The Voodoo Sciences: The Difference Between "Data" and "Evidence"

Go to any U.S. university. You will hear lamentation and wailing and gnashing of teeth. Washington has become unfeeling and stupidly refuses to support higher education: don't those idiots on the Potomac know that education is a investment in the future? Don't they know that human resources are our most valuable resources, that public higher education is necessary preparation for a democratic future? That we must invest in the future?

But now wander about the campus, and look at how our typical university allocates that all-important investment dollar. You will find that the "social science" departments are far larger than the "hard sciences," and indeed have more students than are enrolled in liberal arts. You will find that even in states with tens of thousands of unemployed teachers, the Department of Education is among the very largest departments on campus.

The social sciences will be large and important departments, with many members of faculty and much classroom space. One wonders what it is that graduates in the social sciences are prepared to do. It must be an important skill; we are spending a large part of our scarce but all-important investment funds to acquire it. Oddly enough, though, we're not training so many engineers and scientists, physicists and mathematicians. Why?

But of course the answer is well known. In most universities, our education investment funds are allocated by entering freshmen. They go to a kind of oriental bazaar, where they are seduced into choosing a major; the number of majors then determines the department's share of the university's budget funds. It does seem an odd way to allocate an important resource.

One might suppose a better way: that the legislature, or other public authority, determine the number of engineers, biologists, physicists, medicos, sociologists, etc., that might reasonably be required in the future, and allocate public funds among the departments accordingly. Students wishing to declare various majors could do so; but when the number that the taxpayers will support is exceeded, the next student to enroll in that major gets to pay tuition accordingly. If tax-supported higher education is an investment--and what other theory justifies sending the tax collector, policeman, and eventually the public hangman to extort the funds from the taxpayers? -- then might we have some care in the way that investment is allocated? The present scheme looks like a bad parody invented by an inept science fiction writer. Who'd believe it if it weren't happening?


The real difference between arts and sciences is the difference between data and evidence; and the "social sciences" don't know one from the other.


When this was put to Dr. Paul Bohannon (5), dean of the School of Social Sciences at the University of Southern California, he replied that Mead's value didn't lie in her data-gathering. She stretched imaginations and made people think larger thoughts.

Granted this may be true, but it seems more the job of a science fiction writer than a scientist.


The social sciences have made an art of forgetting embarrassing facts. If a fact doesn't fit the theory, leave the fact for another discipline. Sociology has nothing to learn from anthropology, which has nothing to learn from social psychology. None of these has anything to learn from the mathematics, physics, or chemistry departments.


The poet who believes he knows something of science having taken "Sosh 103" and "Ed Stat" is far more dangerous than ever he would have been if he had remained ignorant.

Meanwhile, novelists have as much right to be called "experts" on human behavior as any social scientist, which is to say we can learn as much about our fellow humans from a good novel as from a sociological treatise; and I know which I would rather read. Similarly, the poet may find beauty in the theory of probability, and will learn something of the difference between data and evidence while studying it; "Stat for Social Scientists" teaches nothing, and is dull in the bargain.

When the social scientists are challenged as unscientific, their usual plea is that their subject matter is very complex and thus the methodology of physical science won't work. This is an interesting argument, but it would carry more weight if students of social science knew something of physical science's methodologies. Granted that the "social sciences" have an intrinsically more difficult job; is this any reason to abandon the tools of science?

The article is long but well worth reading in its entirety. Via The Voodoo Sciences.