William Uttal has had a long and distinguished research academic career as Professor of Psychology at the University of Michigan and as Professor in Arizona State's Department of Industrial Engineering.  When a person of his standing and orientation takes on the question of the scientific status of psychology, it's worth taking a look.  If you do, you won't be disappointed.  Utall's writing is clear, relatively free of jargon and he has the confidence not to fudge his points of view or his conclusions.  The issues he tackles were more actively addressed at the philosophy/psychology border area in the 1970s and 1980s than recently, but that does not render them out of date.  No doubt the absence of Burrhus Skinner has dampened some of the interest.

Uttal's thesis is easily put though not so easily understood.  "Psychology is a science in the usual sense of the word to the degree that it is behaviorist." (245)  Part of this he makes very clear.  By "behaviorist" he means the approach that eschews any reference to cognitive or mentalist concepts.  It is the behaviorism of Skinner and not of Tolman or Hull and certainly not the "behavioral decision theory" of a Daniel Kahnemann or Amos Tversky.  For Uttal it's not enough to anchor mentalist concepts (expectation, belief, desire, mental mapping, intelligence etc.) in empirical observations, one must avoid them.  We'll get back to why.

There are two elements of his thesis that are not clear.  One is: why should anyone care if psychology is a science?  A "Boston cream pie" is two layers of cake, cream between and chocolate on top.  Is it really a pie or a cake?  Who cares?  It would only be interesting if one of the categories held a higher value than the other, as in the invidious, "Let them eat cake."  It matters if psychology is a science only if the label "science" confers a prize, e.g., "superior method of knowing".  Uttal does not argue this, but must be assuming it.  (If I may, I've learned a good deal about people, their types and quirks, from good literature, at least as much as from my travels through psychology or philosophy.  No one asks if psychology qualifies as literature.) 

The second issue is trickier:  How should we characterize, ". . . science in the usual sense of the word . . ."?  The author devotes Chapter I, Lexicographica Scientifica, to this issue (His chapter titles are the exception to my claim above about clear, jargon-free writing.)  Uttal begins by seeking the help of common dictionaries and various websites (Wikipedia, of course), none of which provide assistance.  He rejects "organized body of knowledge" (common in the nineteenth century) since that would apply as well to the studies in Madrassas and Yeshivas (and to Hegel, though he gets no mention.)  He is pleased with, ". . . search for an orderly understanding of some aspect of the natural world."  I doubt however that one could unpack "natural world" in any way other than "what science can fruitfully study" which would make his definition circular.  And again, though Hegel talks about such things as "absolute spirit", when it comes right down to it there's nothing "un" or "non" natural going on within his highfalutin semantics.  Finally, Uttal tries some of the theories common to the positivist traditions of the latter half of the twentieth century and feels more at home, ". . . the ideal of science is embodied in the observation-induction-axiom-deduction-theorem-verification method . . . " (85)  "The axiomatic-deductive method has become the archetype of modern scientific reasoning." (94). Uttal has read some philosophy of science though there is little evidence that he takes seriously the long list of charges against this account.  Moreover, he associates this method with Newton (one could try to make the argument), Popper (definitely not) and Russell and Whitehead (definitely not) (91).  There's a failure to distinguish the true sense of "axiom" in the logico-mathematical sense of "claim taken to be true in some system yet unproven in that system" (Euclid, Peano, and others) from fundamental (empirical) laws such as are found in Newton or some axiomatic decision theories. 

One odd point in Uttall's analysis is that in several places he warns against "physicophilia" or "physics envy", the adoption of the theoretical structures of physics as a model for psychology (57, 58, 247).  This is all to the good and I wish he had carried it through.  The problem is that the model of good science that Uttal erects is exactly the one that is exemplified most fully by physics.  There are numerous axiomatizations of classical mechanics, relativistic mechanics and quantum physics.  Where are the corresponding examples from botany, geology, archaeology, evolutionary theory, physiology, etc?  Ironically the next most likely place to find axiomatic systems in elegant form is in psychology, specifically in decision theory (e.g., Ward Edwards's ed. Utility Theory: Measurement and Application, 1992).  It's difficult to escape the conclusion that according to Uttal an activity is scientific to the degree that it resembles what Isaac Newton did and its results are science to the degree that they resemble the structure of Newton's Principia.  I think Uttal's book would have had a cleaner focus (and been a good bit thinner) if he simply addressed directly how psychology could more effective at its tasks and forgot about whether it's a "real science"?

A personal digression.  Why is physics better at what it does than psychology is at what it does?  Simple, physics is about the dumbest things in the world (billiard balls, electrons, quarks) and psychology is about the smartest (pigeons, rats, humans).  And even given that fact, physicists select their problems with care, namely those with simple environments of few variables (moon shots, linear accelerators, etc.).  Take a bag of confetti, mark one piece with your initials in pencil and empty the bag from the top of the Sears Tower in the Windy City.  Can physics tell you where the marked piece will land?  LaPlace would have said yes, but only if the physicist has perfect information to accompany his copy of the Principia but that's the point of careful selection of problems.  In the end, one could say that physics is to psychology as checkers is to chess.  Physics envy indeed!

Let's suppose that Uttal has an adequate account of what science really is.  What's his case that most of psychology doesn't make the grade?  He has a number of charges but overwhelmingly what has him riled up is that psychologists posit and seek to investigate mentalist or cognitive processes.  They introduce into the discourse of their theories and research references to states of consciousness that can be directly experienced only by those whose states they are.  What's wrong with that?  Here are few answers: 

(1) It assumes an ontological dualism and organized religion does that.  Okay that's a weak one. 

(2)  Such states are inaccessible, meaning that they cannot be publically observed (except by their owners).  Uttal recognizes that this charge of inaccessibility to direct observation applies as well to physics, "Quarks, gluons, hadrons and other particles of the micro world … are just as invisible …" (68), yet he does nothing with this recognition.  He doesn't argue that this is bad for physics.  He doesn't argue that it works for physics but can't for psychology.  This is a gaping hole in his overall argument.  One might even argue that psychology has one up on physics in this regard since at least one person can experience excruciating nerve pain and no one can experience a single photon of light, or a collapsing star.  And to, my inability to experience another's inner states is more a technological shortcoming than a conceptual one.  That is, it doesn't take a science fiction genius to envision a technological advance by which my nervous system will be wedded to yours so that I could truly and literally, in the words of William Jefferson Clinton, "feel your pain".  In that not very far off world mental processes will be pubic events – no doubt to be recorded by satellite and reviewed by our intelligence professionals.  

(3)  Cognitive states are not measurable.  This is a surprising charge since the greatest advances in the understanding of measurement in the twentieth century came out of psychology (e.g., the multi-volume Foundations of Measurement by such authors as Patrick Suppes, David Krantz, Duncan Luce, and Amos Tversky.)  And one of the most elegant definitions of measurement belongs to psychologist S. S. Stevens, "The assignment of numbers to things or events according to a rule."  Stevens is talking here about the mathematical concept of a function, and about measurement as a one-to-one or one-to-many correlation between numbers and some feature of the world, to achieve some goal.  And certainly psychology is filled with measurements of cognitive properties such as IQ, heritability, JNDs and many others, measures which are justified under Stevens' definition.  To make his case that cognitive properties cannot be measured Uttal attacks Stevens' approach in favor of one that requires that all measured quantities by definition be transitive (he uses "ordinal"), that is, where A>B and B>C then it must be that A>C.  Uttal claims that all physical quantities exhibit this property and many cognitive ones do not.  For example there are many men who would prefer Mary to Jean and Jean to Sue but prefer Sue to Mary, where "prefer" is the intransitive cognitive state in question.  Does this imply that "preference" cannot be measured?  Not at all.  First, one could use the transitivity principle as a limiting condition such that the theory applies only to transitive preferences; Luce and Raiffa do this in Games and Decisions (1957) for "indifference".  A second approach is to develop a refined scale of measurement that includes intransitivities as Duncan Luce did in his theory of semi-orders (1964) that handled certain intransitivities.  It's not unusual in science that theoretical or empirical problems generate new systems of mathematics.  These are complicated matters that take a good deal more analysis than the author provides.

Uttal has many very interesting observations about psychology.  I'll just mention one, which is the penchant for researchers to grab onto a technological development and transfer it as metaphor into psychological discourse. (53)  I recall a teacher claiming (in the mid 1970s) that there is no longer an issue of the relation of cognitive to physiological states now that we have the example of the software/hardware distinction.  Uttal's rightly criticizes the importation of neural metaphors in a manner reminiscent of Skinner in Behavior of Organisms (1938) translating tongue in cheek Hull's CNS as the "conceptual nervous system".  Uttal has many wise things to say about the state of scientific psychology that do not rely upon his arguments, unsuccessful I think, that psychology should not be afforded the honorific title of "science".

I'm not sure who the audience is for this book.  If this were used as a stand alone philosophy of psychology text to psychology students I'd like to see an instructor well versed in the debates that took place within the philosophy of science from the early Rudolf Carnap to the late great and eccentric Paul Feyerabend. 

© 2008 John D. Mullen

John D. Mullen is Professor of Philosophy at Dowling College in Oakdale, New York.  He is the author of Hard Thinking: The Reintroduction of Logic into Everyday Life, (1995) co-author with Byron M. Roth of Decision Making: Its Logic and Practice, (1991) and the author of the widely read Kierkegaard's Philosophy: Self-Deception and Cowardice in the Present Age.(1995)