Shock Tactics: Why Banning E-Collars is a Poor Solution to the Problems in Dog Training.

The use of electronic collars ("e-collars") is one of the most divisive issues in world of dog training and welfare.  The collars are already banned from sale and use in some countries, particularly in Europe, and recently they have been banned in Wales, which marks the first ban in an English-speaking country.  Both sides of the debate over e-collars are entrenched, and vociferous in their criticism of the other. Those in favor of the ban claim that e-collars are "barbaric" and those who use them are de facto animal abusers. Those against claim that their opponents have a sentimentalized view of dogs, seeing them as babies to wrap in cotton wool, and that they lack adequate knowledge of how the collars are supposed to work.  

Tools designed to be painful or frightening to dogs should never be used as a first-line approach in behavior modification.  This is the position of the AVSAB, the IAABC, and the CCPDT. For that reason, I believe that shock and prong collars should not be widely available in pet stores or online. They should not be as easy to pick up as a bag of treats, and we should work to make them less available by lobbying wherever they are sold. But banning them entirely won't work. 

I am going to argue against calls from the UK Kennel Club, among others, to place an outright ban on the sale and use of e-collars.  This does not mean I am arguing in favor of their use - I have never used an e-collar on my dog, or anyone else's. I am merely arguing that banning a piece of equipment as an attempt to change the way dog trainers operate would be a mistake. 

Practicalities 

Obviously, electronic collars are not the only way to cause physical suffering to a dog.  Other kinds of corrective collar are not covered by the proposed ban by the UK Kennel Club.  Kicking the dog is not covered by this ban.  By delivering a short, sharp sensation to the dog, e-collars are more likely to be associated with the undesirable behavior.  To contrast, jerks on pinch collars, alpha rolls and kicks in the ribs are less easy to get the timing right and more likely for the dog to associate with the owner or trainer as well as, or even instead of, the behavior.  By focusing on one item, the "ban the shock� lobby risk taking the best tool out of the compulsive trainer�s toolbox, making room for things that are more invasive and more unpleasant.  They are not addressing the underlying issue, which is, I will argue, that dog training remains an unregulated profession.  

An Ideal World 

The lobbyists often claim that no dog "needs" to be trained using positive punishment.  Any dog - every dog - could be successful with non-compulsive methods, and now that the information on how to use these methods is widely available, there is no excuse for using any other technique.  They point to listings of force-free trainers and organizations like the Pet Professional Guild, as well as all the online resources owners have at their disposal, and argue that because these exist, no dog ought to be subjected to any other kind of training than completely force-free. At the present time, however, such claims amount to utopianism. 

There are still parts of many countries that are not covered by force free trainers at all, and there are places where dogs need training outside a home or class environment, like dog shelters, who do not have access to skilled, experienced force-free trainers for whatever reason.  In real life, compulsive trainers who use e-collars might still be the only option for some people and dogs, if those trainers are the only ones who are accessible.  Taking away their e-collars is not going to provoke some kind Damascene moment of conversion to clicker training; it's more likely to lead to a fall back on more aggressive, more violent techniques.  Not all owners are capable of training their own dogs, especially for difficult or dangerous issues like predatory drive and aggression.  Expecting owners to train their own dogs using methods they are not skilled in is not only unfair, it's potentially dangerous.  

Banning e-collars doesn't change these realities, it merely makes force-free trainers look like they�re doing something.  Just because "something must be done" and you are doing something, doesn't mean you are doing what must be done!

Positive Punishment within a LIMA Framework

Force-free lobbyists often use emotive language in their arguments.  The Kennel Club even refers to e-collars as "barbaric".  We should remember, however, that the use of positive punishment can be a part of the LIMA (Least Invasive Minimally Aversive) approach, widely considered to be the standard of best practice in the ethics of dog training.  LIMA is a principle to guide decisionmaking about which approach to take to achieve a training goal.  It is most commonly, although not necessarily, related to the Humane Hierarchy.  If no other approaches are possible given the trainer's ability and the dog's specific problem, and if the training goal is required for safety, then neither the LIMA approach nor the Humane Hierarchy rules out the use of an e-collar or other method of delivering positive punishment. 

The Association of Pet Dog Trainers, and International Association of Animal Behavior Consultants� position statement on LIMA makes this point clearer:

LIMA requires that trainers/behavior consultants work to increase the use of positive reinforcement and lessen the use of punishment in work with companion animals and the humans who care for them. LIMA protocols are designed to be maximally humane to learners of all species. In order to ensure best practices, consultants/trainers should pursue and maintain competence in animal behavior consulting through education, training, or supervised experience, and should not advise on problems outside the recognized boundaries of their competencies and experience.

Competence is the most important factor in dog training.  We need trainers with a flexible set of skills and a commitment to continuing education.  But we should take something else from this statement too: trainers have different competences and the LIMA approach does not mandate any particular technique, only that the technique is the least invasive and minimally aversive given the situation.  This includes the owner�s abilities, the dog�s temperament and issues, and the trainer�s skills. 

Regulation, Regulation, Regulation

A much better solution than banning e-collars - or any other tool - would be to make dog training a regulated profession.  I�ve made some arguments about this in a previous post.

Ian Dunbar famously said:

To use shock as an effective dog training method you will need:

• A thorough understanding of canine behavior;

• A thorough understanding of learning theory;

• Impeccable timing.

And if you have those three things, you don't need a shock collar. 

If this is true, then making sure that all dog trainers have demonstrated their education and skills to a required level would theoretically lead to the end of e-collar use and of all compulsive training, which is really the goal of the force-free movement.  Such regulation would also make for higher-quality training of every other sort, too - many positive reinforcement trainers mistakenly believe that they "cannot do R+ wrong�.  And, on those rare occasions where trainers do feel that suppressing a behavior is the best approach, the e-collar would still be available with all the benefits it has over other corrective tools. 

In demanding that governments ban e-collars, all the force-free lobby is doing is risking the alienation of a whole group of trainers who use them, including as an adjunct to mark and reward and as part of a LIMA approach.  Until we regulate dog training as a profession, there is no way to minimize the kinds of abuse that go on behind closed doors, and we won�t be able to make progress towards regulation unless there�s widespread support among trainers in all camps.  Dog trainers would need to show that regulation has strong support, and put pressure on governments.  

Furthermore, a regulation system will only be truly fair if it is consented to by as many trainers as possible.  Getting rid of one tool and using language like �barbaric� and �cruel� to describe everyone who uses it is not a step in the right direction because it will ultimately and lead to more divisions, less willingness to co-operate and less likelihood that regulation will happen.  Instead of focusing our energy on banning tools, we should be demanding higher professional standards for the people who use them. 

What is an "Obedient Dog"?

Everyone knows what �obedience training� is, it�s the process of teaching a dog to reliably perform a behavior when you give them the cue to do so. But this doesn�t actually mean you�re training a dog to be obedient, even though it�s often described as him �obeying your commands�.  Some trainers baulk at such authoritarian language, because it is suggestive of a hierarchy, which is a highly contentious concept.  

Without the concept of a hierarchy, however, the idea of an obedient dog means nothing.  

The definition of obedience is to submit yourself to commands given by someone in authority.  On this definition, you can�t be obedient to somebody without holding them in authority - the authority doesn�t have to be a person, but it has to be elevated, like the law or moral principles.  

So, an obedient dog is a dog that dutifully does as he is told to do by someone he recognizes as an authority. Without the authority relationship, there can be no obedience. But this doesn�t necessarily mean teaching the behaviors we now call �obedience training� are impossible to teach a dog, merely that the process should rightfully be called something else.  

Two analogies 

The most common example of human obedience training is in the military.  A commanding officer asserts his authority over his platoon by instilling them with respect for his authority.  The private is expected to obey her commanding officer, not to ask �what�s in it for me?�  If the commanding officer had to remind her that she was being paid for her service every time she told her to drop and give her twenty, the private would no longer be called obedient because she was not obeying for the sake of the authority relationship.  Not to mention being charged with insubordination.

Drop and give me the ball! 

In the opposite way, a worker might obey his boss's demand that he stay later with his team to prepare for an important deadline, but the worker is only doing so insofar as it makes his life easier in the long term. Maybe he�ll look for a better, faster way to accomplish whatever the boss needs so he can go home earlier, even if that means not technically doing what he's been told. In this case, we can hardly say that the worker is being obedient, acting for the sake of the boss's authority.  A better description would be "savvy"; knowing when to keep his head down and hoping for a raise come Christmas. There is nothing insubordinate about a worker reminding himself that he is here because he is being paid, or about his boss reminding him of this.  This is because in this case, there is no need for submission to authority simply because it is authority. 

Is training a dog more like a commanding officer and his private, or like a boss and her worker?  Both models are in use by professional trainers and pet owners, but only the latter has traction in the "force-free" community.  We are told that our dogs should want to work with us, and that we should make it so that our dogs want to do what we ask them to do because they have a history of enjoying doing that thing, then in a way, we are not developing obedience, we are building willingness, memory, and reliability. 

If we use the analogy of �paying� a dog his �salary� to work, then we�re effectively precluding the idea that our dog is being obedient.  He�s being savvy, like the worker. Even if sometimes you withhold his paycheck or give him less than he expects, the fact that he is expecting a reward at all means he is not acting for the sake of you as a perceived authority figure. 

Should we reject the concept of obedience? 

Do we need our dogs to be obedient at all, given that we have an alternative model? This is a question for more experienced trainers to answer with evidence; all I can do is give some criteria to meet if we are to reject the notion of obedience with no loss of function.  If we can teach them to do everything that an �obedient� dog does, but in the spirit of willing co-operation rather than dutiful submission, then we haven�t lost anything by losing the concept of obedience.  

If, however, there is a point where our dog is not willing to co-operate, and is determined to do something we don�t want him to do, then we might find that we do have a need for the authoritarian model of obedience after all.  The question of whether �obedience training� should be renamed, and the idea of the perfectly obedient dog retired, hinges on whether there are moments where only the imposition of one�s will on a dog is sufficient to attain compliance.  

"My Dog Ought To Sit"

Owners of dogs with problems are often confronted with the charge that they just haven�t done a good enough job of training their dogs.  Sometimes, the owners themselves believe that if their dogs were less stubborn and more obedient, they wouldn�t bark and lunge at joggers, or charge the door when guests come around.  The perceived failures of obedience in their dog leads them to feel resentment and guilt. In short, many people see all dog behavior problems as obedience problems, and are liable to blame the owner and the dog for having them.  

A Misconception

This tendency to see frightened and/or aggressive dogs as poorly-trained dogs is actually symptomatic of a deeper misconception about how we should see the dog-human relationship. As humans, we exist in a world of agency, responsibility, promises and obligations, and we�re liable to try to impose as much of the model we live in on how we relate to our pets.  But in fact, allowing our human-centric focus on what we owe to each other to creep into what we expect from our dogs can lead to frustration on both sides. 

The key is in how we understand just one simple word: �ought�.  When I say, �when I tell my dog to sit, my dog ought to sit�, what I mean by this shapes my whole understanding of our relationship.  As I�ll explain, the appropriate way to understand �ought� precludes the kind of resentment that�s so toxic to our lives with our dogs.  This is why such an abstract enquiry is, I hope, useful work. 

Two Kinds of Ought 

When I say someone ought to do something, there are two ways to understand �ought�. One is statistical: �when I toss a coin, it ought to land on either heads or tails�.  The other is moral: �when you make a promise, you ought to keep it�. 

The difference is important because moral �oughts� are related to what the philosopher P.F. Strawson calls �moral emotions� like resentment and guilt.  We feel resentful towards someone, Strawson contends, when they fail to fulfil an obligation we believe they have towards us.  We feel guilty when we become aware of our own moral failings.  Lying, stealing, humiliating and breaking promises are all examples of failures to meet our moral duties, and all of these things make us feel resentful towards others, and guilty when we do them ourselves.  

When I say, �my dog ought to sit when I tell her to sit�, it could sound like a moral ought.  If we believe that dogs can be deliberately spiteful or can fail to respect us, it certainly sounds like we�re saying that failing to sit means failing to fulfil the obligation to sit.  So it makes sense, on this logic, to believe that a dog that doesn�t sit when a skateboard passes by, or won�t lie on a mat when guests come over, is worthy of our resentment. 

However, it�s a lot more profitable and parsimonious to see �my dog ought to sit� as reflecting a predictive, statistical kind of ought.  An ought that does not depend on imputing bad character to our dogs when they fail to do what we ask of them.  A coin ought to land heads or tails, but very rarely it will land on its side.  My favorite tea is oolong, so when I visit a tea shop I ought to choose oolong - but sometimes I�m in the mood for genmaicha.  When a dog is under pressure, the likelihood of what ought to happen actually happening will go down.  Seeing �ought� as purely practical means we don�t resent our dogs for failing to perform as we expect - if any frustration or guilt is in the picture, it is us feeling bad for not setting our dogs up for success.  

It�s All Semantics, Anyway

Often in debates with trainers and behaviorists, the charge of �it�s all semantics� is levelled at anyone trying to clarify the meaning of key terms.  I hope to have shown that there is a lot of unpacking to be done even within a single, non-technical word, and that this conceptual work can shed light on how we see the world with our dogs.  The better we get at understanding the underlying concepts at work when we talk about dogs and dog training, the more effective our communication will ultimately be.  

Cognitive Dissonance and Respectful Disagreements

In an argument, it is fundamentally disrespectful to the other person to assume, let alone assert, that the reason they don�t agree with you is because they are suffering from a defect in their reasoning.

Increasingly, psychological terms like �cognitive dissonance� are being used by dog trainers as a way to explain everything from why some people choose to use e-collars to issues with client compliance.  Whilst an understanding of psychology can undeniably help us understand how to effectively communicate with clients and help them through roadblocks in their training with their dogs, I believe that some of these psychological concepts are being over-used. More specifically, concepts like cognitive dissonance are used to immediately explain what is happening inside the mind of the person we are trying to communicate with - whether that�s another trainer or a client - and I think this is essentially insulting.

Cognitive dissonance is, at bottom, a defect in the reasoning process.  It�s a common problem, born from our reliance on heuristics to quickly draw conclusions from available information and beliefs, and it is local, meaning that a person can suffer from it in one area of their reasoning but not in others. In using concepts like cognitive dissonance to account for disagreements, we are essentially privileging our point of view over that of our opponent at a level that our opponent cannot argue with.  We are saying that our opponents are suffering from a defect in their reasoning, and that - tacitly - if they were free from this defect, they would have agreed with us all along. 

If you accuse me of suffering from cognitive dissonance, it�s almost impossible for me to prove otherwise.  Saying �No I don�t� only makes me look defensive, but there aren�t many other things I can say.  It drains an argument of any use and makes it impossible for you to lose, since my statements are automatically made irrelevant by your �diagnosis� of my defect.  In short, diagnosing your opponent with cognitive dissonance in an argument is a cheap way to avoid actually justifying your own claims.  It is an ad hominem attack dressed in a tweed jacket - it appears intellectually profound but in fact it�s deeply dishonest.

Instead of assuming that people who disagree with us or who won�t comply with our requests are suffering from cognitive dissonance or related problems, we ought to start from the belief that everyone is equally reasonable, and work on showing that the arguments themselves are faulty.  If our arguments are strong, we should be secure enough in our ability to muster the relevant facts, lay out our inferences and allow our opponent to try to criticize us.  This is the only way to show our opponents that we respect them as equals.  Failing to respect people who disagree with us as equals - or making them prove their equality - shows a kind of arrogance that will only lead to laziness. 

x

LIMA in Practice: Efficacy and Welfare

As people who work with dogs, we have usually amassed knowledge of many different training techniques.  Our toolboxes are overflowing with products, techniques, protocols, proprietary systems, theoretical concepts, half-formed ideas - all designed to teach a dog how to do what we want.  With so many tools at our disposal, we have to ask ourselves what the best way is to train a desired behavior.  Where do we start?  We need a principle of help us decide what kind of intervention we should select.  

The principle of LIMA was created for just this purpose.  LIMA encapsulates a set of criteria on which we can choose between interventions - we should choose the Least Invasive, Minimally Aversive tool at our disposal.  LIMA, however, is only a practical principle, it is not a justification.  LIMA by itself cannot explain why we should accept LIMA; that would be tautological.  The reason we care what is the Least Invasive, Minimally Aversive intervention is because we care about the dogs welfare.  Training is supposed to make the dogs world a better place.

So, LIMA is an answer to the question, how do I train with welfare in mind?  Seeing it this way, as part of a larger picture, allows us to fold in other important concepts that are not easily captured by seeing LIMA as a standalone principle.  

One of these concepts is efficacy.  Recently, Eric Brad argued that efficacy is a vital consideration in choosing an intervention, and that LIMA is only part of the decision-making process.  He offers an excellent treatment of the concept of efficacy as it relates to dog training, however, I think he is too quick to relegate LIMA to a great place to start. Rather, I think that a complete conception of LIMA necessarily contains the idea of efficacy.

LIMA is based on welfare.  In many cases, dog training aims at improving the dogs welfare.  Let's look at an example:

A dog is afraid of loud noises, but his owner lives right next to a railroad track, with trains making the house shake every couple of hours.  As a result, this dog is living in a constant state of stress and barks like crazy when a train goes past, annoying his owner and their neighbors.  His owner hires a trainer to try to make the dog stop barking.  The trainer notices that the dog is barking because he is afraid, therefore, his aim is to improve the dog's welfare by making him feel better about the trains, which in turn ought to make it easier to stop the barking.  What kind of intervention should the trainer try first? 

The absolutely Least Invasive, Minimally Aversive option in this case is something like an Adaptil pheromone diffuser.  The trainer could instruct the dog's owner to plug it in and hope for the best.  But there is only patchy scientific evidence that these diffusers work, and no evidence that is specific to this situation.  Choosing this approach, then, has a reasonable chance of not working.  This means, the dog has a reasonable chance of continuing to live in misery for as long as it takes for the trainer to decide to try something else. That can't be right. 

By contrast, behavioral modification by classical conditioning and desensitization has been proven to work for decades, in hundreds of situations.  It can be invasive, as it involves setting up the environment to restrict exposure to the trigger.  It can also produce unpleasant feelings if it is done improperly.  But it is effective.  

So, should considerations of efficacy mean the trainer should skip a step in LIMA at this point?  I would say yes, absolutely.  The risk of the dog continuing to experience fear and pain justifies using something that is more invasive.  Considerations of overall welfare have to be paramount.  Effective solutions should take priority over ineffective solutions.  In real life, this means we have to choose solutions that have evidence behind them.  

This isn't an argument in favor of using aversives as a quick fix - the framework of LIMA still applies; don't use a more invasive or aversive technique than you really need to.  As I've argued elsewhere, if it is true that one never needs to use and aversive technique in training, then LIMA doesn't justify doing so. But what a trainer really needs to do must always be determined by the situation of the dog you see in front of them, not by a rigorous structure of interventions.  

Considerations of efficacy can be brought in through positioning LIMA as one part of a holistic, welfare-focused approach to dog training.   The ethical approach to selecting a technique is to choose something that will lead to the dog spending as little time as possible in a state of physical or psychological pain. 

Should Dog Training Be A Regulated Profession?

This week, a German magazine reported that Cesar Millan had been banned from training dogs in Germany because he failed the local exam for dog trainers.  Germany is the first country to make animal training a protected profession, that is, to regulate who is allowed to call themselves a dog trainer.  The story has been widely discussed on social media, where two conflicting viewpoints have emerged.  One is that it's good to have a central place that certifies dog training.  The other is that it is not.  I believe that regulation for animal training would be an overwhelming positive anywhere it doesn't already exist, and I'm going to show why those who oppose it are mistaken.

First I'll rule out the idea that there could be a conclusive reason not to have certification.  That means a reason that can't be outweighed by anything - like, if it went against some ultimate value.  I'm going to assume that certification, if it is wrong, is wrong in virtue of the risks outweighing the benefits, not because it's inherently immoral. 

There are two interdependent strains to the argument.  One is that regulation of all kinds is always bad, that is to say, it has bad consequences of some kind.  Specifically in this case, that it fails to ensure only good trainers can train dogs, and that it takes power away from the public, or really, the market.  Second is the claim that because there is debate over what classes as science-based and as ethical in dog training, a central authority would be forced to impose a controversial value judgement on trainers and the public.  Such an imposition would go beyond the mandate of any such authority, therefore, we should not centrally regulate dog training. 

I will argue that although there are grains of truth in both claims, neither amount to a convincing argument for reversing the decision to regulate.  They merely draw attention to the need for caution in developing the institution and setting the standards. 

1. Regulation Restricts Choice 

The situation right now in the USA is that there is no central regulation but a lot of different certification bodies and schools, all with different standards and all claiming to be a metric to set some trainers above others.  Why might this be preferable to a centralized regulatory body?  One reason is that it provides choices for individual dog owners, and more choice is good.  Individuals in need of training certainly do have a lot of choice. But the question of whether this much choice is always good is still open; if we assume choice is valuable in part because of a link to a more important good like wellbeing, then being able to choose between dog trainers is only valuable to the extent that it does in fact allow individual dog owners to choose the best fit for them and their dog.  

I would argue that there is too much choice with too little guidance, and that marketing is unduly influential in prejudicing that choice.  Properly implemented, regulation would provide a basic standard; the choice would be between educated, qualified professionals.  The benefits it would provide outweigh the cost of restricting the number of trainers between which a dog owner can choose.  All responsible dog owners want to choose a high-quality trainer who will effectively train their dog; stripping out the charlatans, abusers and delusional people isn't going to cause a decrease in their welfare or that of their dog.  

2. Regulation Won't Promote Quality

The second objection is that for all its lofty aims, regulation has shown itself to be unsuccessful in stripping out the worst of the worst.  This is usually explained as an analogy. We can all remember that one schoolteacher whose lessons we always dreaded or who seemed to teach us nothing at all.  Not all teachers are equally good - some are terrible - but all teachers are subject to the same regulatory standards.  If teaching quality isn't correlated with regulation, we have no reason to assume dog training quality will be either.  Regulation won't succeed in its stated aims.

This is a strawman, however.  Nobody is claiming that regulation in itself will purge training of people who aren't very good at it, or who can pass their exams and then proceed as if they never had an education.  Regulation isn't a silver bullet, but it could be one important barrier to entry.  Moreover, with the best approach to regulation would come a system of formal complaints, so those animal abusers and ineffectual instructors could be subject to sanctions.

This argument is often packaged with a related claim, that market forces or "the power of the public" provides a more effective way to sort the wheat from the chaff than top-down regulation.  This relies on a rose-tinted view of how individuals get their information about dog training.  With clever marketing, coupled with a lack of public knowledge about alternatives, a bad trainer can convince people she's fantastic.  Cesar Millan is doing pretty well for himself despite his recent Teutonic oversight. 

3. Regulation Can Never Be Neutral On Ethics

This objection is less about the theory of whether regulation is a bad thing per se, and more about whether dog training in particular ought to be steered by any one group of people. 

The "training wars" show us that there is extreme disagreement about what counts as ethical, humane training. "Force free" has been labelled as a lifestyle choice, and P+ is often unreflectively compared to animal abuse.  On the other side, trainers who use corrections claim their position is mischaracterized because their techniques are poorly understood. Both sides feel oppressed and victimized by the other.  In this climate, how can we hope to create a judicious, clear-eyed group of regulators?  Is dog training the kind of enterprise where making any judgements on ethics is likely to lead to such bad consequences that regulation is not worth any potential benefits it could bring? 

First off, it is clear that we don't want regulatory bodies to be dead neutral on ethically-sensitive matters.  Euthanasia, abortion, stem cell research, removing children from their biological parents, all highly sensitive and divisive issues and all subject to guidance that comes both from ethics and law.  The fact that a regulatory body has to come down on some issues to do with what it is regulating, however, does not mean they are committed to binary pronouncements on all such issues. 

Nevertheless, there is a grain of truth in this worry.  Ethical pluralism is an important commitment in a broadly liberal, democratic society and regulatory bodies ought to be hospitable to more than one view of what counts as ethical action.  Putting a blanket ban on one style of training or one kind of tool in place risks alienating a lot of trainers and creating a backlash to the enterprise of regulation in the first place. 

Furthermore, questions of what 'counts' as animal abuse are open to a degree of interpretation - proponents of electronic collars, for example, claim that their methods don't cause a significant impact on welfare when they are properly used.  Biologists, psychologists, and philosophers still disagree on the theory of welfare, and there is not yet overwhelming scientific evidence that correctly-applied positive punishment meets the definition of abuse.

A good regulatory system would take into account ethical differences and try to minimize substantive judgements.  In setting a standard of knowledge and skill, regulation could ensure that only trainers who demonstrate a commitment to animal welfare and knowledge of animal behavior can train professionally.  There is no need to come down one way or the other on issues like the use of electronic collars.

Psychotherapy could be an informative analogy.  There are many different schools of psychotherapy with radically different approaches to their patients; many practitioners are vocal in their disagreements with some of their colleagues' chosen methods. Regulation exists, however, and differences of opinion are preserved.  The regulation of dog training could be like the regulation of psychotherapy, with a plurality of methods and different basic assumptions united by a standard of professional competence that conceptually functions as a lowest common denominator. 

Efficacy could play the same role in dog training as professional competence does in psychotherapy - define efficacy and you have a standard of competence. We could design regulation around the potential to be an effective trainer; what sort of basic knowledge would someone need to train effectively?  With this, the professional bodies we see now like PPG could still play a role, that of setting guidelines for best practice within their chosen theory. 

The benefits of regulation, for dog trainers, owners, and dogs, are clear: overall higher standards of competence, a framework for dealing with complaints and less reliance on marketing.  Provided it didn't fall victim to mission creep or the efforts of lobbyists, a regulatory system could allow trainers of all schools to uphold the same standards of competence.  And competence, with its attendant benefits to dogs and owners, is paramount for all trainers. 

Informed Consent in Dog Training

Informed consent is not something we hear of much in dog training circles.  I believe, however, that dog training just is animal educational psychology - we are trying to bring about changes in behavior and, in many cases, in emotional state.  This means we should consider it subject to the same ethics of consent as human educational psychology. 

When parents are making this kind of decision about a educational or psychological intervention for their child, they are acting as their child�s proxy. They have their child�s best interests at heart, but also have to consider the effects of the intervention on their own household.  Analogously, dog owners who are making the decision to undergo training are overwhelmingly doing so with their dog�s best interests at heart, as well as their own and those of other family members. 

So what does informed consent look like between the dog trainer and the dog owner?  We can use the same three key elements that we find in human medical ethics, which are disclosure, capacity and voluntariness.  Of these, disclosure is the most important. 

Disclosure

The dog owner has to know what the risks and consequences of the trainer�s program is before he consents to it.  The trainer has to therefore be transparent and honest about what she plans to do to the dog.  Simply put, dog owners need to know what will happen to the dog during the training process, both physically and psychologically.  You can�t consent to something that you don�t know will happen.  

The second important piece of knowledge concerns whether the trainer is using the most humane and effective method possible. The huge majority of people want a happy dog who no longer does things that get him into trouble, and they don�t want to make their dog suffer unnecessarily. Owners need to know whether there are alternatives that have fewer risks, work faster, or are less painful and invasive. Could the same results be achieved with rewards rather than punishments?  

On the subject of alternatives, owners also need to know whether there is an approach that requires less disruption to their lives, or poses less risk of harm to the dog and the rest of the family.  This is especially important when we are dealing with potentially dangerous behaviors like self-harm, ingesting objects or aggression, where there are risks associated with management during training as well as with the training process itself.  Balancing the risks and benefits of a training protocol for everyone involved is a large part of making an informed decision to consent to it. 

There are some things that trainers have to know too, if they are to be in a position to offer full disclosure of all the relevant information to their client so that he can choose whether to consent. Trainers need to be aware of the alternatives to their own preferred approach and be able to justify why they choose the methods they do.  I wouldn�t trust a therapist who had only read medical textbooks from a century ago and wanted to cure my depression by sticking me in a freezing bath.  Nor would I trust a dog trainer who had never heard of a front-clip harness.  Techniques and protocols change even if the subject matter remains the same, and it is therefore important to understand one�s own approach in relation to the state of the art.  This is true no matter how the trainer identifies herself, whether �force free�, �purely positive�, �natural� or �balanced�.  

Simply claiming that your method works is not enough to secure informed consent because it's not giving the owner all the information he would want.  Especially if pain or suffering is involved in the training process being proposed, because we can fairly assume that most people prefer less pain and suffering to more for their loved ones when they are given a choice.

Capacity and Voluntariness

Capacity is less of an issue than disclosure.  We can assume that if someone has come to a dog trainer with their dog, they have the capacity to understand basic concepts of training.  But one rule is important - the necessity of phrasing descriptions of what is happening in language that anyone can understand.  It is important for the owner and trainer to be on the same level of understanding, even though the trainer's knowledge will be deeper.  A trainer who overuses scientific jargon or the unique terminology from her own particular theory is making it less certain that her client has the capacity to consent to what she is proposing, and introduces scope for manipulation.  

Voluntariness is also related to manipulation.  Although we could assume that because the owner has sought your services, they must be voluntarily agreeing to whatever the trainer is offering, if the trainer uses sales-talk and obfuscation to explain something, she is deliberately creating a picture in the owner�s head that has diverged from what�s really going to happen.  The owner is saying �yes� to what�s in his head, not to what the trainer is offering, and therefore the consent isn�t valid.  

An example of this is showing examples of dogs the trainer says are calm but are really shut down and miserable, by way of a testimonial.  The owner says, �Yes of course I want a calm dog who doesn�t jump up or bark�, but in reality that�s not what he�s getting.  Selling something you�re not really delivering is unethical.  

Trainers therefore have a responsibility to show that their method is the best one - the most humane and the most effective - because they are in the role of an authority.  Owners can read up, but they are not authorities and can�t be expected to know which side of a debate is right if they don�t know the science behind the issue.  There is a lot of information about psychology - human and animal - out there on the Internet, but just as we don�t expect patients in therapy to know the risks and consequences of one kind of counseling technique or another, we don�t expect dog owners to know whether clicker training is better than shaking a can, or whether pack theory makes more sense than pressure/release.  The responsibility has to lie with the trainer, to inform the owner about how their method works in the clearest way possible so that they can be aware of the consequences.  

It's also important for the trainer to respect the beliefs and desires of the dog owner, even when they differ.  For example, an owner may believe a prong or a slip collar to be the best for him and his dog whilst out walking, but may hire a clicker trainer who recoils at the sight of such equipment for help with an unrelated behavior happening in the home.  A trainer seeing people who don't believe the same things as she does as deficient in reason or in compassion is not only disrespectful, it limits that trainer's perception of her own responsibilities.  People who use the information given to them to reach a conclusion the trainer disagrees with, or who are only interested in help for a specific problem and not general advice, are still entitled to that information and they are entitled to reach their decisions voluntarily, without manipulation.* 

The vast majority of dog owners who seek help from trainers just want everyone in their household to be happy, including the dog himself.  If an owner learns that the behavioral outcome they are looking for from their dog trainer could have been achieved with rewards and not punishment, they may well be entitled to claim a lack of informed consent.  

*Think of a Jehovah's Witness in an emergency room; all the doctors may strongly believe the best medical decision is to give him a blood transfusion, but his personal religious conviction is that transfusions are wrong.  The doctors should respect his decision even though they don't agree with it, and they should continue to give him as much medical treatment as he allows.  

The Simple Ethics of Receiving Unsolicited Advice

Yes, sometimes I am forceful with my opinions.  But that�s only because I�m passionate.  And right.  And passionate about being right!  - Frasier Crane.  

As dog owners, and especially owners of dogs with behavioral problems, we often find ourselves on the receiving end of advice.  Sometimes well-intentioned, sometimes less so, but almost always bullshit.  

I�ve lost count of the number of times I�ve been approached by someone I have never met, who has proceeded to tell me everything I�m doing wrong with my dog after having observed the situation for all of ten seconds.  And, invariably, the phrases �pack leader�, �show him who�s boss� and �coddling him� spring up within the first ten words. 

I�m sure a lot of you have experienced the same thing, and I know how upsetting it can be.  But, there�s good news!  I have discovered a simple way to rid yourself and your dog of these irritating know-it-alls.  Better yet, it can be distilled into one simple phrase: �shut up�. 

Yes, I know, that does sound rather harsh.  But really, when you get down to brass tacks, you have no obligation to pretend to a stranger that you have anything other than disdain for the unsolicited, usually-bullshit advice he�s accosted you with out of nowhere.  Why waste your time - more importantly, why risk your dog reacting?  

For those of you who are trainers, perhaps you feel like you ought to be nicer.  Perhaps, you think, these people are presenting you with a �teachable moment�, a chance to engage with someone, save them from the clutches of Cesar Millan and maybe convert them to the Gospel of Positive Training?  

Here�s the thing: most of the time, it isn�t.  Chances are, if someone is so sure of themselves that they�ll approach a stranger to offer them advice, you�ve got about as much chance of having a reasonable discussion with them as you have turning a street preacher into an atheist.  People don�t offer advice in these kinds of situations because they�re looking for personal growth through generous critique, they offer it to feel good about themselves.  To show off.  I�m not going to feed into their ego trip by being anything less than honest about how useful I feel them to be.  

Probably, that makes me a bad person.  Almost certainly, those strangers who freak my dog out and spout crapola think I�m pig-headed as well as a bad trainer.  Whatever.  My first, indeed my only, duty, is to do the best I can for my dog. If that means people don�t like me, it�s a consequence I�ll have to live with.  

Why Aren't Dog Licences Like Drivers' Licenses?

What if we couldn�t just go out and buy a puppy whenever we wanted one?  Is it really okay to pick up a Chihuahua with your cheeseburger, or a Groenendael with the groceries?  

In this article, I�ll be arguing the case for making it more difficult to own a dog.  Dog licenses should be like drivers� licences - granted subject to passing a test to demonstrate that you know enough about basic animal welfare to give a dog a good life, and that you don�t intend to engage in any dangerous practices.

The question of whether dog ownership should be denied to some people is often put in terms of whether dogs are rightly seen as property.  Legally, all animals are property, and so people argue that if we have a right to own property, we must therefore have a right to own a dog.  On the other hand, some people claim that animals are living beings and are therefore better seen as something closer to a family member than an object.  

I�m going to nearly sidestep this particular thorny issue by offering one argument for each side.  I�ll argue that if dogs are fairly seen as property, this still doesn�t mean that people have a simple right to own one - the legal system distinguishes between different kinds of things we can own, and puts restrictions on some of them.  And, I�ll argue that if dogs are better seen as family members rather than property, this doesn�t give us an automatic right to own them like we have a right to family life, because the basis for the right to have our own children doesn�t apply to dogs.  

After I�ve given these arguments, I�ll then discuss what I see as the major obstacle to introducing a dog license, which is the question of how to avoid prejudice. 

Dogs Are Property

Earlier this year I read a blog post that claimed that the reason we all have a right to go out and buy a dog whenever we want to, is simply that dogs are property, and we all have the unconditional right to own property without government interference.  Unfortunately, the author of this article overlooked a class of rights that are conditional, which are rights to own dangerous articles.  

The most common example of this is a car.  Of course, everyone has the right to buy and own a car, but only people with drivers� licenses have the right to drive.  We can only legally do what the car is designed to do when we�ve proven to the government that we understand the rules of the road in theory and in practice.  Or, think about a gun license - in many countries a person can only own a gun when they can prove that they�ve undertaken some training, otherwise they are committing a crime.  

In both cases, the danger of letting someone have an unconditional right to own or operate an object merits interference.  Dogs can also be dangerous, the frequency of attacks on children show this.  The danger an untrained, poorly cared-for dog can pose also merits putting up a bar for people to clear before they can legally own one.  This doesn�t mean that people don�t have the right to own a dog, only that dogs ought to be classified as a different kind of property for the protection of others.  If we make getting a dog license contingent on proving you have a basic knowledge of dog welfare and skillset for handing a dog, we could potentially reduce the number of attacks.  

Dogs Are Not Property

This last argument is unlikely to convince people who just reject the idea that a living being can fairly be seen as someone else�s property, subject to the same rights as couches or minivans.  

So let�s say that any right we have to adopt or purchase a dog doesn�t stem from property-rights.  Where else might it come from?  One common paradigm is that dogs are family members.  The Universal Declaration of Human Rights accords all humans the right to found a family, so therefore we all ought to have the right to take a dog into our homes.  

The key point in an argument against this is that the right to found a family was originally intended as a right to have children of one�s own.  This right to reproduce - to have a part in the creation of a life that carries some of your genes forward into the future - is not quite the same as a right to take a child into your home.  This becomes clear when we look at adoption agencies.  Most agencies from governments and charities have criteria that any potential adopters have to meet before they can be allowed to even meet a child in their system.  In fact, the stringency of these guidelines has been a hot topic in the UK in recent years, because it has been argued that they prevent children finding permanent placements in families.   

My point is, the reason it is okay for adoption agencies to lay out rules and make prospective adopters go through an often arduous process, whereas there is almost no restriction on who can conceive and raise children that are biologically their own, is because having a child of one�s own is seen as a fundamental part of human life.  To deny people that would be to remove their genes from the species in the future, something that veers dangerously close to eugenics.*

It�s clear that buying or adopting a dog has no such biological distinction attached - requiring stricter testing for people who want to buy a dog rather than rescue one seems more like levelling the playing field than denying a fundamental part of human life.  There�s no relevant difference between acquiring a dog from a rescue or from a breeder, so there should be no issue in being required to pass the same tests no matter where you get your dog.

A Problem with Prejudice

So far I�ve argued that making it more difficult to own a dog through requiring that potential owners demonstrate basic knowledge and skills is compatible with seeing dogs as property and seeing them as family members.  Now I will raise a worry about how such a system could be used to reinforce socioeconomic bias.  I�ll make the general case by giving a specific example.  

When my dog was experiencing gastrointestinal issues as a puppy, I signed up to a lot of dog food forums and groups. I was surprised by how heated some of the debates got - whether raw food was better than cooked, whether vegetables and carbohydrates were necessary, anything that mentioned bones - but the one thing everyone seemed to tacitly agree on was that if you didn�t feed your dog expensive food, you were a Bad Person.  If it wasn�t grain-free, artisan-made, air-dried, or fresh from the local organic farm, you might as well be feeding Rex shredded cardboard.  This made the price of admission so high that basically only middle-class people were thought of as good dog owners.  

Dog food is a vital part of dog welfare, so any kind of dog license testing would probably involve a question about feeding.  And, the armchair nutrition experts are a vociferous group, so if there was any consultation about the substance of the test, you can guarantee they�d try to weight it in favour of their own views.

The same goes for dog training.  Some shelters mandate that adopters sign their new pups up for obedience classes, which aren�t cheap.  Nor are they convenient if you don�t own a car - when I got my puppy, the nearest obedience class was a half-hour bus ride away, so by the time we arrived she was already over threshold.  Should all dogs have to go to (expensive) training as a condition for ownership?  What if the family doesn�t speak English, but the trainer doesn�t speak any other language?  Should all dogs have an (expensive) yard rather than an apartment, and an (expensive) insurance policy?  The list goes on.  

The point is, I believe that it�s possible for people from any background to give a dog a good life, and I worry that although introducing a dog license system like we have for drivers� licenses could be beneficial, there is also the risk of making dog ownership an elite activity that people from tougher neighborhoods, with less money and less education, are automatically precluded from.  

So, although there is a case to be made for tougher restrictions on owning a dog, we have to be careful that these restrictions are really doing what we want - weeding out potential abusers and minimizing the risk to dog and human welfare.  

This being said, if the proper care were taken to only include the most important parts of dog welfare and safety on the test, I believe that a dog license system could do a lot of good.  It would eliminate impulse buying of puppies, which in turn would impact puppy mills, and go a long way to identifying potential abusers.  There is no rights-based argument against taking such action, although I highly doubt there will ever be a public base of support for it. 

--------------------------------------------------------------------------------------------------------------------------

*Note here that I�m not saying adoption �doesn�t count� as having a child of one�s own, merely that there is a genetic basis for the right to reproduce that can account for the reason we don�t need to pass any tests to conceive and bear children, whereas denying inappropriate families the chance to adopt a child seems like the right thing to do.  

Dogs Are Not People, But That's Okay

In the last few months, there has been a slew of fascinating new data about the emotional capacities of dogs.  The first comes from scientists� new-found ability to image canine brains whilst they are conscious and alert. We can now be more sure that dogs share many of the same emotions humans do, including love. This new data was summarized in an article in the New York Times, entitled �Dogs are people, too�, a sentiment that has been picked up by animal rights activists, dog lovers and dog trainers.

The second stream of data comes from Alexandra Horowitz and her team of canine behavioral researchers at Barnard.  Horowitz designed an experiment that showed the �guilty look� people see when they come home to find their dog in the middle of a pile of shredded couch cushions, for example, only happens when the dog is being scolded.  It is not correlated with whether the dog actually did anything �wrong�, only with the owner�s reaction to the dog.

Trainers have taken this to show two things.  First, that Horowitz has proven that dogs do not feel guilt at all.  Second, that because dogs cannot link actions to consequences over more than a few seconds and because the �guilty look� isn�t an indicator that the dog know it has done something wrong, we shouldn�t punish dogs when we get home to find a roomful of cushion stuffing.

Furthermore, trainers have taken this to indicate that we shouldn�t punish dogs at all, since it�s not true that they �know what they ought to do�.  Dogs shouldn�t be held responsible for their actions like humans are, even though dogs can behave as if they feel guilt, which is a moral emotion.

The way these two streams of new data about what dogs are have been interpreted has lead to a tension.  On the one hand, there is the desire to show that dogs have so many similar emotions to humans that they ought to be seen as people too. On the other, we want to show that dogs ought not be blamed or punished for their actions like we blame and punish people, because doing so is unjust. 

If dogs are people, but we don�t hold them responsible, then it�s meaningless to claim that dogs are people in the first place.  To say that a dog is a person diminishes what personhood is, and ultimately threatens to undermine both the welfare of dogs and the unique responsibilities of humans.  

What is a person? 

How do we define a person?  The core of my preferred definition is that only a person can be held morally responsible for their actions. Only a person can make promises and be expected to keep them.  Without persons, there would be no good and evil, no right and wrong, just nature.  Being able to be held accountable for failing to live up to moral standards of behavior is what separates persons from the rest of life on Earth, and what gives persons a unique responsibility to do the right thing. 

On this definition, most animals are not persons and not all humans are persons.*  We can use promise-making and accountability to draw a bright line between normal adult humans and everything else.

But wait, doesn�t this collapse into �speciesism�, the unfair privileging of one species above others?  Any assertion that persons and non-persons belong in different categories - that they are different in kind as well as in degree - is now confidently waved away as a simple error in reasoning.

My definition of a person does not amount to speciesism for two reasons.  First, because it is an irrelevance that many people claim only humans can be persons - persons could be aliens, angels or apes and still have this unique place in the world.  Second, because there is a difference between being expected to act a certain way, and being entitled to a certain kind of treatment.  We don't have to claim that only persons deserve to be treated fairly and without cruelty, but we can claim that only actions that persons do ought to be appraised as good or evil, fair or cruel.  

To say that persons are unique is not fundamentally unfair.  Persons are expected to act morally, and they are also entitled to be treated well by other persons.  Non-persons are not expected to be able to make promises, uphold the rights of others or do their duty, but they are still entitled to good treatment because of other features they have.  

Dogs aren�t people because dogs can�t make promises or be held morally responsible for their actions.   This doesn�t mean that dogs aren�t entitled to be treated well; far from it.  Most dogs deserve better than their lot in life.  To say that they dogs are people threatens to have the opposite effect to the one intended by dog trainers.  Instead of increasing understanding and empathy for dogs, we are in danger of forcing them into a human scheme of values.

An insidious speciesism 

When I adopted my first dog as a child, I spent the journey home in the car telling her all the things I promised to do for her.  I promised to feed her, walk her, be her friend, and make her happy.  I didn�t expect her to promise anything in return; how could she?  Dogs don�t choose who rescues or buys them, and they don�t have a say in the rules of their new home.   Dogs make no promises about their own behavior - they don�t promise not to guard the food bowl, or growl at the baby, or come with you everywhere without fuss.  Seeing dogs as people is an insidious harm because it threatens to see dogs as parties to a covenant, and therefore subject to blame and the loss of entitlements if they break that covenant.  

Seeing dogs as people also gives them little room for those behaviors that we can�t fit into the human-like rhetoric of victimhood, past abuse, or mental illness - the kinds of aggression that don�t come from fear, the prey drive, in short, the �animality� of the dog.  When a dog deviates from the human-like mould we�ve force him into, when he chases and kills the family cat, does he deserve less good treatment?  Is he a �bad dog� because he no longer appears to be the simple little person, full of love, that the �dogs are people too� message would characterise him as?  This is closer to dangerous speciesism than the claim that there is something unique about humans that gives us special responsibilities to other creatures.

We humans have a responsibility not to see animals as worth bothering with only to the extent that they provide us with good feelings or resemble the parts of ourselves we like. Animals are an Other that we have to learn to relate to on their terms as well as on ours.  The way a dog, pig, rabbit, or snail perceives and processes the world is radically different from our own worldview, and we should never lose sight of this difference in our dealings with them.  We have a duty to treat living things well because we as persons have the capacity to be good that is unrivalled anywhere else in nature, not because they are people.  

*Great apes and cetaceans are candidates for personhood; although their lack of language makes it difficult to know whether they can keep promises, their cognitive abilities are such that there is an argument to extend personhood to them.

The Case for Genetically Modifying Dogs

Since the creation of the first transgenic mouse in the early 1980s, there has been a great debate over whether it is ethical to deliberately alter animals at the genetic level.  In the current climate, just mentioning the term �GMO� is enough to send many people into a slavering frenzy.  Pictures like this are met with repugnance as well as with wonder at human scientific ability:

The first glow in the dark rabbits were produced at the University of Hawai'i - Manoa in 2013. 

These initial research projects sparked discussion over what the limits of our interference in animal genetics ought to be; there are arguments about the wrongness of �playing God� as well as the threat of a �slippery slope� into Huxley�s Brave New World for humans. 

Although the ability to create new animals �from the ground up� is still a long way off (further still for humans, and there are immense legal roadblocks), the first steps have already been taken towards genetically modifying living animals for purposes beyond research.  The AquAdvantage salmon, for example, has a gene from a different species of fish in its genome, so that the salmon can grow all year round instead of seasonally.  This is, at least, proof of concept that genetic modification is possible on a large scale.  If we can do it in salmon, we could do it in dogs.  

But why would we want to?  In the case of the AquAdvantage salmon, the reason is profit: farmers want bigger fish that are ready for the table faster.  For dogs, the reasons are more diffuse, and so are the potential modifications we could make.

Why modify?

Dogs have a number of roles in human society; they are beloved pets, service animals, sports competitors, livestock guarders, herders, and law enforcement operatives. Historically, dog breeding has been for the purpose of developing animals that fit the roles humans have set out for them.  More recently, dogs have been bred for even more fine-grained purposes, for example, �labradoodles� were originally bred as guide dogs for visually impaired persons with dog allergies.

We are already modifying the genes of dogs though careful breeding, but the ability to modify individuals and breeds of dog at the genetic level has the potential to give us unprecedented power in this project of tailoring dogs for roles in human society. Biotechnology could let us �zoom in� and isolate the genes that play the greatest role in the expression of behavioral and physical traits in the dog, and then decide whether to switch off these genes, switch them out for new ones, or boost their function.*  The technology could allow us to create new and better dog breeds, even more able to do the jobs we ask of them.

More importantly, genetic modification has the potential to increase the welfare of dogs, at both the individual and the species level.  It is this set of reasons that I will be focusing on, rather than on how we humans could benefit - although, the two are not cleanly divisible since dogs do better when they are able to perform better.

In this article, I will make the case for genetically modifying dogs from the standpoint of the welfare of dogs.  I believe that because of the way we have bred dogs for our own purposes, with little regard for their wellbeing, we have an obligation to explore genetic modification as a tool to restore dogs to health.

Classifying Genetic Modification

Genetic modification can mean anything from removing the possibility of passing on a gene for a serious disease, to the creation of human-animal chimeras.  These are two ends of the spectrum, however, and raise very different moral issues.  We need to be clear what we are talking about before we can see the problems with them.  According to the Nuffield Council on Bioethics, genetic manipulation can be divided into �therapeutic� and �enhancement�:

�The way to distinguish between those interventions which count as �therapies� and those which count as �enhancements� is by reference to the condition that is to be altered: therapies aim to treat, cure or prevent diseases and to alleviate pathological conditions which place someone outside the normal range, whereas enhancements aim to improve already healthy systems and to advance capacities which already fall within the normal range.� 

This is a good baseline, although there are some potential interventions we could make in dogs specifically, which would count both as therapy and as enhancement.  I will divide up the interventions we could make into three classes; purely therapeutic; therapeutic enhancements and purely enhancement.

1. Purely therapeutic interventions

Creating different dog breeds has lead to �genetic bottlenecks� within the dog species, which have lead to some breeds being at an elevated risk for certain genetic diseases.  Cavalier King Charles Spaniels, for example, are at risk of syringomyelia, a terrible condition that leads to persistent headaches, seizure and death.  Chihuahuas are particularly vulnerable to luxating patella, a progressive condition of the knees that can lead to episodes of severe lameness.

Almost every breed has a particular set of genetic diseases associated with it.  Our ancestors, in creating today's breeds of dog by strict artificial selection, did dogs a great harm.  A harm that we ought to try to put right, however we can.  Genetic modification gives us the chance to eliminate the diseases we put dogs at risk of, both from individuals and from the species as a whole.

Responsible dog breeders are already working hard to rid dogs of these diseases, but the tools they have to use mean that there is still a long way to go, meaning that many more animals will be born blighted with genetic disease in the meantime.  If there is a duty to remove genetic diseases from the dog's gene pool, there must be a duty to do so as efficiently as possible.  With genetic modification there is the possibility to abbreviate the process with minimal loss of life.  Therefore, when the technology becomes available, breeders have a responsibility to use it.

2. Semi-therapeutic interventions 

On my definition, an intervention is semi-therapeutic when it takes an individual outside the range of normal functioning for a species in order to treat a pathological condition.  This is important because it takes the dog's capacities away from what is "natural" by introducing genes from other animals, which marks a critical moral boundary for some people.

For example, the squashed nose of the pug makes it more difficult for them to breathe properly, which leads to an increased risk of various medical conditions that come from a lack of oxygen.  We could counterbalance this by modifying the pug�s hemoglobin so it can carry more oxygen; making every breath it takes more efficient.

We know, to give another example, that giant breeds of dog such as the English Mastiff can suffer from a variety of health problems that come simply from being bred to be so big.  If we could make the bones of giant dogs stronger than they naturally are, we could prevent many of these problems.

Adding genes from other animals, or even creating genes from scratch*,  could be the key to preserving the health of these breeds without changing their appearance or behavior.  Given that we humans have created these problems though more primitive breeding technologies, and given that any attempt to remove these breeds would probably be met with huge resistance, we have a duty to do the best we can to make the lives of these dogs comfortable so that they can flourish rather than being, as Alexandra Horowitz puts it, �captive in their own bodies�.

Or, to take behavior rather than physiology, we could create a pet dog that is adaptable and calm enough to tolerate the constant imposition of city life; kids, other dogs, being left alone for long periods without developing the kinds of problems that can lead to being surrendered to a shelter.

Doing so would improve the lives of dogs without demanding that humans make the kinds of changes in lifestyle that history has shown many are unwilling or unable to make.  Many people get a dog and then have children, or get a full-time job, or expect their dog to be as comfortable at a sidewalk cafe as on a hiking trail no matter what breed it is.  Although it is possible to teach pet owners to manage their expectations, education is a slow process.  Creating dogs that can do these things without stress is a way to increase their welfare in the intervening time.

3. Purely enhancing interventions

The final class of interventions are the most morally problematic.  Enhancing a dog would mean giving it abilities that are outside the normal range of functioning for the entire species without aiming to treat any medical condition.

With pure enhancement, we could give guide dogs new senses like magnetoception, heightened intelligence, and full color vision so they can understand more street signs.  We could give detection dogs even better scenting abilities, or pre-program them to recognise certain odors.  We could make dogs that live as long as humans.  The question, as the biologist Lee Silver puts it, is, �Why not seize that power�?

Despite the public's association of genetic enhancement with Frankenstein or Superman, there are some morally sound, welfare-based reasons for undertaking such a project.

The needs humans have for dogs often put them under stress and at risk of harm. Dogs that do demanding physical jobs are at risk of injuries; dogs that work in combat and disaster zones are at risk of extreme stress.

If we could engineer dogs so that they were more tolerant to stress, less likely to get injured, faster, stronger, smarter, more long-lived than any dog could possibly be right now, perhaps we could increase the welfare of dogs even further.  We could give them new, enjoyable experiences that today's dogs could never have, or allow them to develop new skills to cope with their world. Enhanced dogs could be a part of society in ways that demand we recognize their rights and prevent their suffering - given that the science is still in its infancy, the only limit is imagination.

Pure enhancement raises the most moral questions about the limits of our creative power over another species, but it is easy to wall off from the other two classes.  We do have an obligation to use every available means to alleviate the harms we have done to dogs through breeding and through expectations, but this obligation does not simply translate to a duty, or even a permission, to enhance non-therapeutically.

Conclusion

I have made the case for genetically modifying dogs based on the claim that we have a unique set of duties towards them given our history of modification through artificial selection.  Genetic enhancement is still science fiction right now, but it is coming ever closer to science fact.  The practical and ethical implications of genetic modification for dogs must be taken seriously by the people who love them now, so we can get clear on what we believe is right before the technology catches up.

--------------------------------------------------------------------------------------------------------------------------

*Of course, the actual process will be a lot more complicated than this - I am simplifying because I�m not an expert in biotechnology and because many of the breakthroughs that would let this happen are yet to be made. 

*This is much further out than the classical method of modification by taking the genes from one animal and transplanting them into another. However, in the last few months a team at the Scripps Research Institute has successfully created the first artificial DNA base pair, suggesting that synthetic genes are a possibility for the future. 

"Dog Sense" and the Appeal to Experience

Just because you�ve had dogs for ten, twenty, even fifty years, doesn�t mean everything you say is automatically credible.   Of course, it�s true that there is some plausibility to the claim that more experience is better - nobody wants a novice dentist - but there�s also the possibility that being experienced just means �being set in your ways�.  There are some ethical pitfalls with the appeal to experience that we need to be careful not to fall into, whether we�re using it or listening to it. 

Confirmation bias

The first is that experience can be colored by confirmation bias.  Often there are arguments from experience on both sides of a debate, but experience only seems worth listening to if it lines up with what you already believe.  A trainer could claim that you should only name your dogs Harold, because in her fifty years of dog training all dogs named Harold were good dogs, but we wouldn't see this as a good argument just because it comes from experience. Experience only has weight when it's combined with our own beliefs about how good the argument is. 

Secondly, as an experienced trainer, if you�ve had dogs for many decades, you�re not likely to remember absolutely everything that�s ever happened with them. Confirmation bias also means it is a lot easier to remember facts that support the position you�re trying to argue for. Experience alone doesn't make you an authority. 

Experience is not a substitute for science

Relatedly, having experience and being sure that your approach works well can lead to a rejection of more systematic, scientific approaches, either because of confirmation bias or because of a feeling that science can�t test for the �natural dog-sense� that decades of experience gives a person.  This can be a problem.  Although most people would agree that an experienced doctor, for example, is better than a novice, we regard a doctor who hasn�t kept up to date with any developments in medicine since qualifying with some suspicion. Having a �feel for� how sick a patient probably is can be useful but not as a substitute for knowledge. This is especially important if the patient is presenting with symptoms that would have been written off as �malingering� or �hysteria� not too long ago.  

New discoveries are also being made in the science behind dog-human interactions; for example, most trainers now believe the majority of aggressive behaviors in dogs are grounded in fear, not anger or competitiveness. This has lead to a revolution in how these cases are treated.  In dog training, many different approaches can �work�, but not all of them are equally ethical or effective.  It could be that a very experienced trainer found an approach that worked and simply stuck with it, without seeing the need to try anything else. This doesn�t mean their approach is better, just that it meets a minimum subjective standard of �working� for that trainer.  We can let our dogs down by assuming that we already have all the answers or that there is nothing new left to learn. 

Subjectivity and specificity

Subjectivity and specificity are the final suite of problems with the appeal to experience. Just because a trainer has had decades of experience raising herding dogs on a smallholding, does not mean they are experts in how to raise a French Bulldog in a studio apartment. Breeds differ, living situations differ, individual dogs and owners differ, so simple experience may not be relevant. 

We need to take special care with picking out relevant from irrelevant experience when we�re talking about behavior modification rather than more day-to-day issues in training. Dogs with emotional and behavioral problems pose very specific additional challenges. Whilst finding someone with the right kind of experience can be invaluable, the wrong kind of experience can make the problems worse. 

If you, as a human, were having problems with a relationship, you might consider talking to a trusted older friend, or your parents - someone who has successfully navigated relationships and has probably dealt with most common issues.  If, however, you found yourself having fits of uncontrolled rage, OCD or panic attacks, you would probably want to go to a psychiatrist - someone who has been trained to deal with these specific problems.  People with depression, for example, are often told by loved ones just to think positive, or to try harder - well-meaning advice that may have helped non-depressed people get over a bad time, but can have detrimental effects by instilling a sense of guilt and failure in people with clinical problems. 

So, how do we know when an argument from experience is a good argument? The answer is in whether the person using their experience to persuade you that they are right, is also giving you enough information to justify their argument independently. The explanation an experienced trainer is offering should be acceptable even if it came from a novice. Experience comes in as extra weight in favor of accepting this explanation rather than another - as I have said, talking to someone who has been there before can be a great help, so long as we take the time evaluate their claims.