Tag Archives: bias

ethics free-speech freedom of speech hate speech Incendiary Speech journalism marketplace of ideas neutrality objectivity Programming social construction social software

Notes: Bias in computer systems

Friedman, B., & Nissanbaum, H.  (1996). Bias in computer systems.  ACM Transactions on Information Systems, 14(3), 330-347.


In this article Friedman and Nissenbaum look at bias in software systems. Although the word bias can cover a number of related concepts, the definition used here is something that systematically and unfairly discriminates toward one party or against another. The authors see three main classes of bias in computer systems: Preexisting bias, when an external bias is incorporated into a computer system, either through individuals who have a hand in designing the system or via the society the software was created in; Technical bias, where technical considerations bring about bias (from limitations, loss of context in algorithms, random number generation, or formalization of human constructs); and Emergent bias, where bias emerges after design when real users interact with the system (for example, when new information is available but not in the design, or when systems are extended to new user groups). A number of illustrative examples are given, and the authors look at a number of specific software systems and point out existing or potential biases. One of the systems is the The National Resident Match Program (NRMP), used to match med school graduates to hospitals. In this system, if a student’s first choice of hospital and hospital’s first choice of student do not match, the students’ second choices are run against the hospitals’ first choices. Overall, the result favors the hospitals. Two steps are proposed to rectify bias – diagnosis and active minimization of bias.

This is an extremely interesting subject, and and I doubt most users and programmers are any more aware of it now than they were in 1996. One more recent article, (http://web.mit.edu/21w.780/Materials/douglasall.html) which sought to turn literary criticism toward video games by pointing out cultural biases, also mentions the lack of study in this area. With so many people spending so much of their day interacting with software, why do these kinds of articles seem so few and far between? On the other hand, the particular examples chosen are illustrative but not very current. All three of the systems were large-scale, mainframe-type software that users interacted with in a very small sense. Would the risk of bias be even greater for a system which is largely a user interface?

One clear implication is shown in the diagnosis stage of removing bias—to find technical and emergent bias, designers are told to imagine the systems as they will actually be used and as additional user groups adopt them, respectively. So the charge is one-third ‘know thyself’ and two-thirds ‘know the users.’ The very notion of looking for bias is probably foreign to many user interface designers (in fact, few of the programmers I’ve met are even aware that accessibility guidelines exist for blind, deaf, and other users). The authors’ proposal that professional groups offer support to those designers who detect bias and wish to fight it is a nice thought but doubtful. Few programming or UI organizations can exert any kind of pressure or drum up much bad publicity, or if they can, I haven’t heard of it (which I suppose means they can’t).

Hate speech and freedom of speech

Jonathan Rauch argues in his “Defense of Prejudice: Why Incendiary Speech Must Be Protected” that despite all the pain hate speech causes to the groups and individuals it targets, it must be protected.  In “There’s No Such Thing As Free Speech and It’s a Good Thing Too,” Stanley Fish argues that free speech is a sham-it’s a phrase tossed around by those seeking to use it to gain power.  When their opponents use it, it’s no longer important to them.

Rauch comes to his conclusion from something close to the marketplace of ideas theory.  Scientists, for example, although they seem to be impartial and authoritative, are actually all fueled by biases.  It is the competition and collaboration of all these biases which leads to an unprejudiced field as a whole, and the best ideas win.  Rauch seems to thin that letting society or the government pick which speech is to be tolerated and which not would be similar to the old-fashioned way of forming consensus used by kings and dictators: eliminating the opposition.  He also sees the movement extending past hate speech and into prejudice in all its forms-from hate speech codes in universities to laws treating the same act more harshly if it was motivated by violence.  In the end, minorities can only lose out when dissident opinions are banned.

Fish points out, however, that even Milton’s Aereopagitica, a famous defense of speech and tolerance, he ends by excluding Catholics.  Fish thinks this is indicative of a larger notion; that by defining a space of tolerated or free speech, we inherently provide barriers against what is intolerable or gibberish.  For example, the First Amendment does not allow freedom of action, and therefor allows speech which is also action to be regulated.  Furthermore, pure free speech can only be maintained when no one has any stake in what they’re saying-Universities might say they’re for free expression, but if so, why have classes and tests?  Because those things work toward the University’s actual purpose.  If speech works to that purpose to some degree that is the degree to which it is tolerated.  On another level, answering hate speech with more speech would only work if contrary opinion could take away the pain, fear or humiliation.

I don’t think I could pick one of the authors here as the winner of the debate because their arguments diverge.  This does remind me, though, of the debate running in the Transcript last semester about hate crime legislation.  On one side was the In Righting columnist and some letter writers, and on the other was the From The Left columnist and some letter writers.  Though the debate went on for weeks, neither side presented a good case (or anything as intricate as the arguments here).  Most notably, the side for hate crime laws failed to come up with any rational justification what-so-ever.  Although I think many hate crime laws as written are too restrictive and in general I don’t like content-based restrictions, I can think of one rational that makes some sense.  Take a cross burning and compare it to two 13-year-olds playing with matches on their neighbor’s property.  With no hate crime laws, both receive the same charges-trespassing, destruction of property, probably not arson.  But now ask: which action does society have more of an interest in preventing/punishing?  This is getting to Fish’s argument.  Does the accidental act of two kids harm society as much as a KKK cross-burning?  Doubtful.  And the reason is content-based.

I’m still not sure that overcomes the necessity of free expression Rauch argues for.  And it could be said that by confining his argument (more or less) to this country, Rauch’s arguments could be slid into Fish’s schema-proving how free expressions serves the underlying purpose of our society within confines.

Defending and defining objectivity

A response to Mass Media and Society (James Curran and Michael Gurevitch), Chapter 11

In “In Defense of Objectivity Revisited,” Judith Lichtenberg argues that most critics of objectivity are actually criticizing the concepts associated with objectivity or specific practices which are thought to be objective and therefore we cannot abandon objectivity completely.

Lichtenberg points out that critiques of objectivity often take three forms: saying journalism is not objection, should not be objective, and can not be objective.  Though many critics try to make more than one case at a time, they are logically inconsistent.  If objectivity is impossible, why complain that journalists are objective?  When a piece of reporting is accused of being not objective, we have ways to gauge it (by reading Iraqi news during the Gulf War or pointing out over-reliance on military sources, for example).

She makes the point that critics who say a journalist’s social situation dictates certain biases cannot also say there’s no way to get around those biases or analyze them-that is, after all, what the critic is doing.  It makes no sense to call something biased unless there is some sort of humanly-accessible objective reality to compare it to.  Also, different cultures are often able to find common ground.  None of this is to say that journalists don’t write horribly distorted, biased stories, but Lichtenberg suggests if we can recognize the problems there must be ways of making them better.

Lichtenberg says that objectivity does not mean there are right answers for all questions, but that there are right answers for some questions and wrong answers for all questions.  The answer to “is the moon made of green cheese?” is clearly no, and it’s not a theory (in this day and age) or socially constructed.  It allows for different interpretations, but all interpretations must agree on that fact at least.

I agree with Lichtenberg in that there is some sort of objective reality we can reach, at least for some matters.  You can’t breath water, and no matter what culture you’re from, if you try, you die.  I also think there’s a sliding scale of possible objectivity.  Some things can have objective truth and do, such as the water example.  Other things have an objective answer but it is not known and is therefore open to speculation.  And some things have no objective answer, like value and opinion judgements.

This all makes sense, and if you define objectivity the way Lichtenberg does I will happily pursue it when applicable.  If you include some of the notions she does not, however, like neutrality, then it’s a stickier question.