Is experimental psychology science?

Maybe I should clarify that.The term "soft science" is pretty old. The traditional sciences have often been called "hard science", and I assume that "soft science" is a back formation from that. The better term is "social science", but I was just continuing with terminology that had already been used in the other thread.My comment about predictability was a bit too simple. The real problem in the social sciences, is that the type of experimentation that is possible is very limited due to ethical considerations. There are occasional natural experiments, such as when a feral child is discovered. But such natural experiments are rare.For myself, I don't have any doubt that experimental psychology is science.

The terms "soft science" and "hard science" are probably older than I am. These are coffee room terms. When people want to sound a little more professional, the distinction will be between "social sciences" and "physical sciences". Often mathematics is included in the physical sciences, even though it is neither physical nor a science.There is an informal pecking order in the sciences, roughly

physics > chemistry > geology > biology > social sciences.

Feynam had a somewhat low opinion of biology, in "Surely You're Joking, Mr. Feynman!".I'm inclined to think that a lot of the distinction between soft and hard sciences has to do with the extent of the theoretical basis, and the degree to which mathematics is used in that theory.Biology has changed since the time of Feynman's experience with it. By now it has a larger theoretical component, and is more mathematical. But I guess I should mention that there is a pecking order in mathematics too, with differential equations ranked considerably higher than combinatorics and graph theory. The mathematics used in physics and chemistry is thus rated higher than what is used in biology. Statistical hypothesis testing, much used in psychology, is probably ranked lower than graph theory, and that may have to do with the perceived softness of the social sciences.In some ways, economics (sometimes dubbed "the dismal science") is more mathematical than other sciences. That's because it has economics has its own clear measuring systems (money, GNP, etc). From the perspective of measurement, the best of psychology is in psychometrics, where psychology develops its own measuring instruments. Whenever you have a measuring instrument, there is a possibility for mathematical analysis of that instrument. Unfortunately, psychological measuring instruments give considerably less precise answers than even the economists' instruments. And they are mired in controversy (does IQ measure an innate ability, or is it culturally biased).In our campus library (Library of Congress classification system), the psychology literature is adjacent to and intermingled with the philosophy literature. It isn't even close to where the science literature may be found.Thus there are all sorts of cues which tend to hint that psychology is not as much of a science as the physical sciences.As for the work that experimental psychologists actually do, some of it is quite good. But psychology gets a bad reputation from its mistakes, such as some of the work on repressed memory.

Research psychologists discuss inner feelings. But their scientific reports are typically reports about behavior that is usually believed to be related to feelings. If you read the reports, they make it clear that they are reporting behavior.Note that verbal reports of having feelings are behaviors.

This is a combined response to several of Ben's posts in this thread. Note, however, that I am not a psychologist. I am commenting from the perspective of my understanding of the nature of science, and my own amatuer studies in cognitive science.In Message 39, Ben writes: It isn't easy to come up with good working definitions. If you look at other sciences, you will see that there are periods where scientists are struggling to find usable definitions and concepts. The early scientific investigation into electricity and magnetism comes to mind as an example of this. I see psychology as going along that path, struggling to come up with clear formulations of what is to be investigated. But that is how it has to start.In Message 40, Ben writes: I suggest you take a look at the book "How the laws of physics lie" by Nancy Cartwright. No need to seriously study it. Just browse through it briefly in your campus library. Cartwright discusses how the laws of physics are about things such as chargeless particles, massless charges - things that don't actually exist. And such laws are nevertheless useful because we can piece them together to solve what we are trying to investigate.You are right, that a lot of psychological research is done in labs that don't correspond to real life. Still, that can yield an understanding that is applicable to real life situations. I'll grant that it will be hard for psychology to piece the various components together. But there is value in what is being done.In Message 40, Ben writes: I am as skeptical of folk psychology as you are. I expect that it will eventually be discarded from psychology. But the history of science shows that you cannot discard a theory until you have a good replacement. Even poor theories can be the basis of scientific discovery. Let's remember that the discover of oxygen was made under the assumptions of the phlogiston theory. In a sense, modern chemistry can trace its origins back to work done under phlogiston and alchemy assumptions. Scientific investigation has to start somewhere.Also in Message 40, Ben writes: Fitt's law is an example of an empirical law - a relation found from empirical observations. Newtons laws claimed to be causal laws.Doubtless, causal laws are to be preferred where available. I'm like you, I don't find Fitt's law very satisfying. But this is where psychology runs into difficulty. It might not be possible to find strong causal laws for psychology.The radical behaviorists (the B.F. Skinner school) claim to be finding causal laws of behavior. Count me as a skeptic. I don't think such laws exist.In his book "Beyond Freedom and Dignity", Skinner argued against the idea that we have free will. Radical behaviorists seem to assume that a person is pretty much like a mechanical automaton, and causal laws of behavior do exist and are waiting to be found. I think Skinner was mistaken about this. I think we do have something like free will, in the sense that we are not at all like mechanical automatons. I don't expect that we will ever have strong causal laws of human behavior.In Message 44, Ben writes: I expect that we will eventually come up with "mechanisms" that underly human behavior, but this still will not give us strong predictive laws. Your mechanisms will be different from my mechanisms, and both of our mechanisms will be under constant pressure to change as we continue to learn. We will come to know these mechanisms only in the form of broad principles, because it will be too complex to catalog all of the details for any individual. Moreover, I expect it will turn out that our mechanisms are such as to make us exquisitely sensitive to very subtle details in our environment, and this alone might rule out any hope of being able to predict individual human behavior. The kind to broad statistical laws that you see in Fitt's law might be about as good as you can do in terms of laws.The kind of understanding of mechanisms that I see possible, while it might not lead to strong predictive laws, could still be very useful. It could help design better teaching methods, perhaps it could help design better therapies for various conditions. So I look forward to some sort of success.In Message 46, Ben writes: I think your concept of science is a little too narrow.

I'm taking the liberty of spreading those basic questions over several lines. It depends on what you mean by "language". If "language" means what the Chomskyan school studies, then apes don't have language. And humans probably don't have language either. This depends on what you mean by "conscious". Personally, I think it quite obvious that other mammals are conscious (except when asleep, etc).Many people claim that other animals are not conscious. And then there is Julian Jaynes ("The Origins of Consciousness in the Breakdown of the Bicameral Mind") who thinks that even humans were not conscious until relatively recently. I haven't a clue as to what these people mean when they use the term "conscious". Again we have the problem of what is meant by "conceptual knowledge". I would say "yes" to this question. However, what I mean by "conceptual knowledge" is clearly different from what most people mean. Most people seem to tie it to language. The problem is more basic. A science cannot properly investigate its subject matter, unless that science controls its own definitions. Psychology is still a fledgling science. It really hasn't come fully to grips with the need to define its terminology and to control its own definitions. In some areas (I mentioned psychometrics in an earlier post) it has done this. In other areas it too readily accepts the terminology that comes from the culture.

It is my impression that there is a lot of skepticism of cogsci among research psychologists. I will be interested in what Zhimbo has to say about it.