Archive for the ‘shy dogs’ Tag
I stood in front of the copy machine, not so silently cursing the manufacturer, the store where I purchased it, and the salesman who recommended it. The darn thing wasn’t working. I pressed the number of copies button, hit the start button and nothing. It didn’t work and if that wasn’t bad enough, after setting it up I was going to have to pack it back up again and return it.
The owner’s manual sat unopened on the desk next to the machine. I hadn’t bothered to read it. Why should I? I’ve been using copiers since they were called mimeographs. I practically grew up using them helping my father produce newsletters for his business. It was just a copy machine for heaven’s sake, how difficult could it be. Fortunately I didn’t embarrass myself by picking up the phone to complain to some poor tech support in India. Instead I read the manual. As it turns out the “start” button was not the same as the “power” button. Tucked behind the machine, out of my sight was the all important on/off button that changed the course of my day. This was nearly as bad as the time I called in an electrician to repair a light fixture because I hadn’t screwed in the bulb tight enough.
Many of us assume that because we have lived with dogs all of our lives that we know how they work, what makes them tick. And unfortunately for many of us if we picked up an owner’s manual written by someone without the requisite background and understanding themselves, we’ve been led astray and our lack of success in getting dogs to do what we want is seen as their flaw, not ours or the method we are employing. When this happens the labels start getting slapped on the dog. They’re dominant, submissive, red zone, vindictive, stubborn, lazy, stupid, etc., ad nauseam.
When we are trying to help fearful dogs not be so fearful the way we do this is through counter conditioning, which means we change how the dog feels about the stuff that scares them. It’s not easy and depending on what it is they are afraid of, we may have limited success, but at the end of the day, it’s what we’re doing. How we go about trying is important. The most important piece of this training puzzle is that the scary thing needs to predict a good thing, before the dog has a chance to experience the fear of it. Given how quickly brains and bodies respond to things that scared them in the past, this isn’t always easy or possible. Sometimes we can get away with having the scary thing be not so scary by keeping it further away from the dog, or making it go away sooner rather than later. But we have to quickly follow its appearance with whatever we are using to counter condition. This is usually some kind of yummy food or a toy the dog loves.
We know that we do not reinforce fear by providing a dog with comfort, food or a toy. This is because when we present something to the dog that they like, immediately after or while they are experiencing the scary thing, we are counter conditioning, not reinforcing. But this will only be the case so long as the scary thing is not so scary that the dog can’t begin to feel good about the treat or toy. If I was in a car crash and someone walked up to me, my knees shaking, tunnel vision setting in, heart racing and stomach turning, and they handed me my first Publisher’s Clearing House check for a million dollars, I’m still not likely to learn to love being in car crashes, even if I wasn’t killed or injured. We also know that the emotional response of being afraid can be made worse if we don’t intervene soon enough or do something that contributes to it, such as yelling at the dog, poking them, yanking on their collar or shocking them.
A common error that handlers make is not providing the treat or toy (the US or UCS) soon enough after the appearance of the trigger (the CS). One of the reasons this occurs is because they are waiting for an appropriate behavior to reinforce. This is not to say that rewarding a dog for an appropriate behavior is wrong, but that if you wait too long for that behavior you run the risk of the emotional response the dog is experiencing, becoming stronger or more intense so when you finally do introduce the reward its counter conditioning “power” is lost. This is the case whether you are using positive or negative reinforcement to create an alternate or incompatible behavior. For some dogs even waiting for them to turn and look at their handler takes too much time and their negative emotional response is too strong to change given where you are and what you are using as a reward.
Once the treat or toy has been paired with trigger it is often possible to switch to rewarding for behavior so long as the dog continues to feel happy and safe in the presence of the trigger. When this happens we can start to build duration in the dog’s ability to remain in proximity to the trigger, or to changes in the trigger’s behavior. When it comes to addressing fear in dogs, what are you waiting for?
One of the reasons I go on like a broken record about the importance of using reward based training methods that have been designed based on the evidence available garnered through the study of animal behavior and research is because working with fearful dogs can be so darn challenging. So challenging that if you don’t start seeing improvements soon you might become frustrated and disillusioned and the dog’s behavior can continue to degrade.
It’s the same reason I repeatedly remind people about behavioral medications that can help the process of changing how a dog feels about things that scare them. The risks of putting a dog on an approved behavioral medication for a few months, following the protocol recommended by a veterinarian, may be fewer than the risks we take by continuing to expose a dog to triggers without them. We can add more fears to a dog’s list of triggers, or further sensitize them to the ones they already have. It’s something to think about.
The gold standard for working with fear based behaviors in dogs is to use a combination of desensitization and counter conditioning. These are easy enough to understand, but not always easy to implement successfully. When a dog’s behavior does not improve, though the handler is employing these techniques, there are some common “fails” that may be occurring.
One common fail is to expose the dog to what scares them at a level that overwhelms them. It could be that the scary object or event is too close, too big, too many, too loud or around too long. Being able to eat treats is not a guarantee that a dog is what we commonly refer to as under threshold. It is possible for a dog to be motivated enough by something, to tolerate something scary or unpleasant to them in order to get it. It’s why it’s not recommended that a dog who is afraid of people be invited to take treats from a stranger. The same way you might be willing to pick up a paycheck every week and still hate your job, a dog may be willing to snatch a treat from someone and still wish they weren’t there. This does not mean that we can’t help a dog who is routinely over threshold, sometimes we have no choice, but until you have a good relationship with a dog and have given them coping skills it’s best to strive for less bothered rather than more.
Another fail is to assume that you are actually counter conditioning a dog to what it is they are afraid of. Our understanding of classical conditioning is based on the work of Pavlov, the man who turned getting dogs to drool into an art form. Classical conditioning is learning by association. We all do it, all the time. Counter conditioning is changing an already established classically conditioned response. A dog who is afraid of ________ learns to love children, loud noises, other dogs, car rides, vacuums, getting their ears cleaned, men with hats, etc. The scary thing which once predicted being scared now predicts cheese or a frisbee toss. It can take countless repetitions for some dogs to get this new association to replace the old one. A handler may be feeding steak in the presence of a trigger for years and not make this switch. The problem may be that the trigger is not what is predicting the treat for the dog.
Life is not always orderly. What can seem obvious to us is not to our dogs. If there is something that is relevant to us we often assume it is relevant to others, and it is not. There can be things and events in the environment that take precedence over another for a dog’s attention. We may be assuming that because we noticed the trigger and fed our dog treats, that the dog will make the association that it was the appearance of the trigger that made the treats appear. This isn’t always the case. Even if the dog notices the trigger it might not be the event in the environment that the dog is learning makes treats appear. If this goes on long enough, you reaching for a treat when a kid on a skate board goes by, the dog may eventually learn to feel ok about the skateboarding kid, but not as quickly as he would if it was the kid on the skate board that predicted the treat, and not your hand movement or that you stopped and turned in a particular direction.
Another common fail is that whatever is being used to counter condition is simply not good enough. Many dogs will eat anything, any time. I have no trouble motivating my dogs for a training session using treats, after they’ve had a meal. This is not true of all dogs, but by my dogs’ reactions to seeing me gather up training paraphernalia; clicker, treats, target stick, toys, bait bag, you’d think they’d never had a square meal in their lives. One of the reasons for this is that it’s not just the food that they enjoy. Figuring stuff out is fun for dogs too. But when you are working with a dog who is really afraid of something whatever you are offering them to create a positive association, needs to be amazin. Sometimes this is tough, and is why we combine counter conditioning with desensitization, to tip the scales in our favor. Suffice to say if someone wanted me to feel good about seeing Rush Limbaugh walk into a room they’d have to take out a loan. It’s not always easy to change how a dog feels about something or someone.
As the dog’s emotional responses change we can increase the level of their exposure to a trigger and we may find that what used to require filet mignon to get a tail wag only requires a smile and word of praise from us to get a positive response from our dog. If what you are doing isn’t working, it’s not that the process of desensitization and counter conditioning doesn’t work, it’s that your technique may need some work.
There is so much progress that needs to be made in regard to how people “think” about animal behavior and training that it can seem overwhelming. But seven years ago I had to seek out and search for information regarding the most humane and effective ways to help dogs with fear based behavior challenges, whereas today it streams on my Facebook page and twitter account. Articles like this one, sharing the research done on how our intervention when a dog is scared can help alleviate their fear, is becoming mainstream. The science and research is being repackaged for mass consumption. It’s about time. But don’t think that everyone is buying it.
There are those who, due to an inconsistency in the terms we use to “talk” about behavior, will go on endless semantic journeys to dispute the claim that “comforting dogs does not reinforce fearfulness.” Comforting will be termed coddling and the methodology for applying either will be criticized. In one way it’s good. It means people are thinking, but when pieces of the puzzle don’t belong there, it’s difficult to come up with the right picture.
One such piece is the misunderstanding that people have regarding the use of the term “reinforcing” and how it is applied to behavior versus emotional responses. Behaviors that are reinforced can be expected to increase. Behaviors based on powerful emotional responses, if paired with what one might label a “reinforcer” (the same bit of cheese that increased the chances that a dog would sit when asked) cause a decrease in the emotional response, and subsequently we are likely to see a decrease in the behavior associated with the emotion. This is because we are counter conditioning the emotional response. Not all behavior is created equal.
Let’s use “hunger” as an example. Though not exactly an “emotion” if we feed an animal who is hungry, their hunger will decrease, and unless there is an eating disorder involved, the behavior associated with hunger, eating, will decrease. We do not reinforce hunger by feeding an animal. We do not reinforce fear when we comfort an animal. In both cases what constitutes food or comfort is dependent on the animal’s definition of them. One dog may find being stroked and held comforting, another might find it annoying. A hungry lion would not look at a bale of hay and see a meal, but a horse would. In either case if we know what a dog finds comforting or an animal thinks is tasty, and give it to them, we are likely to see a decrease in the behaviors associated with either being scared or hungry (after they’re done chewing of course).
I am going to propose that since the word “comfort” seems to be difficult for people to accept, even though it can be clearly defined-
1. To soothe in time of affliction or distress.
2. To ease physically; relieve.
1. A condition or feeling of pleasurable ease, well-being, and contentment.
2. Solace in time of grief or fear.
3. Help; assistance: gave comfort to the enemy.
4. One that brings or provides comfort.
5. The capacity to give physical ease and well-being: enjoying the comfort of my favorite chair.*
can be replaced by the term “to support.”
My aging border collie Finn was walking down a flight of stairs in our house when I could hear his nails scrambling on the wood. I got to him in time to prevent him from tumbling down. I helped him right himself and supported his hind end as he continued down. At the bottom I gave him a cheer and opened the door so he could go outside. I provided him with what he needed to get down the stairs unharmed. It is likely that he will avoid the stairs, and until I can put a runner down I would prefer that he did. But I did not want him to be injured to learn that lesson. Ultimately, when it is safe for him to do so, I want him to continue to go up and down the stairs on his own.
When we offer a stressed and scared animal our support we do so based on the needs of that animal. It does not make sense to state unequivocally that we should not attend to these needs because we may not be clear on what that support should look like. Or having provided that support inexpertly in the past it is proof that it doesn’t help. Our goal is to help a dog develop the skills and confidence they need so that continuing support becomes unnecessary, but until that happens it would be foolish to stop providing it.
Advising that supporting someone trying to learn to swim will keep them a life-long non-swimmer doesn’t make sense, and it’s dangerous. Someone may not need you to hold their hand as they walk into the waves, but someone else might lest, they be swept under and drown.
It would seem that it is too much to expect that people who decide to make a living from “working” with dogs, actually spent some time learning about them. There are plenty of jobs out there that don’t require any education for someone to excel at them. If you want your lawn mowed and raked hiring a neighborhood kid with the tools is probably not a big risk. But if you want your fruit trees pruned you’d better be careful before you give that kid a pair of clippers.
Awhile back I caught grief for suggesting that as admirable as her intentions might be, having a 15 year old run an animal rescue might not be the greatest thing in the world to happen to animals. A similar thing happened when I suggested that stopping to let a newly freed group of rescued laboratory beagles out of their crates to explore the great outdoors, might not be the best choice. That any animal copes and thrives when handled inexpertly is not an excuse for the handling method.
The pet industry is booming. Anyone can label themselves a trainer or behaviorist. Log on to any chat board and it’s apparent that many of these so-called trainers base their understanding of dog behavior on what they’ve seen on television. There’s a saying that if you can’t “dazzle them with brilliance, baffle them with bullshit.” One of the problems is that many of the bafflers have no idea that they’re doing it (some do and are laughing all the way to the bank).
I was asked to go through the hand-outs that a vet gave to new clients with puppies. The information came from another vet who made the rounds on the speaking circuit. I was aghast at what I read. Despite the lack of evidence that dogs form rigid social hierarchies or live in packs, the literature was full of advice for owners on how to securely position themselves at the top of said hierarchy. The methods for doing so ranged from cruel to absurd. This came from a vet who, like the rest of us, has access to information about animal behavior and the science of learning. That they were unaware of this information, or chose to ignore it in favor of their own, scientifically unfounded hypotheses on behavior, is inexcusable.
Anyone can start a day care center, dog walking or sitting service. These are people who will be in direct contact with an animal, or numerous animals. If you’ve got a happy dog who only needs to get out for walks, finding someone who enjoys being with dogs may not be a bad decision. But once someone sets themselves up in business they should be held to a higher standard of behavior. I recently heard about a day care center that employed one of the most misunderstood forms of punishment, the time-out. A dog who became overly aroused and did not play well with others was removed from the group and isolated in an area where he could see, but not get to the group of dogs continuing to play. Did the staff expect that by doing this the dog would come to the conclusion that he needed to play more appropriately?
Since I brought it up, the time-out is a tool that is used to end a dog’s ability to continue to perform an inappropriate behavior and then set them up so they can be rewarded for performing the right one. Putting a dog into a time-out and expecting that they’ll come out when they decide to apologize is as silly as it sounds, but yet, it is not far off from what people think is going to happen.
Someone who labeled themselves the “top behaviorist” in their country (I discovered that it’s crowded up there at the top with other self-appointed “tops”) called me an idiot for suggesting that comforting a fearful dog did not tell them they were correct in being afraid. Though they had an impressive history working with and training dogs, they held no certification or credentials as a “behaviorist.” This is not uncommon in the dog world and pet owners would be wise to cotton on to it.
When someone decides to label themselves a surgeon, and goes on to perform surgeries, they end up in jail when they are finally caught. And within the medical industry a surgeon who practices psychiatry, without first taking (and passing) the requisite courses is also frowned upon. If you are inclined to suggest that performing surgery and training dogs are completely different things, maybe you should think about it from the dog’s side of the equation. Screw up a gall bladder operation and you might end up with a dead patient. Screw up teaching a dog to stop resource guarding and there’s a good chance you end up with a dead dog. As someone who takes animal behavior and training very seriously, this thought is never far from my mind.
If we truly love and care about our pets as much as we claim to, we have to put our money where our mouth is. The hope I hold in my heart is that when the art and science of dog training and behavior modification is respected for what it is, and people who put the time, effort and money into learning about it, are both respected and compensated for it, one day the knowledge that we have will filter out into the general population. It will replace the misinformation and myths currently touted and adopted as truth, and dog trainers will have fewer behavior “issues” to deal with and can focus on teaching dogs to open the refrigerator and get their owner a can of soda.
On my book shelf is a CD by Wayne Dwyer on the power of intention. It’s an inspirational presentation. But as powerful as intentions are, given the right set of circumstances, habits will win out.
Recently I moved the app icons on my iPod around. I deleted a few and moved some off the main screen. In the process a couple of the ones I use regularly shifted their position. My finger moved to tap the icon for an app that was no longer in the lower left hand side of the screen, but to the right and up a row. I managed not to open the wrong app, and stopped my finger in time. My intention was to open an app, and my habit was to tap the lower left of the screen to do it. I still catch myself aiming for the old, now incorrect, location. Old habits are hard to break.
For up to a year after we moved from one house to another, about a mile down the road past the old house, I would on occasion discover that I was preparing to turn into the old driveway. One day I even made it up the driveway to the house before it dawned on me that I had made a mistake. It wasn’t that I was sleeping at the wheel, I’d not driven off the road into the river, so some part of my brain was doing its job. My intention had been to drive home, but an old habit kicked in.
We want to take advantage of the power of habits when we are working with our fearful dogs. If we can create behaviors that require little thought on the dog’s part, it will be easier for them to behave appropriately in scary situations. If we create those behaviors using positive reinforcement, performing those behaviors can include a positive emotional response at the same time- more bang for our buck.
Sitting and looking at me is a “trick” I’ve worked on with Sunny for years. He will plop his butt down and look at me with the slightest prompting on my part. When the pressure is on, he will do it. It’s become what we call a “default” behavior. If he’s not sure of what else to do, this is his fall-back behavior. I’ve rewarded him with food and praise, a lot, for doing it. It’s proven to be useful at the vet’s office and at the groomer’s. When we’re out and there are people around, he will do it and I can step in between him and the approaching monsters.
Consider helping your dog create the following habits:
- Look at you regularly for feedback
- Feel good when they hear their name
- Sit or lie down easily wherever they are
- Stay or wait when asked
- Come joyfully when called
- Play daily
- Become addicted to learning new tricks
There are some things I take for granted, things that have become part of the fabric of my life with dogs, that I forget that at one time they were new ideas to me. Creating conditioned reinforcers to use in a variety of ways with my dogs is one of them. Here’s a quick primer on what conditioned reinforcers are, how you create them and what you can do with them.
A conditioned reinforcer is anything that is paired with a primary reinforcer, often food, in a way that imbues the conditioned reinforcer with the impact of the primary reinforcer. The conditioned reinforcer causes the dog to feel the way they feel about the primary reinforcer. The sound of the clicker is a conditioned reinforcer. There are other names for conditioned reinforcers, such as secondary or tertiary reinforcer, but all are conditioned reinforcers. A conditioned reinforcer, before it’s been conditioned, or the dog has learned to make the association between it and the primary reinforcer, has no intrinsic value or does not necessarily create a positive emotional response in a dog. The sound of a clicker, the words “good dog!”, a dog’s name, have little or no meaning to a dog, but we can change that. Once we do, a conditioned reinforcer can be a powerful tool in training a fearful dog.
The way we can create conditioned reinforcers is simple, we follow the presentation of whatever we want the dog to feel good about, with a food treat, repeatedly. The dog hears, sees, feels or smells whatever we are conditioning them to and they anticipate a treat will follow. Eventually, even if a treat does not follow, once something has been conditioned, the good feeling occurs. For years an aunt would send me a birthday card with a check in it. The check, also a conditioned reinforcer, made me feel good. When I would open the mailbox and see an envelope from my aunt, I would feel good, even before I opened it. When I got older the checks stopped coming, but before I realized that the routine had changed, I still got that oh goodie a check feeling, when I saw the envelope.
Dogs fearful of people need to be taught that certain things should feel good to them, unlike dogs who grew up having positive interactions with people from puppyhood on. Praise or petting, which can be reinforcing for many dogs who have positive associations with people, are not necessarily loaded with enough of the feel good effect to cause a fearful dog to work for them. By following words of praise, ear scratches, a smile, your dog’s name, or clapping your hands with a treat, you will have a variety of things you can use to make your dog feel good. If something makes a dog feel good enough, they will try to figure out what they did to make it happen, so they can do it again. This is how we train dogs to do all kinds of things.
How can you use conditioned reinforcers? They can be used as rewards to reinforce behaviors in dogs we would like to see more of as I’ve described above. We can also use a conditioned reinforcer to mark a behavior to point out more clearly to the dog what it is we are rewarding them for. This is how the clicker is typically used. When teaching my dogs to take a nap, I can mark, with either a click or a word, when they lower their head. Then I toss them a treat. In training this is the deal I make with dogs. They do something, I point it out with a conditioned reinforcer and then I give them a treat, or other primary reinforcer. For some dogs the opportunity to chase a ball or play tug, can replace the food.
A third way to use conditioned reinforcers with fearful dogs is to use their value to change how a dog feels about an event or trigger. Take a phrase, silly boy, for example. Say it and then give your dog a treat, over and over again. The next time something spooks your dog and they startle or cower you can tell them him he is a silly boy, and if he is not too terrified the words can help change how he feels. Some people call this jollying.
I use conditioned reinforcers to help scared dogs stop being scared of me approaching or being near them. I pair my arrival with a conditioned reinforcer, often the click of a clicker, and follow that with the toss of a treat. So long as the dog is happy to hear the clicker, go after a treat, my arrival predicts both of these will occur. This is not the same as having a treat predict the appearance of a trigger, which can ‘poison’ the value of the treat. This is an important distinction to make. Think about luring a dog into the bathtub with a piece of bologna. Imagine how the dog will respond the next time you try to lure them anywhere with bologna. You don’t want this to happen with your conditioned reinforcer.
Pick a few things you can turn into conditioned reinforcers for your dog. Choose words or actions you normally do without thinking about it. Is there something you say when a dog does something you like? Do you cheer or give your dogs a thumbs-up? Grab a bowl of their favorite treats, say it or do it and toss your dog a treat. Repeat until you use up the treats. During the next few days carry treats with you and do the same thing whenever you say or do what you are conditioning. You can work on several throughout the day; smile-treat, yahoo!-treat, good dog-treat, happy hand waves-treat, etc. You will be creating a collection of feel goods and if there’s anything a fearful dog needs, it’s to feel good.
I love this. Flat out love it. At the Delaware Valley Golden Retriever Rescue in Pennsylvania adult survivors of puppy mills are provided with skills to make the transition to being pets. Not only does DVGRR accept mill survivors they actively seek them out and make relinquishing these dogs easy. Their no questions asked acceptance policy means that more dogs will find their way to them for help.
It’s difficult to imagine what it’s like to have spent every day or your 5 or 6 years of life in a cage. Understanding the challenge of leaving that cage and being confronted with the stimulation of the world outside the mill is important for handlers. As well-meaning as many of the people involved in rescue and foster care might be, the implications of early deprivation are often not fully appreciated or acknowledged. In the rush to give a dog new life dogs are frequently overwhelmed and frightened.
Project Home Life provides mill survivors with the opportunity to learn about houses and the things in them. Several times during the day trainers and volunteers bring dogs into an apartment set up specifically for this purpose. Common household sounds can be scary to dog who has never heard them. A refrigerator door opening can be startling, as can chairs moving, doorbells ringing, and footsteps on stairwells. Assuming that a dog will somehow get used to these and the myriad other new objects and events they will be exposed to, is naive. Pet owners or foster care givers, as committed as they might be to a dog, may not have the skills for effectively desensitizing and counter conditioning to home life and the complete immersion may be too much for a dog regardless of their handler’s skills. Some dogs will adapt but many others will spend their first days or months cowering in a back room. Project Home Life minimizes the likelihood of the latter occurring.
I wish I had known about this program when Sunny first came to live with us. I might have been more sensitive to his needs and his early life with us might have been easier. Even if you are unable to recreate Project Home Life exactly, given your access to resources, understanding why this step is important may help you create new protocols for special needs dogs.
Language is important. The words we use to convey ideas matter. Times change and language changes with it. It is helpful to know that when someone is describing something as fat, they mean it’s phat. There’s nothing wrong with being gay and happy, or gay and homosexual, but using the word gay as an insult, as in that’s so gay, should be discouraged, even if the kid saying it does not realize its implications.
Frequently I am asked for my opinion on trainers who I have never met or have seen working with dogs. When someone with a fearful dog is going to consult with a trainer, often a coup in itself, the skills of that trainer matter. With nothing other than a website to go on I have to make assessments as to whether or not that trainer has the ability to help a dog struggling with what may be extreme fear based behavior challenges. And helping the dog means helping the owner understand and work with the dog. I am well aware of, and share with owners, the limitations that exist with my long distance appraisals. One of the things I take into consideration is the language a trainer uses to describe the relationship between the owner and their dog.
Years ago, some of the best trainers in the world used the term pack leader to describe that relationship. But times have changed and like a poisoned cue, the term has become outdated and potentially dangerous. There can be endless debates regarding the different definitions of leadership and how we implement that leadership, however one need not have a shred of leadership ability (whatever the heck that means anyway) in regard to dogs in order to effectively look at and come up with ways to change their behavior.
A trainer who advises dog owners to act as leaders may do no harm, and even some good, when dealing with dogs who are only lacking in basic skills and manners. But once you move on to dogs who need more help in changing their emotional and behavioral responses, the leadership recommendation is often sorely lacking and frequently misleading. Owners don’t need to be better leaders, they need a better understanding of what is setting their dog up to behave the way s/he is and the steps to take in order to change that behavior. Even the parent model, or otherwise benign leader model does not give owners the skills they need to effect the changes they want to see.
Dog owners don’t need to become professional dog trainers in order to help their special needs dogs, they need information about behavior and what ends or maintains it. It’s a much simpler and safer solution than encouraging owners to come up with ways to be respected as pack leaders, which is something even dogs don’t have a definition for.