC OF SCIENCE Section 13.12.4 at 424 (2003):

[I]t now appears that from a fundamental standpoint loss functions are less firmly grounded than are prior probabilities. This is just the opposite of the view that propelled the Wald-inspired development of decision theory in the 1950s, when priors [prior probabilities] were regarded as vague and ill-defined, but nobody seemed to notice that loss functions are far more so. For reasons we cannot explain, loss functions appeared to workers at that time to be "real" and definite, although no principles for determining them were ever given, beyond the truism that any function with a continuous derivative appears linear if we examine a sufficiently small piece of it.

In the meantime, there have been several advances in the technique for assigning priors by logical analysis of prior information. But, to the best of our knowledge, we have as yet no formal principles at all for assigning numerical values to loss functions; not even when the criterion is purely economic, because the utility function of money remains ill-defined.

Great Law Schools & Great Libraries

My law school does not do badly in the law school ratings game. But the law school rating services play a poor game because they generally ignore one crucial measure of the greatness of a law school: the quality of a law school's library.

We can have endless debates about whether a law school either is ought to be essentially an academic institution or a professional school, or whether the academic-professional divide is a false one. But -- regardless position we take on such issues -- all sensible law teachers and legal practitioners should agree on one point: much of law centers on TEXT. Hence, a great law school, regardless of how it defines its mission, must be a great repository of textual material (cases, treatises, journals, the lot).

If a law school is to grant text its proper role in the life of a law school, the law library must be a sanctuary, and the library ought to be an inviting and alluring sanctuary. For example, the seats should be comfortable and the physical environment should be aesthetically pleasing and warm. The library must be so arranged that its "customers" want to spend time in it.

A great law school must have a great library. Does US News & World Report know this? Does Brian Leiter know this?

Apparently not.

N.B. My law school fares worse -- not better -- if "library quality" is a measure of the quality of a law school. So this post does not serve a narrowly-conceived personal interest.

A Stellar Conjunction -- or a (non)Lunatic One

Will the Red Sox win the World Series just at the moment that the moon goes blank? If so, are the Red Sox responsible? Or is the moon responsible? Which way does the chain of causation run? Is non-Luna pulling the Red Sox or are the Red Sox eclipsing the moon? I need a Latin phrase here. ("Post hoc, propter hoc"?)

European Morality

"Hussein's government killed an estimated 300,000 people, most of them Shi'ite Muslims or ethnic Kurds, rights groups say. The Iraqi government has identified about 40 mass graves, but until now none has been scientifically exhumed -- in part because European forensic teams won't collect evidence that might be used to win death penalty convictions." Thanassis Cambanis, "In Iraq grave, evidence of regime's horrors," Boston Globe (online) (October 13, 2004).

What Is "What Is Evidence?"?

Some scholarly discussions of the law of evidence begin with a question such as "What is evidence?" or "What is proof?" See, e.g., I Wigmore on Evidence Section 1 (P. Tillers rev., 1983).

Questions such as these have a mind-numbing quality; they have a tendency to paralyze thought.

Why? Is there a better way to consider the nature of matters such as "evidence" and "proof"?

Perhaps questions of the form "What is ... [some thing or phenomenon in law]?" induce mental or intellectual paralysis in part because such questions incline the observer to launch a search for attributes which, when properly assembled and arranged, could constitute -- the observer may hope -- a correct or adequate definition of a phenomenon such as "evidence" or "proof." If a non-solipsistic observer conducts a non-circular search for the attributes of a (complex[!]) social(?!) phenomenon such as "legal evidence," (s)he is likely to generate a very long list of attributes. Such a list of features may end up being a mere aggregation of attributes that resembles a serving of thick porridge unaccompanied by any explanation for the identity or quantity of the ingredients found therein.

I do not wish to overstate my objection to starting discussion of the law of evidence with a definition: it is unlikely that conceptual mush is an inevitable effect of launching an investigation with a request or search for a definition. I only wish to suggest that acquiring an understanding of the nature of a legal phenomenon or practice such as the law of evidence is not best promoted by formulating and then pondering assertions such as "legal proof is an epistemic process," "judicial proof is a legal process" and "judicial proof is a symbolic process," and that the search for an understanding of a matter such as legal proof is better promoted by formulating topics of discussion in the following fashion: "scarcity in proof," "time in proof," "evidence in legal proof," "argument in legal proof," and so on.

A restatement of the general question under discussion in this post:

Is it fair -- or is it instructive -- to begin a discussion of the law of evidence or proof with a definition or definitions of matters such as "evidence" or "proof"?

A tentative answer to the (reformulated) general question:

A question such as "What is [legal] evidence?" is in part an empirical question: unless one is a Platonist -- or unless one denies the possibility of social variation --, a good answer to such a question always requires in part an account of what is conventionally considered to be a thing such as evidence.

True, a definition of phenomenon such as "evidence" or "proof" should not be a mere catalogue of the matters that are considered "evidence" or "proof": a good definition crystallizes a wide diversity of phenomena; a good definition resembles a rule that generates or explains (perhaps only by and large) a wide and diverse collection of phenomena that might be considered instances of a matter such as "legal evidence." But it does not follow that one ought to begin a scholarly discussion of a legal field such as the law of evidence with a rule or formula that (putatively) specifies the essential or important attributes of a matter such as "evidence" or "proof." It is probably better instead to proceed quasi-empirically and quasi-inductively: judgments about the important or "essential" attributes of matters such as "legal evidence" should emerge out of ruminations based on our observations of the real-world workings of matters such as "the law of evidence" or "proof in legal proceedings." (Such ruminations may, but need not, devolve into bare quasi-statistical statements of the relative frequency of various attributes in a process such as "proof in legal proceedings.")

N.B. Is it not the case that for some purposes -- including the present one -- a good "definition" of a social phenomenon and practice such as "judicial proof" must include an account of the motivation(s) for the phenomenon or practice? (Construe "motivation(s)" broadly: make it encompass "function(s).")

Postscript #1: Definitions -- properly and broadly understood -- are important. One needs them to understand the spirit of a social phenomenon or practice that, because of its variety and diversity, may otherwise seem bereft of rhyme or reason.

Postscript #2:The ruminations in this post are intended only for (actual or aspiring) authors of legal texts. (These ruminations are, in any event, unlikely to interest anyone else.)

Sir Richard on Sir Arthur

Judge Richard Posner finds little to admire in Sherlock Holmes' methods. See R. Posner, "CSI: Baker Street," New Republic (October 11, 2004 [which is, BTW, surely a fictitious publication date {since I am reasonably sure that today is October 4, 2004, and in my limited human experience the clock and the calendar do not run backwards, but who am I to say they could not?} -- explain this, ye besotted publishers of journals a/k/a ye purveyors of 21st century versions of Soviet-style histories]).

Judge Posner thinks some or many of Holmes' deductions are little more than a shot in the dark -- and are therefore unscientific.

Verily, verily, I say unto ye ("thee"?; i.e., "to all of you out there"): science requires shots in the dark, shots that are merely(!) prompted, or suggested, by evidence.

Verily, verily, I say unto thee: science depends on careful deductions but it also depends on abductions a/k/a imaginative hypotheses suggested but not dictated by evidence or logic.

Verily, verily I say unto thee -- or ye, or y'all: imaginative reasoning is not an oxymoron -- and, besides, where would we be without Einstein's imagination, an imagination par excellence but an imagination that could work its wiles only because of Einstein's meticulous attention to details and matters such as clocks and their synchronization in distant places [for train schedules and other such purposes]? {You surely don't expect me to answer this last question, do you? The best answer I can give you now: not quite where we are now.}

Take Two Samples ...

Take two samples of handwriting, practically any two samples. Make sure the two samples are made by different people. Question: If you look close enough and long enough, how probable is it that you will find the same extraordinarily rare combination of characteristics in the two samples?

Many years ago I followed exactly this procedure. I had a group of students write down the same phrase twice on two different pieces of paper and then throw their handwriting samples into a hat. I then picked two samples that I knew -- or believed -- had been written by two different people. (I think I asked the students to write their names on the back of each piece of paper with their handwriting samples on the front and I had the students do this before they knew what I was up to.) As I said just moments ago, I picked, more or less at random, two handwriting samples that had been made by different people, by different students. I then scrutinized these two handwriting samples for a while. After doing so, I found about a dozen handwriting quirks that occurred in both samples. I pointed out these similarities to the class. I then did some product rule calculations and I asked the class to do the same with the probability values (and dependencies) that they thought were appropriate. I then asked the students in the class whether they thought the two samples were written by the same person. Everyone (in a class of ca. 25) answered in the affirmative. (I had somehow managed to instruct the actual authors of the two samples to keep their mouths shut.) When I told the students that in fact two different students had produced the handwriting samples, about five students found my confession to be both astonishing and unbelievable; and despite my confession of trickery, they argued that the two sample had been written by the same person.

As I recall, statements by the two students who (I think) actually made the two samples overcame the objections of the dissenting students.

I believe I successfully tricked the class. But I can't really say that the dissenters were completely befuddled or irrational, can I?

Dreyfus redux?

My pedagogical trick would not have worked if one writer had written English and the other, Arabic. It would not have worked if one author had been 25 years old and the other, three years old. Therefore?

Improbable DNA

Jennifer Mnookin, "Fingerprint Evidence in an Age of DNA Profiling," 67 Brooklyn L. Rev. 13, 49-50 (2001)(footnotes omitted):

[I]n a 1999 case in England ... Raymond Easton was charged with burglary after authorities made a "cold hit" with his DNA in a DNA database. His DNA matched the crime scene DNA at six loci. Because there was only a one in thirty-seven million chance that a randomly selected person's DNA would match, Raymond Easton was charged with burgling a house 200 miles from where he lived. However, after Easton, who had advanced Parkinson's disease and was unable even to drive a car, offered an alibi for the night in question, the DNA was eventually tested at four more loci. This more sophisticated test showed there was no DNA match after all. All charges were dropped.

Investigating Multiple SIDs Deaths

Question for the day:

If the number of multiple SIDS ("sudden infant death syndrome") deaths within single families within some large population is exactly what one would expect if chance alone governs the distribution of SIDS, should government authorities investigate for possible wrongdoing if the only thing they know is that there were, apparently, three SIDS deaths within a single family?

Further questions:

(i) Are three such deaths within a single family ever sufficient for a criminal conviction of a person who alone had access to the children when they died?

(ii) If not, would four deaths suffice?

(iii) If not, would five or ... n deaths ever suffice?

Dice, Probability, and Law

There is more to probability than dice and games of chance. Nonetheless, I think it is probably(!?) useful to use dice to introduce law students to some basic points about probability theory. I like to use a set of large "fair" dice. (Later I will perhaps post a [true!] story about my unsuccessful attempt to buy magnetized dice.) This Monday I will also try to use the nifty applet at the following web site to make several points: Introduction to Probability Models.

Magnetized dice would be a nice way to illustrate dependent probabilities.

Repeated rolls of dice (with, e.g., the applet mentioned above) can be used, I think, to show, by analogy, some of the problems that can arise with the use of statistics about the relative (in)frequency of SIDS to prove criminal guilt or, even, with the use of such statistics to justify "just" coercive investigation by the state.

A Timely Closing Argument

In a baby murder trial in which the baby finally died of suffocation, the prosecutor -- Michael D'Andrea -- "asked the jury to look at the clock while one minute ticked by -- the amount of time it would have taken [the baby] to suffocate. As the time elapsed, D'Andrea stared directly into [defendant's] face from across the defense table." Michaelangelo Conte, "Jury takes just three hours to convict mother's boyfriend in baby's death," The Jersey Journal pp. A1 & A10 (September 22, 2004).

I saw some fancy lawyering when I practiced law in Texas (many years ago). But New Jersey lawyers, it seems, have their own bags of tricks.

Some Gentle Readers out there can surely relate stories about similar forensic tricks they have seen; I doubt that Mr. D'Andrea is the first trial lawyer to ask a jury to literally watch a clock for 60 seconds or so. In civil litigation the best-known parallel, now generally frowned upon, is for plaintiff's counsel in a personal injury case to ask a jury to imagine how much suffering plaintiff must endure each second of his or her life, put a dollar value on each second's suffering, and then tote up all of those dollars and return a handsome verdict for plaintiff.

How Should We Think about a Person's Possible Recklessness?

What kind of a fact is "He was reckless"?

"Recklessness" is a concept.

But "He was reckless" is not merely a reformulation of the notion "recklessness."

"He was reckless" is an assertion that there is (or was) an instance of "recklessness."

We want grounds for thinking or asserting, "He was reckless."

We do need the concept, or notion, "recklessness."

But we also need something more.

What more?

The concept "recklessness" can be decomposed. In law it usually is decomposed.

Let's do so.

One legal definition of recklessness might run this way: "A person acts recklessly if he or she consciously disregards a substantial and unjustifiable risk." Cf. New York Penal Law § 120.25."A person is guilty of reckless endangerment in the first degree when, under circumstances evincing a depraved indifference to human life, he recklessly engages in conduct which creates a grave risk of death to another person." Add the gloss: A person consciously disregards a risk only if he is consciously aware of the risk. Accord, e.g., Knapik v. Ashcroft, 2004 U.S. App. LEXIS 19445 (2nd Cir., Sept. 17, 2004) (noting that a federal agency concluded that a legal definition requiring conscious disregard of risk "requires an actual awareness of the risk created by the criminal violator's action"). So: assume that one of the ingredients, or elements, of "recklessness" is "conscious awareness"; i.e., one can be called "reckless" only if it can fairly be said that one is "consciously aware" of a risk of some kind.

When does one have an instance of "conscious awareness"?

One cannot indefinitely decompose a notion such as "recklessness."

But we (e.g., a jury) still must decide when to call a state of mind "conscious awareness" and when not to do so. We want to have the jury have grounds for doing so or not doing so.

Aha, evidence surely comes into play now!

"Awareness" -- or, in any event, some state of mind or consciousness -- may be reported by some person -- or there may be a report of a state of mind that suggests awareness: e.g., "It occurred to me that there was a risk of death"; "I heard him mutter, almost under his breath, that he realized there was a serious risk of serious harm."

We might take such a report as a signal.

We might even try to take the thing reported -- the state of mind reported by the reporter -- as a signal.

But if the report is taken as evidence rather than as an exemplar of a matter such as "awareness," ... hmmm ... then what?

Let's use this terminology:

m-1 = state of mind
m-1* = report of a state of mind
A = conscious awareness

If we are to think frequentistically and along Bayesian lines, and if we are therefore to think p(m|A) over p(m|~A), i.e., h/f; we must somehow, in some way, separate m and A: we cannot think of instances of m as being, pro tanto, instances of A; instances or states of m must be distinct from A, nicht wahr? And if that's the way things go when we think frequentistically, perhaps that's also the way things (our thought) must go if we think probabilistically.

But what if we do see A in m -- what if we [already] see A in m and what if we wish to figure out how much of A we should see in m? [Is this a coherent way of thinking?]

Consider a reporter. Suppose the reporter is also the actor; i.e., we wish to know if the reporter was "consciously aware" of a risk. The reporter may be one of our sources of information about his awareness. The reporter must decide what to tell us. How does the reporter decide what to report?

It is possible -- no? -- that the reporter will reason with himself (before reporting), "How much did my state of mind partake of awareness? How much did it have that quality?"

Here our focus is on the reporter's -- i.e., the witness' -- actual belief about his own mental state or condition. We can ponder later how & whether the reporter decides to report what he [thinks he] believes about his mental state or states.

Or is this possibility impossible? Must the reporter believe, "Either I was aware or not aware of a risk. I must decide which, on balance, I more probably was".?

Or can the reporter think, "It's not a question of one or the other. It's a question of the degree to which I was aware, of the degree to which I had conscious awareness. In most instances, even with complete information, I will not be able to say, with fidelity, that either I had conscious awareness or that I didn't have it. The most accurate way for me to report the state of my awareness is for me to report my belief about how much awareness I had."

By posing this thought-experiment involving a reporter-actor ruminating about the form of his report of his own mental processes, I [Peter Tillers] am not ruling out a probans separate from the probandum: I am not saying that evidence in the [modern] classic sense is useless when the matter in issue is some mental state or process. For example, even if I am a reporter-actor I may want to have some evidence of the occurrence or non-occurrence of some thoughts or notions inside my head at some earlier time -- because, for example, my memory may be frail and I may not remember what I thought or felt at some earlier time; I might want, therefore, to look at my diary if I have one. Similarly, an outside observer -- a trier of fact, a jury, for example -- may have to decide whether to believe that a reporter, a witness, is attempting or not attempting to tell the truth as the reporter sees it. In these situations the only question is not how someone should characterize the thoughts and feelings that are thought to have occurred in someone's head, or brain.

*** Why do fuzzy predicates (sometimes) work? Why do fuzzily-measured linguistic variables work? For example, why are they sometimes great as instruments of control? Compare Judea Pearl's approach, which emphasizes that any understanding that intervenes in the world must be an understanding that sees the causal connections among events -- and, for that reason principally, evidence must consist of events that have a discernible causal connection with the matters or hypotheses in issue? There is none of that in Zadeh's version of fuzzy sets. We control the world by working with it at a kind of surface level!

*** Where do we go from here? What is the issue?

Perhaps there are three basic forms of evidence (or signs):

(i) an element is a sign or evidence because it stands in a discernible causal relationship or relationships to pertinent hypotheses;

(ii) an element functions as evidence because it is associated with pertinent hypotheses in some fashion; and

(iii) an event or state serves as a sign of a hypothesis because the event or state partakes or seems to partake of a possible defined state -- and here an event is not so much "evidence" as it is a condition that absorbs or gathers the status of some hypothetical state to some degree, and the thinking or perception here is quasi-Platonic, with the proviso that in Zadeh's world one thinks of events as having varying amounts, or partaking in varying degrees, in names and concepts rather than in being.

This third line of thinking, if it is not entirely absurd, necessarily leads to another category of questions, a category that Zadeh explores (though how successfully for my purposes -- for an understanding of law's reasonings -- I cannot yet tell): we now wish to know to what extent we can reason about our sense of how far an event or set of events -- e.g., "mental states" -- partakes in a concept such as "awareness." Is the judgment about awareness akin to a primitive and irreducible perception -- or is the judgment about a matter such as "awareness" in some (fuzzy) sense an "inference," albeit a fuzzy inference? For example, perhaps the judgments about (degrees of)awareness are and ought to be influenced by matters such as the duration of certain kinds of mental processes and the amount of feeling that accompanied some mental state or states.

But if we are to pursue this kind of inquiry -- and where it will lead I do not know -- it is crucial to play a certain mental trick. The trick is to recognize that a relationship R between events such m and possible states such as A is not always to be resolved by further analysis or elaboration of a concept or notion such as A, "conscious awareness." If we are to think in a Zadeh-like fashion about the implications of E (some item of evidence) for the degree to which m (some mental process) is A (conscious awareness), it is crucial to keep in mind that the appropriateness of seeing some m in A and the question of the extent to which m should be thought to be in A are not exhaustively specified by the definition of A; we must assume that factors outside of A, so to speak, influence the degree of m's membership in A.

But if any of this makes sense, what could a fuzzy analysis of legal reasoning and factual inference in legal settings accomplish? This is a very big question. Perhaps fuzzy analysis would have predictive value. See the work by L. Philipps. But perhaps the theory of fuzzy sets in this connection (factfinding in law) could also serve to reassure us that the way that we generally expect to resolve issues such as "D's recklessness vel non" or "D's conscious awareness vel non" is not irrational.

Consider this specimen phenomenon and problem:

Walter Witness believes and reports: "The thought that my car might skid flitted through my mind. The thought gripped me, but, as I said, it only flitted through my mind. I had other things on my mind, mainly my wife's illness and my bad day at the office. They were in the forefront of my thinking, I was preoccupied with those things."

How should we decide and how do we decide whether or not and to what extent Walter was reckless and consciously aware of a risk that his car might skid? *** Reference material:

Knapik v. Ashcroft, 2004 U.S. App. LEXIS 19445 (2nd Cir., Sept. 17, 2004) (Review of order for deportation of a lawful permanent resident; defendant pled guilty to the crime of attempted reckless endangerment; Knapnik's plea and conviction arose from "an incident in which, while intoxicated, Knapik drove at an excessive rate of speed against the flow of traffic on the Staten Island Expressway"; the question on appeal was whether defendant's conviction was a conviction of a crime of "moral turpitude"; New York statute defined reckless endangerment thus: "A person is guilty of reckless endangerment in the first degree when, under circumstances evincing a depraved indifference to human life, he recklessly engages in conduct which creates a grave risk of death to another person."; the court noted that in an earlier and different proceeding involving Illinois law, the BIA found persuasive that under Illinois law a "person acting recklessly must consciously disregard a substantial and unjustifiable risk, and such disregard must constitute a gross deviation from the standard of care which a reasonable person would exercise in the situation. This definition of recklessness requires an actual awareness of the risk created by the criminal violator's action."; the court of appeals further observed that "the BIA [Board of Immigration Appeals] limits moral turpitude to crimes in which a defendant consciously disregards a substantial risk of serious harm or death to another"; but appellant was convicted of attempted reckless endangerment, and the federal court of appeals held that the concept of attempted reckless endangerment is "nonsensical" -- because, it reasoned, a reckless act is unintentional and one cannot intend to commit an unintentional act -- and that a conviction for such a crime with such oxymoronic elements does not support the conclusion that every person convicted of such a "nonexistent" crime was convicted of a crime of "moral turpitude").

A Unified Theory of Uncertainty?

Lotfi Zadeh, Abstract of "Toward a Unified Theory of Uncertainty," a lecture presented at the International Conference on Information Processing and Management of Uncertainty, Perugia, Italy, July 4-7, 2004:

In science, and especially within [the] probability community, it is an almost universally held view that probability theory is the theory of uncertainty, and that no other theory of uncertainty is needed. ...

...

What is proposed here is what may be called a unified theory of uncertainty, or UTU, for short. In this theory, the point of departure is the concept of partiality--a concept which has a position of centrality in human cognition. Thus, in human cognition almost everything is partial, that is, is a matter of degree. For example, we have partial knowledge, partial understanding, partial truth, partial certainty, partial possibility, partial belief, partial causality, partial information, partial preference, partial independence and partial satisfaction. In the unified theory of uncertainty, there are three partialities that stand out in importance, (a) partiality of certainty (likelihood); partiality of truth (verity); and (c) partiality of possibility.

Lotfi Zadeh, "Toward a Unified Theory of Uncertainty--from PT to UTU," PowerPoint Version of lecture presented at Perugia, Italy, July 4, 2004, at p. 159:

The range of application-areas of fuzzy logic is too wide for exhaustive listing. Following is a partial list of existing application-areas in which there is a risk of substantial activity.

1. Industrial control
2. Quality control
3. Elevator control and scheduling
4. Train control
5. Traffic control
6. Loading crane control
7. Reactor control
....
20. Assessment of credit-worthiness
21. Fraud detection
...
23. Pattern classification
...
34. Internet
35. Library and Information science

Blurb by John P. Burgess on back of Susan Haack, Deviant Logic, Fuzzy Logic (U. Chicago Press, 2nd ed., 1996): "Given the amount of media hype 'fuzzy logic' has received, I am pleased by how informatively and entertainingly Dr. Haack writes in debunking it."

Susan Haack, "Do We Need Fuzzy Logic," in Susan Haack, Deviant Logic, Fuzzy Logic 233 (U. Chicago Press, 2nd ed., 1996): "I also want to raise a question: which of the many applications claimed to the credit of fuzzy logic are in fact applications of the base logics [which, she argues, are not part of fuzzy logic], and which of the more radical systems? It would require a more thorough search of the literature than I have been able to undertake to settle the issue; but I should expect, if my criticisms of fuzzy logic are correct, to find that it is the base logics that have been given practical applications."

Stephen Wolfram, A New Kind of Science 1175 (2002): "The idea of intermediate truth values has been discussed intermittently ever since antiquity. Often--as in the work of George Boole in 1847--a continuum of values between 0 and 1 are taken to represent probabilities of events, and this is the basis for the field of fuzzy logic popular since the 1980s."

Joseph Y. Halpern, Reasoning about Uncertainty Section 2.5 at pp. 40, 42-43 (2003):

Possibility measures are yet another approach to assigning numbers to sets. They are based on ideas of fuzzy logic. Suppose for simplicity that W, the set of worlds, is finite and that all sets are measurable. A possibility measure Poss associates with each subset of W a number in [0, 1] and satisfies the following three properties:

Poss1. Poss ([symbol for null set])= 0.

Poss2. Poss(W) = 1.

Poss3. Poss(U v V) = max(Poss(U), Poss(V) if U and V are disjoint.

...

...Perhaps the most common interpretation given to possibility and necessity is that they capture, not a degree of likelihood, but a (subjective) degree of uncertainty regarding the truth of a statement. This is viewed as being particularly appropriate for vague statements such as "John is tall." Two issues must be considered when deciding on the degree of uncertainty appropriate for such a statement. First, there might be uncertainty about John's actual height. But even if an agent knows that John is 1.78 meters tall ..., he might still be uncertain about the truth of the statement "John is tall."

Criminal Evidence Redux

I repeat the following recommendation to prevent it from being buried in a blizzard of blogs:

If you are interested in the UK way of proving and disproving criminal guilt, you must purchase this book:

Paul Roberts & Adrian Zuckerman, Criminal Evidence (2004), 712 pages, Oxford University Press.

This tome is not merely erudite. It is also well written.

Is a Criminal Defendant's Off-the-Stand Courtroom Demeanor Irrelevant?

Bryant v. Maryland, 129 Md. App. 150; 741 A.2d 495 (1999) (murder trial; defendant did not testify):

During closing argument prosecutor said,

There is so much evidence that corroborates what [the prosecution's witness] told you. When I spoke about her demeanor when she testified, and how she answered [defense counsel's] questions, did you notice the defendant's demeanor when she testified, the way he kept looking down and couldn't look at her? She looked in his eyes several times.

...

... You observed that, members of the jury, you were sitting here. We all saw it. He couldn't sit up and look her in the eye because he knew she was telling the truth. He knew she was telling the truth.

The Maryland Court of Special Appeals said:

Argument that asks the jury to consider the demeanor of a witness when testifying is proper and is consistent with the jury instruction given in this case to consider "the witness's behavior on the stand and way of testifying; did the witness appear to be telling the truth." Argument that comments on the courtroom demeanor of a defendant who elects not to testify is a different matter. Courts that have considered this question have reached different conclusions about when, if ever, comment on a defendant's courtroom demeanor is proper. In State v. Rivera, 602 A.2d 775, 253 N.J. Super. 598, (1992), the court held that when a defendant engages in "testimonial behavior before a jury" by injecting unsworn comments into a trial by word, gestures, display of emotion, or other demeanor intended to influence the jury, the prosecutor may, with advance approval of the court and the making of a record of the defendant's conduct, make a limited argument noting the fact of the behavior and that the comment or demeanor should not be considered by the jury. The court further stated, however, a prosecutor may not comment upon the failure of a defendant to act in a particular way during a trial. Id. at 777. See also State v. Johnson, 576 A.2d 834, 851-52, 120 N.J. 263 (1990) (improper for prosecution to argue that defendant failed to make eye contact with jurors during trial.)

...

The Supreme Court of Delaware, in Hughes v. State, 437 A.2d 559 (Del. Supr. 1981), held improper the prosecutor's comments that characterized a non-testifying defendant's courtroom demeanor as unemotional, unfeeling, and without remorse.

In our view, the courtroom demeanor of a defendant who has not testified is irrelevant. His demeanor has not been entered into evidence and, therefore, comment is beyond the scope of legitimate summary. ... Moreover the practice is pregnant with potential prejudice. A guilty verdict must be based upon the evidence and the reasonable inferences therefrom, not on an irrational response which may be triggered if the prosecution unfairly strikes an emotion in the jury.

437 A.2d at 572. See also Pope v. Wainwright, 496 So. 2d 798, 802 (Fla. 1986) (comments on a defendant's demeanor off the witness stand clearly improper); Blue v. State, 674 So. 2d 1184, 1213-15 (Miss. 1996) (error for prosecutor to comment on non-testifying defendant's demeanor and appearance during trial); People v. Garcia, 160 Cal. App. 3d 82, 206 Cal. Rptr. 468, 472-75 (Cal. App. 1984) (prosecutor's references to defendant's courtroom behavior was improper).

On the other hand, the Supreme Judicial Court of Massachusetts has held that a prosecutor's comment on the defendant's squirming, smirking, and laughing during trial was fair comment and in context did not suggest that the prosecutor had knowledge the jury did not share. Commonwealth v. Smith, 387 Mass. 900, 444 N.E.2d 374, 380 (Mass. 1983). That court has said, however, that a suggestion by the prosecutor that normal courtroom behavior betrays consciousness of guilt is improper. Commonwealth v. Valliere, 366 Mass. 479, 321 N.E.2d 625, 635 (Mass. 1974). And, see Commonwealth v. Pullum, 22 Mass. App. Ct. 485, 494 N.E.2d 1355, 1358 (Mass. App. 1986).

The Supreme Court of North Carolina found no error in a prosecutor's comments on the courtroom demeanor of the defendant, holding that such remarks were rooted in the evidence and that the demeanor of the defendant was before the jury at all times. State v. Brown, 320 N.C. 179, 358 S.E.2d 1, 15 (N.C. 1987), cert. denied, 484 U.S. 970, 98 L. Ed. 2d 406, 108 S. Ct. 467 (1987). See also State v. Myers, 299 N.C. 671, 263 S.E.2d 768, 773-74 (N.C. 1980); Wherry v. State, 402 So. 2d 1130, 1133 (Ala. Cr. App. 1981) (conduct of the accused or the accused's demeanor during the trial is a proper subject of comment, at least when insanity is raised as a defense).

...

Turning to the case before us, we conclude that the argument of the prosecutor concerning the alleged failure of this defendant to look at a witness and the inference of guilt that should be drawn from that conduct was improper. The trial judge therefore erred in failing to sustain the defendant's timely objection.

The defendant in this case did not testify, and the State's argument related only to this courtroom conduct. Nowhere in the record is there to be found any reference to the alleged conduct of the accused. The prosecutor was arguing a fact not in evidence, and compounding that by adding her personal assurance that the alleged conduct occurred by saying, "We all saw it." n2 [footnote omitted] The argument related to conduct of the accused that was entirely passive -- his alleged failure to "look [the witness] in the eye." The argument was not, therefore, a comment on intentional conduct of an accused calculated to influence the jury. Moreover, the prosecutor argued her conclusion that the defendant's failure to "look her in the eye" was evidence of guilt -- "He knew she was telling the truth." -- a questionable inference at best. There may be any number of reasons why a defendant will not fix his or her gaze upon a witness, including a possible earlier instruction by defense counsel to avoid any possible implication that the defendant is attempting to intimidate or "stare down" a witness. The prosecutor should have focused on evidence that was before the jury, which may include fair comment on the demeanor of witnesses while they are on the stand, but which will ordinarily not include comments on the courtroom demeanor of the defendant.

In Wilhelm v. State, 272 Md. 404, 326 A.2d 707 (1974), the Court of Appeals noted that counsel may "make any comment or argument that is warranted by the evidence proved or inferences therefrom" and that the prosecutor is free "to comment legitimately and to speak fully, although harshly, on the accused's action and conduct if the evidence supports his comments . . . ." Id. at 412. We do not understand this statement ordinarily to condone comments of the prosecutor on the passive courtroom demeanor of a non-testifying defendant. ...

The Laughter of Non-Testifying Criminal Defendants

United States v. Schuler, 813 F.2d 978 (9th Cir., 1987):

The prosecutor's comment in closing argument on Schuler's laughter during testimony about the threats Schuler made presents procedural and substantive questions. The procedural difficulty derives from the fact that there is nothing in the record indicating that the laughter did occur, although Schuler nowhere objects that it did not. It is very difficult for an appellate court to review an issue that is not grounded in the record. If counsel considers such an outbreak to be significant, he or she should ask the trial court to have it included in the record.

Assuming that the laughter did occur, we are faced with the substantive issue whether the failure to exclude the prosecutor's reference to Schuler's courtroom behavior constituted reversible error. This issue involves several facets, including whether such remarks (1) introduce character evidence solely to prove guilt, (2) violate a defendant's fifth amendment right not to be convicted except on the basis of evidence adduced at trial, and (3) violate fifth amendment rights by indirectly commenting on a defendant's failure to testify at trial. ...

The Ninth Circuit has not ruled whether it is error to permit a prosecutor to comment on a defendant's courtroom behavior. Several other circuits, however, have held that such comment is reversible error and we find the rationale of those cases compelling. n1

n1 Two circuits have rejected a challenge to a prosecutor's comments on the expressionless courtroom demeanor of a defendant in a habeas corpus context. See Borodine v. Douzanis, 592 F.2d 1202, 1210-11 (1st Cir. 1979); Bishop v. Wainwright, 511 F.2d 664, 668 (5th Cir. 1975), cert. denied, 425 U.S. 980, 96 S. Ct. 2186, 48 L. Ed. 2d 806 (1976). Both cases perceived the major issue to be whether such comments necessarily reflected on the defendant's failure to take the witness stand and both concluded that the comments did not necessarily allude to the failure to testify. We do not disagree with that specific holding. We do not accept, however, any broader implication from those cases that references to a non-testifying defendant's demeanor or behavior present no constitutional issue.

Federal Rule of Evidence 404(a) prohibits the introduction of evidence of the character of the accused solely to prove guilt. The prosecutor's remarks in effect did this by suggesting to the jury that Schuler's laughter was relevant apparently for the purpose of showing that he was of bad character because he considered the charges of threatening the life of the President to be a joke. ... While the offense of threatening the life of the President requires proof of a "knowing and willful" act, 18 U.S.C. § 871, and to that extent involves proof of Schuler's mental state, his laughter at trial could not have any relevancy for that purpose. n2 His courtroom behavior off the witness stand was legally irrelevant to the question of his guilt of the crime charged.

n2 Insanity was not an issue at the trial. There are appropriate legal proceedings for restraining mentally ill individuals who are a danger to others. If Schuler was, at the time of trial, a danger to the President or others, California law provides for evaluation and involuntary commitment of individuals who constitute a danger to themselves or others. See Cal. Welf. & Inst. Code § 5150 (West 1984).

The District of Columbia Circuit, when faced with a similar situation, reversed the conviction based in part on the prosecutor's reference to the defendant's courtroom behavior and the district court's refusal to instruct the jury that such behavior must not be considered. United States v. Wright, 160 U.S. App. D.C. 57, 489 F.2d 1181 (D.C. Cir. 1973). The court stated:

Unless and until the accused puts his character at issue by giving evidence of his good character or by taking the stand and raising an issue as to his credibility, the prosecutor is forbidden to introduce evidence of the bad character of the accused simply to prove that he is a bad man likely to engage in criminal conduct. . . .

This basic principle cannot be circumvented by allowing the prosecutor to comment on the character of the accused as evidenced by his courtroom behavior. That the jury witnesses the courtroom behavior in any event does not make it proper for the prosecutor to tell them, with the court's approval, that they may consider it as evidence of guilt.

Id. at 1186 (citations omitted). We agree that such comments are improper. ...

We also conclude that, in the absence of a curative instruction from the court, a prosecutor's comment on a defendant's off-the-stand behavior constitutes a violation of the due process clause of the fifth amendment. That clause encompasses the right not to be convicted except on the basis of evidence adduced at trial. The Supreme Court has declared that "one accused of a crime is entitled to have his guilt or innocence determined solely on the basis of the evidence introduced at trial, and not on grounds . . . not adduced as proof at trial." Taylor v. Kentucky, 436 U.S. 478, 485, 56 L. Ed. 2d 468, 98 S. Ct. 1930 (1978). We have recognized that a prosecutor may not seek to obtain a conviction by going beyond the admissible evidence. See United States v. Schindler, 614 F.2d 227, 228 (9th Cir. 1980). The Eleventh Circuit in United States v. Pearson, 746 F.2d 787 (11th Cir. 1984), was confronted with a prosecutor's closing argument commenting on the defendant's behavior off the witness stand, as follows:

Does it sound to you like he was afraid? You saw him sitting there in the trial. Did you see his leg going up and down? He is nervous. (Appellant's objection overruled) You saw how nervous he was sitting there. Do you think he is afraid?

Id. at 796. The court found such comments were reversible error, adding that

in overruling [the defendant's] objection and in failing to give a curative instruction, the court, in effect, gave the jury an incorrect impression that [the defendant's] behavior off the witness stand was evidence in this instance, upon which the prosecutor was free to comment.

Id. We agree with the holding of the Eleventh Circuit that such comments in the absence of a curative instruction, constitute a deprivation of the fifth amendment right to a fair trial. See Strickland v. Washington, 466 U.S. 668, 684-85, 80 L. Ed. 2d 674, 104 S. Ct. 2052 (1984); Estell v. Williams, 425 U.S. 501, 503, 48 L. Ed. 2d 126, 96 S. Ct. 1691 (1976). n3

n3 The dissent, to support its position that a defendant's courtroom demeanor is evidence, relies principally on cases where the defendant took the stand. ... When a defendant chooses to testify, a jury must necessarily consider the credibility of the defendant. In this circumstance, courtroom demeanor has been allowed as one factor to be taken into consideration. The dissent has also referred to a case involving a judge's observation of a defendant's demeanor during the sentencing phase of trial. ... At this stage different considerations, including general character, are relevant. The above examples obviously contrast with the present situation where the defendant in the trial phase asserted his right not to testify and where neither his credibility nor his character were relevant considerations.

In addition, prosecutorial comment on a defendant's non-testimonial behavior may impinge on that defendant's fifth amendment right not to testify. We do not accept Schuler's contention that such comments in every case violate the right to remain silent because jurors would naturally take them to be comments on the failure of the accused to testify. ... In fact, we doubt that jurors would construe the prosecutor's comment on Schuler's laughter as referring to his failure to testify. We are concerned, however, that such statements by the prosecutor during trial, or the fear of such statements in closing argument, will tend to eviscerate the right to remain silent by forcing the defendant to take the stand in reaction to or in contemplation of the prosecutor's comments. In effect the defendant would be compelled to testify to explain any actual or possible behavior that the prosecutor might bring to the jury's attention. While this pressure to testify may well be the exception, there is no reason for use of such comments that would justify even a slight opening of the door to an invasion of constitutional rights.

...

When, as here, the prosecutor describes the courtroom behavior of a defendant who has not testified, and then goes on to tell the jury that it may consider that behavior as evidence of guilt, the prosecutor violates those rights.

Hall, J., dissenting:

The majority holds today that a prosecutor's comment in closing argument on the defendant's laughter during the testimony of other witnesses is reversible error. Because I believe that the courtroom demeanor of the defendant in this case is evidence that the jury could properly consider, I respectfully dissent from Section II.A of the majority's opinion.

I

The principle that a defendant's courtroom demeanor is evidence is well-settled. In Russell v. United States, 288 F.2d 520 (9th Cir. 1961), cert. denied, 371 U.S. 926, 83 S. Ct. 296, 9 L. Ed. 2d 234 (1962), this court observed that a trial court is in a much better position than an appellate court to determine whether the evidence in a particular case warrants a new trial. We noted that

the trial judge can see in the conduct and demeanor of one who testifies (or even in the conduct and demeanor of defendants who never take the witness stand, as the defendants did not here) a thousand and one matters impossible for a reviewing court to glean from a printed page.

Id. at 522. Likewise, in Reagan v. United States, 157 U.S. 301, 39 L. Ed. 709, 15 S. Ct. 610 (1895), the Supreme Court quoted with approval instructions that specifically advised a jury to consider the defendant's "demeanor and conduct upon the witness stand and during the trial." Id. at 308 ... . n1 See also Waller v. United States, 179 F. 810, 812 (8th Cir. 1910).

n1 Unlike the defendant in Reagan, Schuler chose not to testify. However, the Russell case, like this case, involved a non-testifying defendant. This court held that the demeanor of such a non-testifying defendant was relevant evidence.

Sound policy reasons exist for allowing a jury to consider the courtroom demeanor of a defendant. As Wigmore noted: "It is as unwise to attempt the impossible as it is impolitic to conduct trials upon a fiction; and the attempt to force a jury to become mentally blind to the behavior of the accused sitting before them involves both an impossibility in practice and a fiction in theory." 2 J. Wigmore, Evidence § 274 (J. Chadbourn rev. ed. 1979). n2 Taken to its logical conclusion, the majority's opinion would require a court to instruct a jury to reach its verdict as if the accused had not been present before it.

n2 See also H. Underhill, Criminal Evidence § 125 n.9 (5th ed. 1956) ("Practically it is impossible to prevent jurors from observing the appearance and behavior of the accused very closely while he is in court during the trial. They will naturally draw inferences therefrom either favorable or unfavorable to him. The information thus obtained is evidence, and, doubtless, many a verdict has been determined thereby.").

The majority relies on United States v. Wright, 160 U.S. App. D.C. 57, 489 F.2d 1181 (D.C.Cir. 1973), United States v. Carroll, 678 F.2d 1208 (4th Cir. 1982), and United States v. Pearson, 746 F.2d 787 (11th Cir. 1984), for the proposition that it is reversible error for a prosecutor to comment on a defendant's courtroom behavior. To the extent that these cases stand for the broad principle that a defendant's courtroom demeanor is never relevant evidence, I believe they were incorrectly decided. However, a careful reading of the cases demonstrates that they are simply restating the standard contained in Fed. R. Evid. 404(a), namely, that "evidence of a person's character . . . is not admissible for the purpose of proving that he acted in conformity therewith on a particular occasion."

In Wright, the defendant was charged with bank robbery. During his closing argument, the prosecutor commented on the defendant's unruly courtroom behavior. The District of Columbia Circuit held that the defendant's behavior off the witness stand was not "legally relevant to the question of his guilt or innocence of the crime charged." 489 F.2d at 1186 ... . Similarly, in Carroll, a bank robbery case, the prosecutor commented on the fact that the defendant had examined bank surveillance photographs with his attorney during trial. The Fourth Circuit held that a defendant's courtroom demeanor could not be introduced "solely to prove guilt." 678 F.2d at 1210 ... . n3 Finally, in Pearson, the defendant was charged with possessing an unregistered silencer and possessing a silencer without a serial number. In closing argument, the prosecutor commented on the defendant's nervousness off the witness stand. The Eleventh Circuit held that "the defendant's behavior off the witness stand in this instance was not evidence subject to comment." 746 F.2d at 796 ... .

n3 The court in Carroll also relied on the fact that the prosecutor's comment violated both the defendant's right not to testify and his right to assist his counsel in his own defense. 678 F.2d at 1209. Neither of these factors is present here. The majority concedes it is doubtful that the jurors construed the prosecutor's comment on Schuler's laughter as referring to his failure to testify. Moreover, it is obvious that Schuler's laughter was not an attempt to assist his counsel in his own defense.

The majority incorrectly assumes that Schuler's laughter is evidence of his character offered solely to prove guilt. In fact, Schuler's laughter is relevant to another material issue in the case, Schuler's mental state. As such, the laughter is admissible under Fed. R. Evid. 404(b). n4 Cf. Commonwealth v. Louraine, 390 Mass. 28, 34, 453 N.E.2d 437, 442 (1983) ("It is an established and universally accepted rule that, when a defendant's sanity is at issue, the trier of fact is entitled to consider the defendant's demeanor in court.").

n4 Fed. R. Evid. 404(b) provides as follows ... :

Evidence of other crimes, wrongs, or acts is not admissible to prove the character of a person in order to show that he acted in conformity therewith. It may, however, be admissible for other purposes, such as proof of motive, opportunity, intent, preparation, plan, knowledge, identity, or absence of mistake or accident.

Fed. R. Evid. 404(b) provides for the admission of other acts of the defendant in order to prove that he had the requisite intent for a crime. In this case, Schuler's intent at the time of the alleged threat was at issue; the defense asserted that Schuler's remark was an expression of anger at law enforcement personnel and not a serious threat to the President. Schuler's behavior in the courtroom was relevant evidence on the issue of intent and, hence, was admissible under Fed. R. Evid. 404(b).

The Ninth Circuit has consistently upheld the admission into evidence of other acts of a defendant where, as here, that defendant's mental state is at issue. For example, in United States v. Hearst, 563 F.2d 1331 (9th Cir. 1977), cert. denied, 435 U.S. 1000, 56 L. Ed. 2d 90, 98 S. Ct. 1656 (1978), this court affirmed the conviction of Patricia Hearst for robbing a bank in San Francisco. At trial, the defense claimed that Hearst was acting under duress at the time she committed the robbery. The trial court admitted evidence of subsequent criminal activities of the defendant in Los Angeles which undercut the defense. This court upheld the admission of the evidence under Fed. R. Evid. 404(b), noting that

appellant's state of mind during the San Francisco robbery was the central issue in the case. State of mind is usually difficult to prove, and the evidence on the issue was sharply divided. The timing and other circumstances of the Los Angeles incidents made evidence of them highly probative on this critical issue.

563 F.2d at 1337.

Likewise, in United States v. McCollum, 732 F.2d 1419 (9th Cir.), cert. denied, 469 U.S. 920, 83 L. Ed. 2d 236, 105 S. Ct. 301 (1984), this court affirmed the conviction of the defendant for bank robbery. At trial, the defense claimed that the defendant was acting under hypnosis at the time of the robbery. This court upheld the trial court's admission into evidence of a twelve year old conviction for armed robbery under Fed. R. Evid. 404(b): "Such a prior act can be probative of intent because the fact that the defendant had an unlawful intent at the time he committed the extrinsic offense makes it less likely that he had a lawful intent when he performed the acts charged as the present offense." 732 F.2d at 1424. We then proceeded to state the general rule of law as follows: "Where the mental state to be inferred from undisputed overt acts of a defendant is the crucial issue, evidence of past criminal acts has generally been found insufficiently prejudicial to warrant exclusion." Id. at 1425.

In this case, Schuler's bizarre behavior in the courtroom lends credence to the government's claim that his threat was indeed serious. Moreover, the admission of Schuler's laughter is certainly less prejudicial than the admission of the criminal acts in Hearst and McCollum. The trial court did not abuse its discretion in allowing the jury to consider Schuler's laughter. n5 [footnote omitted]

.

Courtroom Threat

Michaelangelo Conte, "Baby murder defendant threatens 'snitch' witness in court,"The New Jersey Journal A8 (Sept. 17, 2004):

Jurors were walking from the courtroom when Alvarado [the defendant in a murder trial] turned to prosecution witness Francisco Escribaro [a former cellmate of the the defendant], raised [Escribaro's] written statement in [defendant's] hand, and said, "I am going to make sure people in jail read this because you are a (expletive) snitch."

...

[Trial judge] Nelson told the jury that sometimes attorneys or other participants involved in a trial get excited or angry.

"If it happened in this trial and you heard it, you have to disregard it," he said. "Next week if it's still in the back of your mind, it can play no part in your decision. Everyone understand that?"

But cf. United States v. Mickens et al., 926 F.2d 1323 (2nd Cir. 1991)(held, admission of defendant's courtroom threat against former attorney was not error):

Defendant-appellant Mickens challenges the district court's decision to permit Mickens' former attorney to testify that Mickens had made a hand gesture in the shape of a gun as the former attorney entered the courtroom to testify. Mickens argues that this testimony lacked probative value and that the former attorney's testimony that Mickens had pointed at the court, not at the attorney, was unduly prejudicial. This challenge lacks merit.

...

[T]he standards for admission of Rule 404(b) evidence were satisfied. The testimony about the hand gesture was not offered to prove Mickens' bad character or criminal propensity, but rather to prove his consciousness of guilt. ... The testimony was relevant since an effort to intimidate a key prosecution witness was probative of Mickens' state of mind. ...