The Difference between Reinforcers and Rewards

Some of you are dog enthusiasts. Some of you are interested in neuroplasticity. I hope this post is of interest to both! Let’s find out….

Recently I sent an email to a group of friends and clients who have been using the Dynamic Neural Retraining System (DNRS) neuroplastic program. The main “work” of DNRS is a behavior that takes between ten and 20 minutes to complete, called the Limbic Retraining Steps (LRS). Program participants are instructed to do LRS three to six times (equivalent to one hour) per day.

One of the main problems I hear about from people who are struggling with DNRS — and it was a challenge for me, too — is completing the daily quota of LRS. I used a lot of “behaviorism tricks” to get myself to do LRS (train my brain) as much as possible. So, in the email I sent to other DNRS students, I encouraged people who were having trouble sticking to their daily program to set up a reinforcement schedule for LRS.

While writing this email, I waffled about which words to use — whether to use “reinforcement” or “reward.” I really wanted to use “reinforcement” because it was more accurate, but I worried that if I sent out an email about “reinforcement schedules” and “reinforcers,” it would sound too technical and jargony. Although I mostly spoke of “reinforcement,” in a couple of places, I wrote “reward” to try to be more “accessible.”

I got an email back from someone who said she does not need rewards to do LRS, that she finds doing LRS rewarding in itself, and that she often attaches LRS to another activity in her day. I thought, “Well, that’s because LRS is supposed to be a conditioned reinforcer, so that is working for you the way it was intended! And then you’re using Premack’s Principle to further reinforce it!”

But that probably would not be meaningful to her because she’s not a behaviorism nerd, like me. But I really want more people to know what reinforcement is so they can use it to better their lives and the lives of the people and animals in their care. So I’m writing this post to try to say some of the things I talked myself out of writing in that email.

At the end of this post, I’ll come back around to LRS and brain retraining and why this all matters….

Here’s the first thing you need to know, the definition of a reinforcer:

A reinforcer is an event that, when it occurs in conjunction with a behavior, and contingent on that behavior, makes that behavior occur more often.

Let’s translate that into English and break it down into its component parts.

First component: Reinforcement makes behavior occur more often. 

(I’m going to try to use a dog example and a neuroplasticity example for each point to make this post maximally relevant.)

Example #1: When your dog sits, you give her a lick of peanut butter. Over time, your dog sits more often. This means that getting peanut butter after sits is reinforcing sitting in your dog.

Example #2: When I wanted to improve my memory and cognition, I asked my family and PCAs not to supply me with the word I was groping for when I would get stuck when I was talking. Instead, I asked them to wait quietly until I came up with the word on my own, and then to point out to me that I had come up with the word on my own and say something like, “Yay, Sharon, you did it!” Over time, I was able to speak with more fluency, with less groping around for words. This means that my behavior of persistence in finding the right words was reinforced by others giving me space to find them and then cheering for my success.

But what if it doesn’t work? What if the behavior occurs with the same frequency or less often than before the “reinforcer”? Then, it was not actually a reinforcer!

Example #1: Barnum hates peanut butter. (I know, it’s weird, but true.) If I put a smear of nut butter on my finger and hold it in front of his nose, he will make a face of revulsion and try to get away from it! So, for MOST dogs, a lick of peanut butter after sitting would reinforce sitting, and you’d get sits more often, but if I did that with Barnum, over time, he would sit less often. That would mean that getting a lick of peanut butter after sitting would punish sitting in his case.

Example #2: My PCAs and family said, “Yay, Sharon, you found the word!” With sincerity. They were honestly happy for me that I was recovering and that my cognitive function was improving. Their happiness and support of my hard work came through in their “Yay”s, so I found it reinforcing. But if one of them had not been sincere when she said this — if I thought she sounded sarcastic or grudging, as if she resented the time wasted while I worked to come up with the word — then it likely would not have been reinforcing of my word-finding efforts. I might have felt embarrassed or ashamed of my cognitive struggles and just been quiet around that person or given up on improving my word-retrieving ability around her. If that person’s response was punishing enough, I might even have generalized that bad feeling and stopped striving to find words around anyone.

Second component: Timing is crucial to reinforcement. The reinforcer occurs in conjunction with the behavior it is reinforcing. This means they either occur at the same time or the reinforcer occurs immediately after the behavior it’s reinforcing.

Example #1: Your dog is wild about balls. Every time you call your dog to you, at the moment when he is running his fastest to get to you, you throw a tennis ball in the opposite direction. Your dog, thrilled, speeds off to get it. Over time, your dog would run faster and faster to get to you, even though he is ending up running off in a different direction for the ball. Because the reinforcer (speeding ball) is being delivered in conjunction with “running really fast to get to my human after she calls me,” the ball throw is reinforcing running to you.

If the reinforcement does not occur in conjunction with the behavior, then what?

If you called your dog, and then immediately threw the ball to him before he had moved toward you, you would not be reinforcing running to you. You would be reinforcing standing still, or more probably you would be reinforcing standing still and looking wildly around for the ball. Sometimes people try to time reinforcements too soon, and then the intended reinforcer actually becomes a lure/bribe or a distraction!

On the other hand, if you timed the ball throw too late, such as once the dog had already gotten to you, you had praised him, and then released him, and he was trotting off to sniff something interesting, at that point you are reinforcing moving away from you or sniffing. Timing reinforcement too late is probably the most common mistake trainers make.

Example #2: After I started swimming again, I wanted to do flip turns when I swam freestyle in the YMCA pool.

I used to really enjoy flip turns when I was on swim team as a teenager. They were fun and gave me a sense of speed and mastery. It was fun to get to the other side of the pool and do a partial somersault in order to turn around. Therefore, they reinforced swimming fast and swimming freestyle.

Then, when I was getting sick in 1995, I started having a lot of symptoms when I was in the pool, including nausea and dizziness. These symptoms became the most severe when I did flip turns, which I found particularly frightening because I would lose track of which way was up, and which way was water or wall, and then bump into the wall with my head. It was largely because of these unpleasant flip turn experiences that I stopped swimming.

Thus, when I started my recovery, I had a mixed reinforcement history of flip turns: I had a longer history of flip turns being reinforcing of swimming freestyle for me, but I had a more recent experience of them punishing swimming to such an extent that it extinguished swimming behavior altogether!

I wanted to turn them back into reinforcers. I worked on this with many behavioral strategies just to be able to start to do them again — a story too long to go discuss here. But, after I was back in the swimming pool and starting to do flip turns, every time after my feet pushed off from the wall, I would say, in my head, “Wheeeee!” as I propelled through the bubbles of my own wake.

Saying “whee” in my head was an effective reinforcer because it reminded me that speeding through the water was fun, that I was going fast, and that I was speeding through the water in a fun way BECAUSE I had done a flip turn. It felt silly and enjoyable. (It also automatically made me relax my shoulders whenever I said “whee” after a turn.) Saying “whee” reinforced flip turns because I did it immediately after the flip turn and while I was still experiencing the direct result of the flip turn — speeding through the bubbles off the wall. In other words, I did it in conjunction with flip turns.

What if I said “whee” at a different time?

If I said “whee” as I was gearing up to turn, when I was feeling a little nervous about doing the flip turn, I would be reinforcing “gearing up to turn.” That might work to help propel me into the flip turn except that actually doing the flip turn was my reinforcement for gearing up, so “gearing up” was already being reinforced. It’s possible that if I’d said “whee” before the turn, I’d find it so reinforcing that I would start skipping the flip turn, but I doubt it. It probably either would have eased the transition into starting the flip or just felt irrelevant (and therefore neither reinforced nor punished the behavior of doing flip turns).

If I had said “whee” after I had finished coming out of the turn and then started my stroke again, it might have reinforced whatever I was doing at that moment (arm movements, turning head to breathe, or kicking), or it might not have reinforced anything because “whee” doesn’t feel congruent with the other activities of swimming, so it might have felt irrelevant and not altered my behavior.

Third component: The relationship between the behavior and the reinforcer is crucial. Reinforcement should be contingent on (be a direct result of) the behavior it is reinforcing.

Example #1: (Recycling the previous two dog examples….)

The dog sitting MAKES peanut butter appear. If she doesn’t sit (if she stands up or lies down, for example), no peanut butter. When she sits, she gets peanut butter. Peanut butter (reinforcer) is contingent on sitting (behavior).

For the other dog, running to the handler MAKES the ball get thrown for him. If he meanders to the handler or sits there looking at the handler instead of running over, no ball. The ball (reinforcer) is contingent on running to the handler (behavior).

Example #2: (Recycling the previous two Sharon examples….)

My searching and coming up with the right word LED TO my PCAs or family saying, “Yay, Sharon, you found the right word!” If I had given up, come up with the wrong word, or said, “Just tell me the damn word!” They would not have said, “Yay, Sharon!” I only got my reinforcer (“Yay, you found the word!”) because I did the target behavior (finding the word I wanted).

Likewise, I only say “whee” when I am swimming as I come off the wall in a flip turn.  If I don’t do a flip turn, I don’t say, “Whee!” in my head. And I don’t tell myself, “Whee!” if I haven’t done a flip turn.”Whee!” (reinforcer) is contingent on flip turning (behavior).

Now let’s see how this compares with the definition for reward*….

Definition of reward:

1. Something given or received in recompense for worthy behavior or in retribution for evil acts.
2. Money offered or given for some special service, such as the return of a lost article or the capture of a criminal.
3. A satisfying return or result; profit.

The relationship here between the behavior and the result are a lot looser! For example, “Something given or received in recompense for worthy behavior,” would fit what my parents did after each semester: take me out for ice cream as a reward for getting a good report card. I think this is a pretty typical way that people reward others. How is this different from the reinforcers I described previously?

For one thing, the ice cream came WAY after the behavior it was intended to reward. I had already done all my homework and taken all my tests for the semester before I got that “reward.” In fact, often the new semester had already started by the time the report cards went home and the evening out happened. The ice cream did not occur in conjunction with the studying and test-taking and homework-doing. There was no temporal relationship between the studying and the ice cream.

Secondly, getting ice cream at Friendly’s twice a year did not actually affect my studying behavior, therefore it didn’t act as a reinforcer. I studied because I cared about getting good grades and about achieving. The test or research paper handed back with an “A” on it or the teacher’s smile when I answered a question right was the reinforcement for the behavior of studying for tests or writing papers or participating in class. If my parents had never taken me out for ice cream, I still would have worked hard and gotten good grades.

Third, the ice cream was not actually contingent on my performance. My parents ALWAYS took me for ice cream at some point after I brought home a report card. They did that whether I had all “A”s or whether I had “B”s mixed in there, too. (Or, when I was younger, when I had a mix of “satisfactory,” “unsatisfactory,” “excellent,” “needs improvement,” or “improvement shown.”)

I am not saying there is anything wrong with rewards like this! I really enjoyed those evenings out with my parents. It felt good to have my efforts recognized. We had a good time hanging out and eating our ice cream. It was kind of a feel-good/bonding ritual.

I think this is often how and why we reward our dogs, too: We give the dog a biscuit for being cute. We invite her up on the bed just because we want to be close and have a cuddle. We throw him the ball because we want to play together. There’s nothing wrong with any of this. In fact, it’s some of the stuff that makes life with dogs feel good! But, we should not confuse it with training or with shaping our dogs’ behavior.

Likewise, as I wrote in a previous post, it was very exciting for me to go to a shoe store to get new shoes after I started walking again. I was thrilled to be there, and it felt rewarding in that I had not been able to be in a shoe store in almost two decades and now my hard work was helping to make this happen. But it was not tied to any specific behavior in time or contingency. I did not increase the amount I was walking in order to make new shoes happen. And having new shoes didn’t make me do more rehabilitation work. It was an enjoyable, celebratory event and a benchmark for me, but it was not really a reinforcer for my recovery work.

Also, rewards come with judgment or intention from a person (either yourself or someone else) to you — YOU are being rewarded. Reinforcers, on the other hand, reinforce the BEHAVIOR, not the learner, can come from anywhere, anything, or anyone, and do not necessarily carry a message.

Rewards carry judgment: You did something worthy, so you are receiving a gift. The judgment is that you did something good, so you deserve something good in return. Or, the reward (as in “just rewards”) can be “retaliation” for “evil” behavior. But in either case, there is a giver, and they are acting with judgment and intention to tell you, “You done good.”

Reinforcers can certainly come from other people. In all my examples in this post, the reinforcers were delivered by the dogs’ handlers or by me to myself with the intention of modifying behavior. But reinforcers can come from anywhere and don’t carry any moral weight. In fact, the environment is a common reinforcer.

For example, suppose you are trying to make a habit of meditating first thing in the morning. It’s a struggle — it’s still dark out, you’re tired, you’re groggy, you’re not sure this is worth it. But, the first day, in late summer, you wake up at 5:30 AM, set your timer for 30 minutes, and shut your eyes and meditate. When your timer goes off, you open your eyes and turn and see a beautiful sunrise out your window. The sky is filled with streaks of pink. The birds are singing. You feel calm and happy.

The next day, you are again not sure if this meditating business is all it’s cracked up to be, but you are looking forward to opening your eyes at the end of the meditation and seeing the sunrise. You position yourself so that you will be looking out the window when you open your eyes. After the 30 minutes, you are delighted to see the sunrise again. You start moving your meditation time a little bit later every day so that, as the days shorten and the sun rises later in the day, you will still end your meditation by looking at sunrise.

The sunrise is not “rewarding” you. Nobody is saying, “That person deserves a sunrise for meditating.” It was just coincidence that the sunrise was coinciding with the end of your meditation time when you started, BUT as a person who is savvy about reinforcement, you recognized the sunrise’s reinforcement value for your meditation practice, so you manipulated that variable by moving yourself in front of the window and changing your meditation times.

With dogs, too, reinforcement can come from anywhere. Suppose your dog slips out the door first thing in the morning one day before you have made it to the hens to collect the eggs. The dog, however, does a terrific job of “collecting” the eggs into her stomach. She decides to do the same thing the next day. Eggs have now reinforced escaping in the mornings. The hens didn’t intend this! The dog didn’t know she’d find eggs the first time. It just happened. And the more it continues to happen, the more escaping in the morning will be reinforced by eating eggs.

Another example is barking. Many dogs find the act of barking reinforcing. It’s not that they think, “Barking is good!” Or, “Barking is bad!” Or, “I’m going to reward myself with barking!” They often bark because they’re bored and it gives them something to do, or maybe for no reason other than they got on a roll and now it’s a habit. In a lot of dogs, it seems to be a cycle of anxiety — they bark because they’re anxious, and then more barking becomes both an outlet for the anxiety and a perpetuator of anxiety. Barking reinforces barking because barking seems to be one of those behaviors that is “self-reinforcing.”

Indeed, reinforcers don’t have to be enjoyable!

To support clarity, I have so far kept my examples and explanations of “reinforcement” pretty limited. I have focused on positive reinforcements and primary reinforcers.** But negative reinforcement is also reinforcement. In fact, I think negative reinforcement played a big role in my recovery.

Positive reinforcement means adding something that the learner will work to get (something desirable) as a consequence of behavior which makes that behavior more frequent. The examples in this post — peanut butter, tennis ball, “Whee,” and “Yay” are all positive reinforcers.

Negative reinforcement means removing something aversive (something unpleasant) that the learner will work to avoid as a consequence of a behavior to make that behavior more frequent. The buzzer in your car that sounds until seatbelts are fastened is a negative reinforcer; when the behavior of seat belting is completed, you are reinforced by the absence of that annoying buzz. (Alarm clocks and timers work the same way.)

How this all relates to DNRS and the Limbic Retraining Steps

At the beginning of this post, I mentioned someone saying that she found doing the Limbic Retraining Steps pretty easy because it was rewarding in itself. I’m pretty sure that Annie Hopper, who created DNRS, designed LRS to be reinforcing of doing LRS because the two longest steps, near the end of the behavior, involve immersing yourself in a happy memory and a happy visualization. (That’s not the only reason they’re there, but it’s an important one.) In this case, then, the PROCESS of doing LRS, itself, carries a reinforcer. If you feel good at the end of it, that’s reinforcing. That’s an example of positive reinforcement.

But there’s another potential hidden reinforcer in LRS. The training involves using something mild to challenge oneself (a low-level trigger) at the beginning of the steps. If you have MCS, as I did, you use a scent, for instance. For me, this meant usually when I started my rounds of LRS, I was having unpleasant symptoms as a result of the trigger/challenge item. Because I had symptoms all the time, sometimes I did not use a trigger like fragrance or foods to challenge myself, but would just take a moment to notice unpleasant sensations in my body, making me more aware of the symptoms that were hovering in the background.

However, often (but not always), by the end of the session, the symptoms had lifted to some degree, and then I felt relieved by feeling better. This was a very potent negative reinforcer!

This means that there can be an element of negative reinforcement in the training program: We start out feeling bad, and then doing LRS removes the unpleasant stimulus (headaches or anxiety or whatever), and the relief from that aversive can be VERY reinforcing!

This was most effective when I did not intentionally trigger my symptoms, but something would happen that would make symptoms worse: If I was in the car, and another car drove by that had diesel fumes; or I picked up a piece of mail that had a fragrance to it; or I woke up in the morning and felt lousy (as was usually the case in the past). THEN, when I did LRS, if I felt better at all, that was very reinforcing. I hadn’t done anything to make myself feel worse; the environment (internal or external) had set up the training situation for me to reinforce doing LRS to feel better. In fact, control of one’s environment is a primary reinforcer — something we all want from birth, like food or air or affection.

Also note that even if sometimes you feel better after doing LRS, and other times you don’t, the behavior is still getting reinforced. This means you are on an intermittent or variable reinforcement schedule, which is actually a really potent reinforcement schedule. You keep working at it in the hope that this time will pay off. (This is why gambling is so addictive.)

What this also means, though, is that the process of doing LRS carries an aversive. The tricky part with any behavioral modification program that employs negative reinforcement is that, until the aversive is lifted, and we have the relief of not experiencing the unpleasantness, whatever was happening up till that moment was being punished! So, having limbic system reactivity to the trigger that results in unpleasant symptoms means we are punishing that reaction — we are punishing our brains sending chemicals through our body that cause symptoms, which is a fine thing to punish. But, we might also experientially connect the punishment to starting our LRS training.

Punishment is not necessarily a bad thing in every case. Punishment suppresses behavior, and that can be extremely important in a behavioral modification program like DNRS that is designed for learners to stop using one neuronal pathway and build and use another instead. And I really don’t think Annie Hopper intends for the program to be punishing — quite the opposite! She goes to great lengths to stimulate and support improved mood and other very enjoyable states. I think that’s very successful for most learners most of the time.

But, some participants might experience the beginning of LRS as punishment, and the problem with punishment is that there is always fallout. Fallout in this context means undesirable side effects from the use of aversives. Fallout includes: associating anything present during punishment with punishment and a desire to escape or avoid the punishing situation.

It might be obvious to you that there is an element of punishment if we do something intentional to trigger mild symptoms, but there can also be an element of punishment even if we are using existing symptoms. If you typically use distraction or avoidance to cope with symptoms, then when you trigger symptoms intentionally, or even if you just let yourself connect with and be aware of symptoms that were already present, that might feel like a punisher to you.

How might one avoid the unpleasantness associated with the first parts of LRS? Procrastinating starting LRS in the first place! For most people, the greatest challenge in sticking to the DNRS program is getting themselves to start doing LRS, whereas once they get going, they roll right along and feel better than (or at least as good as) before they started.

So, if you or someone you’re working with is having trouble sticking with a training program (whether you are training your dog, your child, or your own brain), please know the following:

  • This is normal! It does not mean you are bad, wrong, aberrant, or weak-willed. You are affected by the laws of behavior, just like anyone, and this is actually good news because….
  • It might be because there is an aversive in there that is too harsh at its present intensity for you or your learner to withstand, and you might need to dial it down. This is something you very likely have control over! So, this is something that is possible to adjust to work better for you or your learner, to make the program more likely for you to succeed.
  • It might also be that you need more reinforcement to stick with the program. This is something that can be dialed UP. This is also something that is possible to adjust — that you have a say in — to work better for you and help you succeed.

These are also the kinds of things I love to work on with people when I do coaching. If you’re interested, contact me to set up a session. Or, you may find that, armed with this additional information, you are able to figure out some tweaks on your own!

Happy training!

Peace,

Sharon

P.S. Did you enjoy this post? Would you like to reinforce my behavior? Tell me what made it juicy for you in the comments — if you’re a “dog person” or a “brain person” and what spoke to you! Or make use of the Tip Jar in the sidebar; money is a very potent conditioned reinforcer! (I’ll be talking about why money is such a powerful conditioned reinforcer in a future post.)

*Much to my surprise, I have discovered some dictionary definitions of reward now refer to reinforcement. Such as this one, which is from the same dictionary as the definitions above:

Psychology The return for performance of a desired behavior; positive reinforcement.”

But I still think what most people think of when they hear “reward” is what I describe above.

**I plan to discuss conditioned reinforcers, Premack’s Principle, and punishment more thoroughly in other posts.

Upcoming Posts (with extremely fuzzy timeline):

  • The Laws of Behavioral Science Are In Effect All the Time (whether you know what they are or not)
  • Why you procrastinate: Long duration behaviors and delayed starts
  • Punishment: What it is, what it isn’t, how it works to stop behavior (whether you want it to or not)
  • Premack’s Principle, or how to use your addiction to retrain your brain
  • Conditioned Reinforcers, or how brain retraining is like money in the bank
  • Reinforcers vs. Rewards, Part II: Why feedback is better than praise
  • Clicker Train your Brain: Why clicker training and limbic system retraining are a perfect match

 


Comments

The Difference between Reinforcers and Rewards — 14 Comments

    • COol! Thank you for speaking up about this. Can you say more about what you see the link as being?
      I have been loving everything I’ve learned about TAGTeaching so far, and I have been using it on myself. I did clicker train myself (literally) during points in the program when I was recovering. I recently made myself a tagulator, and I’m planning on making more and giving some away as reinforcers for people I work with and for blog commenters. (So stay tuned!) If I start working with people in person (for example, local people who want to do rehab), I will probly incorporate TAGteaching. I’m curious where/how you’re learning or using it, if you want to share.

  1. This was great! I want to add one thing, though (which I’m sure you know already, but didn’t mention) – if a reinforcer can’t occur at the same time as/immediately following the behaviour- bridging is your friend!

    • Yes, I decided to leave out bridges, event markers, and conditioned reinforcers (all clicker things!) for this post because I was trying to keep it focused. I want to do a post on conditioned reinforcers, and then that will probably naturally lead to an explanation of bridges and event markers. But thank you for letting me know that topic is of interest! It’s always hard for me to judge what’s of interest to others and what is just me geeking out on behavior.

      • If you’re trying to appeal to… well, an audience that isn’t me, do the opposite of what I say! I know this is late, but all I do is geek out on behaviour, so-

  2. Love this! As a volunteer “puppy counselor” with Leader Dogs for the Blind, we were introduced to TAGTeaching this spring at our annual training on campus. Cool stuff to help reinforce our puppy raisers when they are learning to reinforce their puppies. Your examples, both dog and brain, go a long way to making things clearer. Can’t wait for more posts!

  3. As a dog trainer I was trying to find something to share with my clients that explains the difference between reinforcement and reward, and I stumbled upon this. I think the misunderstanding of how each works is critically affecting how people perceive +R training – i.e. the ‘cookie pushers’. Thank you for the very complete definitions of the two, and the perfectly clear examples that you give!

What say you? Leave a comment!