Sunday, 29 March 2015

LIMA vs. The Humane Hierarchy: Principles for Decisionmaking

The concepts of LIMA and the Humane Hierarchy are gaining in popularity amongst dog trainers and professional organizations; something I believe is more than past due.  For example, in January 2015 the APDT and the IAABC have put out a joint position statement supporting a Least Invasive, Minimally Aversive approach to decisionmaking in behavioral interventions for dogs.  In both these statements, the concept of LIMA was illustrated with a picture of the Humane Hierarchy.

I believe conflating LIMA and the Humane Hierarchy - seeing both as two  elements within the same concept - is a mistake.  It is an extra step in describing LIMA that seems logical and necessary, but actually can cause problems for the concept.  In fact, it is possible to call oneself a LIMA trainer without subscribing to the Humane Hierarchy.  This is because the Humane Hierarchy is substantial and prescriptive, and LIMA is an abstract principle.  

Defining the terms

LIMA is a principle that states that, for any dog, any situation and any desired outcome, a trainer should always use the Least Invasive, Minimally Aversive intervention possible.  Invasiveness is usually defined both as physically interfering with the dog�s body and personal space, and as altering his routine and diminishing his opportunities for enjoyment.  Aversiveness is defined as anything the dog would seek to avoid if he could - usually punishment and negative reinforcement.

The Humane Hierarchy sets out possible interventions, and groups them according to how �humane� they are, as shown in this roadmap:

At the bottom, there are the most humane interventions; the ones that cause least impact on a dog�s wellbeing.  These are changes in the dog's physical and psychological health, and interventions in environment so that the dog is no longer in a situation where he wants to perform an undesirable behavior.   Next comes positive reinforcement without eliciting any unwanted behavior.  After that, differential reinforcement, where a dog is taught to perform a more desirable behavior instead of a less desirable one, by giving that more desirable behavior a really strong history of reinforcement.

At this point, the Humane Hierarchy starts to move into more controversial territory, with the inclusion of negative punishment, negative reinforcement and finally, positive punishment as a way to change behavior.  Positive punishment is considered the least humane intervention because it is often both invasive and aversive for the dog, and so it is placed right at the top of the roadmap, with several �stop signs� in the way.

It may seem that defining a more humane intervention as a less invasive and/or aversive one just is defining LIMA in terms of the Humane Hierarchy, but this would be an incomplete view of what LIMA can mean as a principle.  In fact, LIMA can exist without the Humane Hierarchy; other sets of possible interventions can be used instead.

LIMA and force-free training

Some advocates of force-free training have criticized LIMA because it supposedly allows for the use of positive punishment and negative reinforcement, thus being a haven for �shock collar apologists�.  The idea is, a LIMA trainer could feel that she is ethically permitted to use an e-collar if she believes that it is the least invasive, minimally aversive option at the time.  In contrast, force-free advocates believe that it is never ethically permissible to use a tools like prongs or e-collars.

One of the main reasons force-free advocates conclude that e-collars are never the right ethical choice, is that they argue that no dog need ever be trained using tools and protocols geared towards positive punishment or negative reinforcement.  They claim that, applied correctly, positive reinforcement-based methods can be effective in every case, so if a trainer is resorting to an e-collar, she either must be deliberately rejecting reward-based methods that would have worked instead, or lacking the necessary education in these methods to have them work effectively.  Any framework that allows the use of positive punishment is allowing something that is, according to force-free trainers, inherently unethical.  Therefore, the framework itself is unethical and ought to be rejected.  

However, force-free advocates who argue that e-collars are never necessary do not need to reject LIMA as a framework - their issue is with the Humane Hierarchy, as I will show.

LIMA simply states that a set of possible interventions ought to be tried in a certain order, it does not stipulate what these interventions are.  Any such set can be ordered in a LIMA way, no matter what it contains.  LIMA just means, use the least invasive, minimally aversive option first.  If there is never a need for a trainer to use an e-collar, as the force-free advocates claim, then an e-collar will never be the LIMA option because something else will work first.

A LIMA trainer does not have to use every tool; they can use any set of tools they believe might be effective.  Force-free trainers can leave out e-collars and tools aimed at causing positive punishment from their set of potential interventions and still use the LIMA principle.  Their problem really lies with the Humane Hierarchy, which includes e-collars, prongs and similar tools as potential interventions even if they themselves would never use them.

LIMA can also be developed as a principle to ground an obligation for trainers to continue their education, as the position statements from the APDT and the IAABC state.  Trainers ought to refine their skills in order to be able to use less invasive and aversive interventions with maximum effect.  If it is true that reward-based methods will work for every dog, anyone using any other methods must be suffering from a lack of education.  Given that there is ample opportunity for trainers to become educated about force free methods, that lack of knowledge can be seen as a contravention of the principle of LIMA training.  For a trainer who is committed to this principle but faced with an inability to find a reward-based method that works for a particular client, the least invasive and minimally aversive thing to do might be to refer to someone she thinks might be better able to help.

Force-free training and LIMA training are therefore not opposed in principle, as the arguments force-free advocates use do not rule out using the LIMA principle to make decisions.  Force-free training can be seen as opposed to the Humane Hierarchy, on the other hand, because of the inclusion of tools that they believe to be essentially unnecessary for behavior modification.  

This is not to say that the Humane Hierarchy does not have a place, or that it is not useful for trainers using LIMA - it is a model decision tree, and without a framework for applications, LIMA would be little more than a wish.  Any application of LIMA will need a set of interventions to reference; it just does not have to be specifically the Humane Hierarchy.  In abstract discussions of the ethics of dog training, LIMA also ought to be considered by itself, as a guiding ethical principle rather than a list of �if this, then that� interventions.  

Thursday, 12 March 2015

Metta and the Shelter Environment

I'm going to do something a little different in this article, and share some personal experiences. In particular, how my practice of Zen Buddhism (on which I am no expert, but I'll try to explain the concepts as best I know how) has informed my volunteer work.  I've been working as a volunteer dog socializer at the Humane Society of Silicon Valley for just over a year, and the experience has taught me a lot about dogs, ethics, and applying Buddhist principles to everyday life. In this blog I'll be focusing on one principle: metta.

Metta is probably best described as "equanimity in compassion".  It means sincerely opening yourself to the feeling of wishing that everyone be happy and free from suffering, without conditions on who they are.

"What do you mean, you can love other dogs??"

Metta isn�t one of the classical area of focus in the Zen schools of Buddhism, but Zen can color the practice of metta by bringing attention to the idea that it's not meant to be practiced with a goal in mind - not to create good feelings for yourself, or to achieve some kind of "enlightenment" or to make your life easier. 

By attending to and trying to develop the experience of metta, we try to bring to mind the essential community of all living beings, and extend feelings of loving kindness without attachment to judgements of who deserves what.

I've found volunteering in an animal shelter a true benefit to my personal practice of metta.  It's really highlighted the roadblocks I face - the conditions I put on how open I am to sincerely wishing others well, which are limiting and damaging.   I'll try to describe two of the main ways spending time at the shelter has deepened my understanding of what it really means to wish happiness for all life equally.

The Difference Between Happiness and Metta

It's easy to wish a living being well, and feel empathy for it, when you see that it has some feeling towards you.  In fact, loving a dog that loves you is almost impossible to resist!  It's much more difficult to give the same love, however, when you're faced with a living being who has no interest in you, or wants to cause you harm.

The feeling of being loved or needed by an animal gives rise to good feelings, and these inspire what I thought was a state of metta, but was actually just attachment.  This feeling of reciprocal love depends on something the other creature is doing.  Metta, by contrast, is ideally universal and independent of context; love for everyone no matter who or what they are.  The difficulty I noticed myself having, in maintaining the same level of compassion and openness towards the dogs that don't care about me and the ones who love me dearly makes it clear where I am falling short in my practice.  Extending openness and love to only those beings who offer me something is the opposite of metta!

When I feel myself closing down, and wishing the interaction to be over so I can go and spend time with my favorite companion, I try to force myself to reopen my mind and remember that the dog in front of me owes me nothing, and deserves everything. 

Love Without Attachment

Another benefit to the practice of metta in shelters for me, is that it has forced me to focus on non-attachment.  Love without attachment is really important in Buddhism, because we believe that attachment causes the world to be full of suffering, and fails to recognize that everything in the universe is transient. 

There is a constant flow of animals through the shelter I volunteer at, which is of course a good thing - it means we�re doing our job well.  This high turnover can be emotionally draining; as soon as you build a bond, they leave with their forever family, their room is filled by someone else and you're faced with building a whole new bond.  It can feel like a repeating pattern.  After this happened a few times, always leaving me feeling sad, I realized that seeing a pattern is itself a mistake and a failure of metta - to the dogs, this is all new of course, and each dog is an individual with its own point of perception.  

Extending the same loving kindness to each dog, approaching the long term resident who covers me with kisses and the stranger who shrinks into the corner or bares his teeth, with an equally open heart can be draining, but it also forces me to recognize that metta cannot be conditional.  Everyone deserves happiness, everyone deserves to be free from suffering.

The first thing I do when I meet a new dog, is to tell them, quietly, "May you be happy. May you be free from suffering".  Extending this wish sincerely to every living creature is a daunting aspiration, and highlights the pettiness and flaws in my character, but it's something I'm committed to, and something I believe will benefit the dogs I am trying to help.