Doug Candland gave the Arthur Staats Lecture for Unifying Psychology in 2010. He was a strong influence during my undergraduate education, and has been a trusted adviser since. Though he has feigned asking me for advice many times -- I say 'feigned', because it often seems to be an extension of an oral exam we started over a decade ago -- he seemed to be actually asking for my advice when we talked about what he might say in his lecture. I'm not sure how much I influenced the talk, but I was quoted once. The title of his talk was "The End of Psychology." During the conclusion, he gave the caveat that he could be completely mistaken, that one of his students had told him "Psychology can not be over, it never really began."
Psychology is broken in many, many ways. It is broken at every level. At the lowest level, Intro Psych is one of the most poorly conceived courses in the college curriculum; at the highest level, the arguments of famous 'top researchers' are based on faulty premises or are more oriented towards public appeal than any scientific value or philosophical rigor. Further, psychology is broken at its core. The dominant framework of cognitive psychology has failed, and while much of the field has accepted that in practice, they deny it in discourse. That is to say, many people are trying to modify, supplement, distort, or even abandon the traditional computer-metaphor for the mind, but they are unwilling to give up the associated vocabulary, and they want to retain the basic questions that the computer-metaphor implied. This is just misguided: Yes, we have memory, but the idea that our having-memories works nigh identically to computer "memory" is wrong. Yes, we perceive, but the idea that our perceiving-the-world works nigh identically to computer "encoding" of "input" is wrong. More than a minor revision is needed.
That said, there is still a 150 years or so worth of experimental work in psychology, which has yielded many stable findings. One reason for psychology's current crisis is because most psychologists know little of this past. When research lines come to a conclusion (either from solving the instigating question, or from gaining insight that renders the question mute) the research drops out of common knowledge very quickly. This, obviously, makes cumulative knowledge production difficult.
Even the fact that the theory behind cognitive psych is faulty, doesn't make it any less tragic that most of the experimental results cognitive psychologists accumulated over the past 60 years are almost completely forgotten. I cannot count the number of cognitive psych talks and cognitive neuroscience talk I have seen that involved a great experiment, with solid results, marred only by an incoherent introduction and conclusion.
There are areas of research that are on a better track, however, and generally those are areas that have a longer memory. The study of comparative psychology generally fits in this category, and within that, especially the developmental psychobiologists and the old-school rat and pigeon behaviorists. The latter get a bum rap because people don't like the philosophy often associated with them, but their results stand regardless. As for the former, dated though a few parts of it are (at now over 15 years old) Michel and Moore's Developmental Psychobiology textbook remains one of the best textbooks in all of psychology, because it is simply full of stuff psychologists know. Page after page of studies most people have never heard of, which definitively answered essential questions in the field. Take that, throw in some sophisticated behaviorism, and some Gibson, and you have the start of fixing psychology. Embodied, embedded, first-principled psychology.
But that isn't just the future, it is also the past. All those pieces were already put together over 100 years ago. They were there in the late work of William James, and the work of E. B. Holt. This was a very sophisticated behavioristic psychology, built on a foundation of 'American Philosophy' - Peirce's Pragmatism, James and Dewey's Radical Empiricism. Direct perception, physiological psychology, and a descriptive account of behavior merged together. Sam Parkovnick has called it a 'meta-theory of psychology', the theory that must be true for other 'theories of psychology' to be coherent. This is the psychology that was never really born. James was criticized by psychologists for being too philosophical, and criticized by philosophers for being too psychological. Holt got the same treatment, but was worse off for lack of James's charisma, for his personality issues, and for the increasingly 'professionalized' state of the field while he was writing. James and Holt's mutual agenda was hijacked first by crude behaviorism and then by cognitive psychology.
Yet threads of it remain. There have always been a small number of behaviorists trying to keep the message going, early Tolman is clearly in this vein, as is the current work of Francois Tonneau. Gibson's work clearly fits in this vein as Harry Heft, Alan Costall, and others have pointed out. It also lead into the developmental epigenetics work of Kuo, Schnerla, Lehrman, Gottlieb, and others that became the modern developmental psychobiology. For example, having recently read a biographical interview with Gottlieb, I was surprised to learn that his career was largely shaped by an undergraduate reading of Bentley and Dewey's 'Knowing and the Known'.
There are other threads that need to be found. But finding thread is not the goal. These threads, strong as they are individually, are not interwoven. Though there seems to be increasing awareness that these seemingly disparate approaches to psychology have some affinity, some common core that makes them compatible, there is little awareness of what that common core is or where it comes from. The birth of a coherent science of psychology requires reweaving the tapestry. Psychology will not be fixed until we can recreate a coherent framework within which the empirical facts of psychology can fit, and can build upon each other.