My understanding about the behaviorist learning theory is this, if we are our observable actions, then learning-wise; the measure of our learning is judged by it’s visibility. Put another way, the more we appear to learn, then the more we have learned. This is like the proverbial duck test;  ‘If it looks like a duck, swims like a duck, and quacks like a duck, then it probably is a duck.’  This school of thought assumes that everything ‘above the hood’ is just as ‘everything beneath the hood’. But this form of abductive reasoning does not wholly apply to learning at least based on anecdotal evidence. First, knowledge in != knowledge out. Second, knowledge present != knowledge always present.
Knowledge in != Knowledge out

A student’s demonstrated performance is not always indicative of the measure of knowledge they have gained; in actuality, it could be significantly more or less. Consider for example the dichotomy of my educational experience as a teacher and student. As a teacher, most of my learning measurement metrics rely on behavioral indicators by way of scores on tests and assignments. The A students are the brightest right? Right. Or are they? As a student I know that so many factors can affect my grade on a given course. It could be how much workload I have at the time of enrollment, my interest in the material I am learning and whether I have a stomach flu on test day. There is clearly a disconnect between the evaluation of knowledge and the assimilation of knowledge using the behaviorist approach.
Knowledge present != Knowledge always present

That a student can demonstrate some evidence of knowledge right now, doesn’t mean they have retained that information. This means that asides from being unable to measure how much knowledge has been gained, they could simply be studying for the grades. And they could discard all that information altogether as soon as that assessment is up. For example; if I promise my toddler ice-cream for behaving well, I cannot guarantee that he has really adopted the new attitude or is just there for the sugar. Given such a situation, one cannot say real learning has occurred.

This is not however to say, behaviorism is all bad. I have utilized behaviorist measures as a teacher because it is what seems to work for the end users of the information (students, school administration, policy makers etc). Students use scores to judge their own learning, whether or not it accurately reflects it. Schools want to know the score break down of their students for their public profile. Policy makers take behavior modification approaches to provide educational reform e.g. (more standardized tests, rationing scholarship opportunities by such scores, et cetera).

Additionally, behaviorism approaches have worked tremendously with those that have limited ways of communicating change or information e.g  the older generation, young children with minimal verbal skills, special populations and those suffering from cognitive disorders. It is also useful for those that struggle with addiction. The strength of this behaviorism application is in its reliance on conditioning through positive/negative reinforcement or positive/negative punishment. It recognizes that true change happens over time and not all at once. On this account, it is still a strong voice in psychology. That said, the conditioning aspects of behaviorism can be beneficial when used responsibly, but caution must be applied. In extreme cases, conditioning is simply a pseudonym for mind control especially among vulnerable persons. Especially when we consider that some of the founding stimulus-response experiments in this domain have been tainted with ethical concerns.

With all of its faults and blessings; the behaviorist theory has paved the way for other learning theories that draw on it such as Bandura’s social learning theory or the theories that dissent it like that of cognitivist learning. I personally do not consider myself to be a radical behaviorist; that behavior is independent of the mind. I am a cocktail of  as many theories as I find to complementarily explain the complex process of learning.  Learning to me is an all-consuming affair that involves behavior as well as thinking, feeling, knowing, et cetera. However, all of this leaves me with some uncertainty as regards the mind-body problem; how much gap lies between the mind and the body when it comes to learning?