Thursday, June 22, 2006

A Familiar but Unresolved(?) Puzzle: Substantive Definitions of Crimes versus Proof Beyond Reasonable Doubt

A very recent Supreme Court decision -- Dixon v. United States, No. 05-7503 (June 22, 2006) -- serves as a reminder that three principles are well-entrenched in U.S. constitutional jurisprudence:
1. The prosecution must prove all essential elements of the crime charged beyond a reasonable doubt. This is required by due process.

2. It is constitutionally permissible for a State or the U.S. Government to make a criminal defendant bear the burden of proving that an affirmative defense applies: due process does not require that the government prove beyond a reasonable doubt that an affirmative defense to a criminal charge does not apply.

3. Legislative bodies (including Congress) have the authority to define crimes. Legislative bodies are therefore free to decide which elements are essential elements of a given crime. Similarly, legislative bodies are free to proclaim that the truth or falsity of a given element -- a given type of fact -- is part of an affirmative defense rather than part of the definition of a given crime.

Propositions #2 and #3 perhaps reflect the notion that legislative bodies are free to choose how a concern for public security and the concern for the role of blame in criminal justice are to be balanced. But if legislative bodies are to have this prerogative, why are they not also free to proclaim that conviction of crime can be had even if the essential elements of crimes -- crimes that those same legislative bodies have defined -- have not been established beyond a reasonable doubt? If the constitution recognizes that legislative bodies are entities that have the constitutional authority to balance the concern for public safety and the concern for protection of the innocent, why aren't those same legislative bodies free to decide how much proof there must be of the essential elements that those legislative bodies have decided should be part of the definition of the crime in question?

I doubt that legislative authority over legal definitions of criminal culpability is compatible with the principle or rule that criminal guilt must be established beyond a reasonable doubt. (In Dixon the Court seems to assume that an accused's mere [pre-existing] knowledge of the illegality of the conduct undertaken ["perpetrated"?] by the accused is sufficient to establish whatever degree of criminal culpability the constitution may demand. But I cannot imagine that even the most hard-hearted member of the Supreme Court will adhere to this harsh principle when push comes to shove -- when, for example, the knowledge of the accused is accompanied by extreme extenuating circumstances such as duress on the accused in the form of extreme physical torture.)

  • The riddle I have sketched above remains fundamentally unchanged if courts (as well as legislative bodies) are thought to have authority to define crimes and affirmative defenses.
  • The tension between the reasonable doubt principle and legislative power to define crimes reawakens in my mind the suspicion that the reasonable doubt principle is more show than substance. I have the suspicion that the reasonable doubt standard is more show than substance because I simply do not understand why the reasonable doubt standard does not itself imply or generate a constitutional obligation on the part of government to structure the entire criminal process (including police investigation) in a way that provides reasonable assurance that convictions will be had only when there is a very high probability of guilt.

    Sunday, June 18, 2006

    Bayesian Spam

    The cleverness of spammers knows no bounds. Today I received an e-mail message from the following sender:

    Bays H. Network

    P.S. I regret to say that the message did not contain a network. What a fraud!

    What Does That Darned Picture Mean?

    Is the following statement true or (better yet!) when is it true (and not true)?:
    If a mental picture is used to represent a thought, it needs to be accompanied by a caption, by a set of instructions for how to interpret the picture--what to pay attention to and ignore.
    Steven Pinker, How the Mind Works pp. 297 (1997).

    It is the case, of course, that instructions for use can be remembered and that people can learn to recall almost instantaneously -- very quickly -- the intended use or meaning of at least some images. Consider (some) highway signs containing pictures. Or, chess players, consider a diagram of a chess position (with icons representing the King, Bishop, etc.).

  • Indeed, my friends, are physical chess pieces different in any pertinent way from "immaterial" icons representing chess pieces? "Real" chess pieces and "real" chess boards also don't usually carry captions describing how chess pieces can be played.
  • Pictures and Thinking

    "My computer screen is festooned with little cartoons that do various things when selected by a click of the mouse. For the life of me I can't remember what the tiny binoculars, eyedropper, and silver platter are supposed to do. A picture is worth a thousand words, but that is not always such a good thing. At some point between gazing and thinking, images must give way to ideas." Steven Pinker, Chapter 4 ("The Mind's Eye"), How the Mind Works p. 298 (1997).

    This thought has many implications for visualizing evidence. One implication is the following:

    The mental labor required by a method of picturing evidence and inference ordinarily should not outweigh the cognitive benefits of using such that method.
    Compare the following lead balloon produced by the ordinarily-superlative John Henry Wigmore: Even though Wigmore's method of picturing inference was far from user-friendly, it is good that Wigmore created his leaden method of charting evidence and inference. For example, his effort helped to get some legal professionals to start thinking about inference networks and it eventually helped to get probabilists to think more carefully about cascaded probabilistic inference and about the peculiar and interesting properties of complex evidential inference.

    Once Again: What Is "Visualization" of Evidence and Inference?

    Oliver Sacks, A Neurologist's Notebook: To See and Not See, THE NEW YORKER (May 10, 1993, posted online June 12, 2006):
    The seventeenth-century philosopher William Molyneux, whose wife was blind, posed the following question to his friend John Locke: “Suppose a man born blind, and now adult, and taught by his touch to distinguish between a cube and a sphere [be] made to see: [could he now] by his sight, before he touched them . . . distinguish and tell which was the globe and which the cube?” Locke considers this in his “Essay Concerning Human Understanding” (1690) and decides that the answer is no. In 1709, examining the problem in more detail, and the whole relation between sight and touch, in “A New Theory of Vision,” George Berkeley concluded that there was no necessary connection between a tactile world and a sight world—that a connection between them could be established only on the basis of experience.

    Barely twenty years elapsed before these considerations were put to the test—when, in 1728, an English surgeon named William Cheselden removed the cataracts from the eyes of a thirteen-year-old boy born blind. Despite his high intelligence and youth, the boy encountered profound difficulties with the simplest visual perceptions. He had no idea of distance. He had no idea of space or size. And he was bizarrely confused by drawings and paintings, by the idea of a two-dimensional representation of reality. As Berkeley had anticipated, he was able to make sense of what he saw only gradually, and insofar as he was able to connect visual experiences with tactile ones. It had been similar with many other patients in the two hundred and fifty years since Cheselden’s operation: nearly all had experienced the most profound, Lockean confusion and bewilderment.

    And yet, I was informed, as soon as the bandages were removed from Virgil’s eye he saw his doctor and his fiancĂ©e, and laughed. Doubtless he saw something—but what did he see? What did “seeing” for this previously not-seeing man mean? What sort of world had he been launched into?

    This tale has implications for the conference on Graphic and Visual Representations of Evidence and Inference in Legal Settings.