When you think about it, it is remarkable that the very people who, in the so-called enlightenment claimed they were freeing the world from such things as the occult, fortune-tellers divining the future and peoples' character and fate from the stars, lines on their palms, card reading, etc. very soon tried to do the same thing using a modified numerology that had the merest trappings of science. Sociology, eugenics, psychology (when it wasn't indulging in math-free psychoanalytical fantasizing) and a myriad of other pretty dodgy and continually invented sciences do just that.
How those things came to be used in criminal trials and the meting out of something called justice is exactly an act of faith in anything officially peddled in academia as science which was never justified by the quality of the results. Those results incorporated biases from the start of the very practice of data collection other formal decisions of what to include and what to exclude as not within the parameters considered in the operation. Such biases are present at every level of the supposed science that deals with human behavior and social interactions.
The very act of studying complex and variable phenomena inevitably include practices guaranteed to embed biases in what is claimed as producing reliable knowledge and the character of the results. For example, basing such decisions in the creation of computer programs such as the one mentioned in yesterday mornings post. Such programs are based on things like a sociological profile of "typical" convicted criminals would, beyond doubt, incorporate the social, class, ethnic and racial biases of society because judges, lawyers, juries and the media that influence them incorporate all of those in deciding who will be convicted, who will be successful in appeals, who will be sent to prison and what crimes done by who which will even be recognized as illegal and whether they are misdemeanors or punished as felonies. It is far more likely that rich people, especially rich white people who commit terrible crimes will go unpunished or even prosecuted. The non-prosecution record of those who committed massive theft and fraud since the start of the century, producing some of the worst economic crises in living memory is certainly a factor in such things
If recidivism is a consideration in coming up with their algorithms for divining the future actions of people before a court, that would inevitably include whatever influences of differences in the kinds of punishments are handed out in previous convictions, the habits of judges who go easy on rich people, the existence of country-club prisons for the rich and connected, the very decision of whether or not to send someone to prison or give them some kind of alternative punishment - if any at all. That would certainly tend to keep some classes of people out of the schools for training criminals, brutalizing people into worse people than when they went in the first time, and certainly it includes such things as whether or not people can reenter society after they get out. Rich people with millions of dollars and ruthless high-power attorneys and other protections money buys will certainly have an easier time avoiding the desperation that a poor person who can't get a job will face.
Our justice system, juries, sentencing, prisons, etc. are not a uniform and guaranteed system that will not include bias in either decisions or in carrying out its various functions, producing a uniform set of people about whom you can make accurate and dependable predictions anymore than the general society does. There was never any reason to believe in the proposed methods that pretend they are.
Even if the creators of such algorithms and computer based excuses or legal requirements for judges to treat people differently didn't want to include racism, class bias, etc. in their programs those and other forms of previous societal and legal bias would be inevitable in even an honest collection of data because the raw material they work with includes it. Any claims made about the product they are selling to courts and legislatures and governors that don't admit to those biases is certainly consumer fraud just as any scientific publication that doesn't include those admissions is scientific fraud. Such scientific fraud is endemic to the social sciences and other, similar forms of it are as endemic to the allegedly scientific treatment of human minds at all levels. If you think that last statement is extreme, consider how much of that relies entirely of untestable, unverifiable reporting by people of their own experiences of their mental states, their perceptions, their ever foggy, ever shifting opinion and their past histories of behavior. There is no way for a researcher who includes reporting of personal thoughts as raw material in their supposed science to verify the accuracy or even truth of what is reported. Giving people a small number of set responses to choose from certainly does nothing to enhance the accuracy of the results, such a thing as even the order in which those appear in the presentation to a subject incorporates biases into the mix.
Such things are inevitable when an alleged scientist wants to treat very complex and, especially, unverifiable, sometimes even undefinable phenomena with a pantomime of science. I have used an admission of the rapidly increasing decay of reliability of that scientific method in even the most precise of physical sciences by the eminent French mathematician, René Thom, to make this point before.
The excellent beginning made by quantum mechanics with the hydrogen atom peters out slowly in the sands of approximations in as much as we move toward more complex situations…. This decline in the efficiency of mathematical algorithms accelerates when we go into chemistry. The interactions between two molecules of any degree of complexity evades precise mathematical description … In biology, if we make exceptions of the theory of population and of formal genetics, the use of mathematics is confined to modeling a few local situations (transmission of nerve impulses, blood flow in the arteries, etc.) of slight theoretical interest and limited practical value… The relatively rapid degeneration of the possible use of mathematics when one moves from physics to biology is certainly known among specialists, but there is a reluctance to reveal it to the public at large … The feeling of security given by the reductionist approach is in fact illusory.
People who peddle sentencing software claiming the scientific nature and reliability of what is behind it are committing fraud. Science can't do what they are claiming it can do, it can't come up with durable and reliable knowledge of the kind claimed about defining human behavior nevermind being used like a set of Tarot cards to divine the character and future actions of any given person. That some means of coming up with an impartial means of making those predictions has not been available and that there is at times grotesque injustice handed down by judges and juries makes such an impartial, reliable oracle desirable doesn't mean that you're ever going to be able to do that scientifically is ever going to be true. That fact is a better argument for logical rigor, not suspending it when someone calls what they do and sell "science".
This practice is a better indication of how people have come to replace a faith in the official occult with just a different form of it, one which fully incorporates human biases into its science mimicking method but covering that with the pretense of objectivity which is certainly not there. That a social scientist, himself, might be excluding his bias doesn't mean that his method doesn't include that of others who provide him with his alleged data. That is how racism is embedded in such commercial products as those studied in that Pro Publica article I mentioned yesterday.
I used to share that faith in the so-called sciences, though the more I learned about the claims of such sciences, the more I became aware that most of it is fraud, conscious or unconscious or buried in the casual non-admission of what they were up to which constitutes the professional practice of academic social science. That kind of thing is bad enough in the general culture but when you endow it with the power given to judges to make often disastrous decisions over the lives of individuals, it becomes extremely dangerous. It is one with a piece in the faith of the power of such alleged science that led to the terrible Buck vs. Bell decision in the United States, racial and ethnic immigration exclusions, etc. and even worse decisions elsewhere. Endowing the social sciences and, especially, their for-profit peddling of their fortune telling wares, with the power of the judiciary is anything but an enlightened act. It is modern-day demonology, astrology and oracular divination calling itself science.
There was never any reason, from the start, to believe that science could do these things.
See Also: The science of psychiatrist James Grigson, Dr. Death, expert witness in capital murder cases who guaranteed that even people later determined to be innocent of murder would reoffend but who were given death sentences on his say-so. He was very popular with prosecutors. He got black balled by a few professional groups but I'm unaware of him ever being excommunicated from the scientific community. He was still allowed to act as a psychiatrist years after the scandal led his fellow practitioners to distance themselves from him, after he'd participated in the judicial murders of certainly innocent people.
The bigger scandal is that such use of "expert witnesses" to give people a death sentence is allowed. They might as well consult an astrologer or, more aptly, a phrenologist. It's the equivalent of that.
I know the arguments are far subtler, but I often point out to my students that 1+1=3, or more, in biology. Sometimes the math is rather curious.
ReplyDelete