The Most Important Thing...
Let's look at an example that is perhaps a little more complex, but one that is no doubt nearer and dearer to your hearts--what happened in class. The diagram below summarizes this interaction:
Let's look at this example in some detail. First of all, this was a contrived demonstration-- I intentionally said a few nonconsequential things, then dropped the biggie, knowing what you would do. I have been around the track a few times, after all. Also, we are clearly jumping into the middle of an ongoing sequence of behavior--other things have gone on with both you and me getting ready for class, etc. Be that as it may, picking up where the diagram begins, I walked in, shuffled my papers, fidgeted, paced for a while, cleared my throat--the normal stuff. My being in class, notes in hand, having engaged in my usual displacement activities, acted as an SD for me to begin talking. This being a Friday, a few of you started to pay some attention to what I was saying, but by and large, there was pretty sluggish activity on the part of most of you. Then, I said, "The most important thing to keep in mind about stimulus control is..." and all of a sudden faces perked up, people found their places in their notebooks, and a couple of you even made a little involuntary jerk as you grabbed for your pen so you could furiously write down the gems that followed. Clearly, my beginning a sentence with "The most important..." acted as an SD for you. Why?
As we talked about, you have a past history in which a professor saying, "The most important..." has been reliably followed by an important contingency-- whatever follows is likely to show up on the next exam. Likewise, your attending to what I was saying was an SD for me to continue with what I was saying. Notice how in social situations, we attend to what the other is doing as a cue for what we should do--usually! How many times have you been frustrated in a social interaction because the person to whom you are talking isn't attending to what you are saying, but is saying what they want to say irrespective of your response? In an important sense, we feel important when we are an SD for the verbal behavior of others. Let's revisit class the other day, and fill in some of the missing components that are discussed above. Below is a diagram of the same class interaction that adds in the Pavlovian components of the interaction:
Let's focus specifically on the last half of the diagram. Notice that my saying, "The most important..." not only acted as an SD, but as a Pavlovian CS as well. That "little involuntary jerk" that some of you made was clearly Pavlovian. Even if you didn't have an overt reaction, that doesn't mean that Pavlovian conditioning isn't involved. Most Pavlovian responses are subtle, rather than a Seinfeld's Kramer-like whole-body reaction. Again, the same stimulus, my saying, "The most important...", is acting as both an SD and CS. In fact, I maintain that it would not control our behavior if it did not evoke some form of Pavlovian response--that's how intertwined the two are. The fact that you can now take notes on the wisdom that follows my saying, "The most important..." reinforces you having found your place in your notebook, etc. Also note that there are multiple SDs in every situation, and I've only added one of many--not only am I responding to your taking notes, but to myself having said, "The most important...". There are many other SDs here--the classroom, my notes, etc.
Intrinsic vs. extrinsic control of behavior.
So far, we've been talking about the extent to which external contingencies (otherwise known as primary reinforcement, extrinsic reinforcement, SR, or token reinforcement which can be exchanged for SR) control our behavior. But throughout the semester, we've alluded to self-, or intrinsic reinforcement. Basically, the concept boils down to this--some things that we are operantly conditioned to do (initially through extrinsic reinforcement) we eventually wind up doing because the behavior produces enough reinforcement on its own that the activity no longer requires reinforcement external to itself to maintain it. This could also be looked at as behavior that is controlled by non-token Sr and may be only rarely or even never connected with SR. The Premack Principle would comfortably handle this, but let's see if we can also incorporate it into a more traditional analysis. Let's take a couple of examples--reading and drawing.
Reading. We learn to read by being extrinsically reinforced to do so. When I think about how my kids learned to read, we spent a lot of time reading to them, and getting them used to stories, and the enjoyment that they produced. At some point, we started pointing to individual words as we read, then praised (reinforced) them for picking out words, then sounding out words, then reading bigger and bigger chunks, until at some point, they could read. Once they could read a rate and with a level of expertise, their behavior produced sufficient reinforcement to maintain the behavior on it's own without praise from us (extrinsic reinforcement). This was clearly a shaping process. Reading, something not genetically induced, came to be self-reinforcing because of it's consequences. It required extrinsic reinforcement to develop, but then became intrinsically reinforcing. But what extrinsic reinforcement can produce it can also destroy. Let's say that we decide that if a little reading is good, lots of reading is better. How do we increase reading? We reinforce it, of course. So, we come up with a program like Read It and Eat It. You read five books, and you get a certificate to get a personal pizza. Does reading increase? Yes--as long as the program remains in effect. Remove the extrinsic reinforcement, though, and not only does reading decrease, it goes below it's previous intrinsically motivated level.
Drawing can be seen in the same way. Kids are at first taught how to draw, reinforced with praise, then come to like to draw all by themselves. However, if children are extrinsically reinforced for drawing (by getting tokens that can be exchanged for treats, etc.), they will reduce the amount of drawing they do when the tokens are removed (Lepper, Greene and Nisbett, 1973).
A recent article by Robert Eisenberger and Judy Cameron in the American Psychologist challenges the general assumption that extrinsic reinforcement destroys intrinsic motivation. Their meta-analysis of over 100 articles suggests that the phenomenon only occurs under very specific conditions, and is easily prevented.
For another view on conditioned reinforcement, you might want to check out Karen Prior's Frequently Asked Questions about Clicker training at her Don't Shoot the Dog site. Her explanation of the use of a clicker in dog training explains conditioned reinforcement well.
Last modified on March 11, 1998.
