Prelude to the dance

here goes nothing...

As a relatively recent convert to the usefulness of a behaviorist perspective in descriptions of learning, I don't feel strongly wedded to the concepts or vocabulary that Skinner proposed so long ago, as resilient as they've proven to be. My appreciation for them is almost exclusively pragmatic: as long as the behaviorist lens serves my aim of unfolding my own latent strengths, making the best of my weaknesses, and helping other creatures to do the same, I'll hold it happily to my eye. But the moment it threatens to obscure the view, I'll toss it.

I'm fickle that way. When I was a theatre scholar (a role I never wore all that convincingly), I loved the idea of "interdisciplinary" study, and I delighted in conversations that spontaneously transgressed the many boundaries - essential, methodological, temperamental - dividing scientific from literary culture (to borrow C.P. Snow's categories). But formal efforts to close the gap seemed almost invariably to encourage the contempt for "fantasy" that's endemic among scientists and academic fantasists (i.e. literary types). Humanities departments are rife with relevance envy, and many of their faculty are eager to prostrate themselves before every promise of effectuality: you mean I might finally say something, do something that touches the world? Wow, hey. Some expunge the embarrassment of being mere doodlers and dreamers by becoming utterly prosaic (and untrained) scientists. Others eventually take up dog training. Ha.

I mean to say that I'm not immune. I caught the neuroscience bug early, and I got excited a few years ago when the contagion spread among my theatrical cohorts. But nothing quashed my enthusiasm like hearing the moderator of a conference seminar sternly remind us that the terms in play were technical, highly specific, and not to be fooled with in any loosey-goosey metaphorical way. Or hearing a director marvel about how wonderful it would be when she could attach electrodes to her actors' heads and discover what a sense memory really was.

Who's going to speak up for the reality of subjective experience if not for the dreamers and fantasists? What can scientists learn from us - why should they suppose we have anything to teach - if we surrender the stage before the play begins? Loose, sloppy, idle, playful, inexact - these are just some of the many strengths of literary culture. We cavort under the banner of "Not Quite." We spill ambiguity and uncertainty in our messy wake. That's our job, and it matters. Big time.

All this because I want to insist that science (generally) and Skinner (specifically) should not have the last word in every conversation about direct physical pressure in training. Next time!

If you can't say something nice...

joys in dissonance?

...don't say anything at all? This question has recently felt very pressing, as I've been fortunate to get into conversation with some terrific, thoughtful people with whom I happen to disagree on a few points that are near and dear to them and to me (namely, the "proper" use of positive and negative reinforcement and positive punishment in training). Simultaneously, I've found myself resistant to one dominant (ahem!) interpretation of positive reinforcement in the community of "R+ trainers" with which I imperfectly identify. Within that community, when it comes to the negotiation of disagreement among human peers, the operating assumption for many seems to be that direct criticism amounts to a betrayal of the positive reinforcement ethos. Here, as in training situations, we are advised to reward what we like and ignore what we don't.

Insofar as this encourages us to focus more of our attention on the things we hold in common, to affirm the essential worth of the people with whom we're at odds, as well as the particular strengths in their experiences, perspectives, and arguments, this makes great sense to me both in practice and theory. In other words, I like the "reward"/reinforcement part of the equation. But I think the "ignore what you don't like" dictum runs roughshod over the ways that dialogue among contending equals differs (ideally) from even the most open and collaborative training process, which assumes superior knowledge and power of influence in one of the parties. I can attest that when I've been on the receiving end of these significant silences, I have found them positively punishing, as they resound (to my ears) with condescension. My feeling is that, yes, we should hold our tongues if we have nothing nice to say, but it's actually not so nice to end our discussions there.

This is a personal, idiosyncratic response - I realize that my appetite for modestly fractious engagement is unusually strong (maybe a symptom or cause of my theatrical vocation). And "modestly" is a key term there. Especially if the point of disagreement is a tender one, it's hard to trust that the other person won't lose his/her temper, won't fall back on ad hominem nastiness when reason runs aground - almost as hard as it is to trust myself! But I'm pretty sure that I'm not alone in craving the challenges that difficult conversations supply, or in finding them tremendously rewarding not just despite of but because of their difficulty.

So it felt like a delicious bit of serendipity to come across Richard Sennett's new book, Together, at the moment that this conundrum is weighing on my mind. I put a hold on it at the library a couple of weeks ago, because I'd loved his last book, The Craftsman, and I just picked it up today. The introduction alone is a revelation. Sennett is interested in what he calls "difficult cooperation," whether it takes the form of a musical collaboration (he was once a professional cellist), a public policy, or an online conversation. He argues that contemporary life conspires in all sorts of insidious ways to dull our appetite and skills for these demanding encounters, and that we become ever more socially fragmented as a result. I should probably read the rest of the book before trying to summarize his views(!), but here are a couple of choice bits from the first twenty pages:

"In addition to material and institutional reasons, cultural forces today work against the practice of demanding cooperation. Modern society is producing a new character type. This is the sort of person bent on reducing the anxieties which differences can inspire, whether these be political, racial, religious, ethnic or erotic in character. The person's goal is to avoid arousal, to feel as little stimulated by deep differences as possible. The withdrawal of which [Robert] Putnam speaks is one way to reduce these provocations. But so is the homogenization of taste... 'Everybody is basically the same' expresses a neutrality-seeking view of the world. The desire to neutralize difference, to domesticate it, arises (or so I will try to show) from an anxiety about difference which intersects with the economics of global consumer culture. One result is to weaken the impulse to cooperate with those who remain intractably Other."

"Reflexive, self-critical thinking doesn't imply withdrawal from other kids; children can be reflexive together. One piece of evidence [Erik] Erikson provides for this process is game-playing. At the age of five to six, children begin to negotiate the rules for games, rather than, as at the age of two or three, take the rules as givens; the more negotiation occurs, the more strongly do children become bonded to one another in game playing... the very misunderstandings, separations, transitional objects and self-criticism which appear in the course of development are tests of how to relate to other people rather than how to hibernate."

"In the performing arts, the sheer need of others can often prove a shock. Young musical hotshots are often brought up short when they begin playing chamber music; nothing has prepared them to attend to others... Though they may know their own part perfectly, in rehearsal they have to learn the ego-busting art of listening, turning outward. It's sometimes thought that the result moves to the opposite extreme, the musician blending in, submerging his or her ego in a larger whole. But sheer homogeneity is no recipe for making music together - or rather, a very dull recipe. Musical character appears instead through little dramas of deference and assertion; in chamber music, particularly, we need to hear individuals speaking in different voices which sometimes conflict, as in bowings or string colour. Weaving together these differences is like conducting a rich conversation."

Sennett's argument about the general need for more listening and less "asserting" struck a nerve with me, because it's something I recognize as a personal weakness. (Hello, fellow bloggers!) I love this passage and want to learn to live by it: "Usually, when we speak about communication skills, we focus on how to make a clear presentation, to present what we think or feel. Skills are indeed required to do so, but these are declarative in character. Listening well requires a different set of skills, those of closely attending to and interpreting what others say before responding, making sense of their gestures and silences as well as declarations. Though we may have to hold ourselves back to observe well, the resulting conversation will become a richer exchange for it, more cooperative in character, more dialogic."

Rethinking negative reinforcement

Animals in a state of relief.

As I wrote in my last post regarding the increasingly fuzzy distinction between classical and operant conditioning, old terminology can hobble new thinking, and given how awkward the language of behaviorism was at its inception, we shouldn't be surprised to discover how creaky it has become in its dotage. (The surprise lies in its holding up at all!) The positive/negative confusion has never really cleared up for many lay training students (positive punishment? WTF?), and no term has given people more trouble than "negative reinforcement," which bundles all the paradoxes and blurred connotations of behaviorist theory into seven dry-sounding but intellectually and emotionally fraught syllables. Technically, it's negative because it describes the removal of some "thing" (which may not be a thing at all). Colloquially, it's negative because the thing that gets removed needs to be nasty or at least unpleasant in order for its removal to be reinforcing, and so the deliberate use of negative reinforcement implies (and carries the ghost of) the deliberate introduction of nasty or unpleasant things, i.e., positive punishment. That's the theoretical tangle as clearly as I can state it (not very!), and it has significant consequences in practice, as teachers and trainers line up on either side of the R+/R- divide (and take occasional potshots at each other over the crevasse that yawns between them).

Does the theory still encompass what we know of reality? Do the terms describe with satisfactory accuracy our growing knowledge of how animals learn? On the contrary, they appear to be busting at the seams. We're patching as fast as we can right now, but I think our best hope of finding our way to a new kind of coherence (to a description of teaching and learning that covers our collective butts once again) may be to pick at the threads where they're coming unraveled. To combine my canyon and sewing metaphors, these may become the ropes that swing us over the training divide. (Ack.)

Some of the most exciting work in contemporary learning theory is being done by scientists and practitioners (e.g. teachers and trainers) who dare to test the boundaries between behaviorism and humanism; between the body and the mind; between emotion and thought; between psychology, ethology, and neuroscience; between biological and historical accounts of the past; between objective and subjective accounts of the present. On the scholarly and/or scientific side, Frans de Waal, Sarah Blaffer Hrdy, Oliver Sacks, Irene Pepperberg, Marc Bekoff, Mihaly Csikszentmihaly, Alison Gopnik, Timothy Wilson, Gerd Gigerenzer, Antonio Damasio, Daniel Kahneman, and V. S. Ramachandran are some of the great "unravelers" I've encountered (if only on the page), and Jaak Panksepp seems like someone who might actually help us knit a new pattern.

But I think all of us who practice learning theory with focused intent and honest reflection can contribute significantly to the radical revision now underway, and a re-examination of the R+/R- split could be an excellent place to begin. I'm not prepared to say that, as a philosophical distinction, it's totally illusory (I'd like to tackle that question in another post), but as a scientific distinction, it may be. This is one of many places where Jaak Panksepp's work is so fascinating and potentially useful, as he's been investigating the physiological and neurochemical bases of approach and avoidance, of appetite and satisfaction, of aversion and reward. I look forward to the publication of his promised book for the lay reader, because I hope it will make his insights more widely accessible. (Temple Grandin's Animals in Translation remains the best introduction to his ideas for the general reader, as far as I know.) In the meantime, I've been making my way very slowly through Affective Neuroscience and highly recommend it despite its density. I hope I don't distort its content too badly here!

In his book, Panksepp describes a discrete number of affective (emotional) processes whose physiological coherence is marked enough that he is comfortable labeling them "systems." These are activated and expressed in more or less predictable ways by animals of diverse species, and we can guess from our common evolutionary history that there are also strong similarities in how they are subjectively experienced. Panksepp is keen to avoid Skinner's mistake of choosing his terms in opposition to common parlance, so he simply capitalizes the colloquial names for these primal emotions/processes to denote their technical use: FEAR, PANIC, RAGE, and SEEKING. While this group may appear heavily weighted to the unpleasant, the SEEKING system encompasses many varieties of pleasurable anticipation.

If I understand him correctly, Panksepp suggests that most of our strongest appetites or drives (and the emotions that accompany their satisfaction or frustration) arise from various kinds of disequilibrium. A truly safe and contented animal is an animal at rest. FEAR is activated by perceived threats to the self, PANIC by social isolation, and RAGE by constraint (especially of one's access to valued resources). The SEEKING system may be engaged when any of these other emotions is in less than full flower. When we're a little anxious, a little lonely, or a little hungry, our minds/brains are primed to seek out whatever will restore our internal equilibrium: an escape route, a friendly touch, a Hostess cupcake.

In such situations, our minds are also primed to learn, to draw connections between environmental circumstances, our own behavior, and the consequences that result from their meeting. Indeed, our capacity to learn has so many advantages for our continued survival that we are primed to find it intrinsically pleasurable. Thus the SEEKING system affords us pleasures that are largely independent from the satisfaction of consuming a good meal or the relief of escaping a fearsome predator. They're compelling enough to be literally addictive - the SEEKING system appears to be modulated primarily by the action of dopamine, and gets easily hijacked by cocaine and methamphetamine among other stimulants.

In addition, while the research remains sketchy, it appears that the (intrinsically rewarding) SEEKING system is activated whether an animal is seeking out the object of some appetitive desire (food, a mate, etc.) or seeking escape from a perceived threat.

Okay, if you've followed this far, I should finally be able to bring the conversation back around to positive and negative reinforcement and the question of whether they're entirely distinct. Once we start thinking about drive or desire in terms of disequilibrium, it becomes harder to draw an absolute line between the internal pressure of hunger and the external pressure of a bit or a leg; it becomes harder to separate the gift of peace from the gift of an apple. It becomes clear that all effective teaching necessarily "exploits" one appetite or another. And it becomes much more interesting and rich to talk about how to do so in a way that best enlists an animal's SEEKING system and taps into our shared love of learning.

I don't want to tax your patience much further in this post, but in closing I'd like to quote a couple of eloquent descriptions of expert horse trainers who supposedly sit on opposite sides of the R+/R- divide, but who clearly overlap in their ability to help other animals to flourish. I already knew I needed to learn more about Alex Kurland's work, but Cindy Martin persuaded me that I'd better do it soon. She wrote in an email, "When the dog world found clicker training, many people abandoned their leashes, vowed to free-shape everything and never touch their dogs. Well, with horses, we're bound to have physical contact. Riding is about tactile cues. Our weight shifts, we squeeze with our legs, we ask with the reins. Alex developed the idea of pressure as information, below the level of a
true aversive. So is it still R-? Probably. But if we very quickly lighten pressure, by highlighting the first approximations of a desired behavior, with the click/treat, then all these kinds of pressure can be information, simply cues for the horse. And they can still learn to work for 'the release.' In fact, the release of subtle pressure can be a low value reinforcer, once the horse gets more sophisticated, and the click/treat can highlight the especially good responses. Alex calls this process, 'Shaping on a point of contact.'"

Emma Kline attended the same Buck Brannaman clinic in Spanaway that inspired me to write my bumptious letter back in November. You can find some lovely reflections on the SEEKING system on her blog, and you can also find her poetic response to seeing Buck at work:  

"At one point Buck was talking about how extraordinary it was to be with a horse that was hunting the feel. He talked about giving the horse what it wants most in the world: PEACE. No wonder this guy doesn't need to use treats.

I could feel the lines in my forehead getting deeper as I strained to see how he was utilizing the laws of science and behavior modification with an accuracy I have rarely seen. And sure enough, he was using a marker and a reward. His marker was the release and his reward was the Peace of Feeling Together.

I think that it is very important to note that this is not a "peacefulness" that comes from robbing the horse of his sense of security or taking away the little peace he, as a flight animal, is born with. Its about adding a peace the horse didn't have before. That's when horse and human become more than what we were separately. So in fact, the release is a marker and not a reward."

Have we outgrown our vocabulary?

A couple of weeks ago now, Professor Jesús Rosales-Ruiz (of the Department of Behavior Analysis at the University of North Texas) gave an esoteric but fascinating talk on the disappearing distinction between respondent (a.k.a. classical) and operant conditioning. By tracing the history of the terms and describing the difficulties that contemporary researchers often encounter when trying to apply them with any consistency, he exposed their contingency and fragility: while they have been extremely useful as springboards to the investigation of how we learn, they may prove not to have any real substance. They might even have brought us far enough that we can safely discard them (and move forward more easily without their dead weight). Wittgenstein once noted how many stubborn philosophical problems are in fact problems of vocabulary; we are sometimes slow to recognize when we've exhausted our terms.

But dying words (and the concepts or categories they name) have something left to teach. By looking closely at their definitional foundations, and then taking note of their specific failures vis-à-vis reality, we can identify some of the perceptual biases that made them so appealing in the first place. The lay distinction between "respondent" and "operant" has always hinged on the question of whether or not a response to a given stimulus (or set of stimuli) is voluntary, whether or not it can be brought under conscious control. But as Jesús described in his talk, even during Skinner's time, the erosion of that distinction was already underway, as physiological responses that had been considered perfectly autonomic (such as blood pressure) were brought through biofeedback under conscious control. More recently (and provocatively), challenges have come from the opposite direction, as individual cells have been observed in response patterns that mimic operant conditioning. As Jesús noted, anytime that relationships between contingencies (in the environment and behavior) grow measurably more consistent, learning is taking place. Our loyalty to the terms "respondent" (or "classical") and "operant" may obscure the complex but unified realities of that process.

Among the phenomena resistant to any simple respondent/operant dichotomy has been the tendency of certain behaviors to wander from unconscious to conscious and back to unconscious "control." We're all familiar with this dynamic as it applies to our assimilation of complex skills. The famous four stages of competence trace the general pattern, from unconscious incompetence (we don't know we can't do something), to conscious incompetence (we know we can't do it), to conscious competence (we can do it with great mental effort and focus), to unconscious competence (we can do it without effort and without conscious focus). If I'm a skilled driver, or soccer player, or surgeon, as long as the given challenge falls within the range of what is well-known to me and therefore predictable, the relationship between the contingencies of the environment and the contingencies of my behavior may be so consistent as to appear reflexive, and I will hardly have the sense that I am making a voluntary decision at any juncture. Only novelty is likely to wake me from the dream of competence and force me back into a state of conscious engagement.

"Respondent" and "operant," like "unconscious" and "conscious," may only describe different modes of energetic expenditure. The brain is a highly economical organ, a regular Bartleby when it comes to the heavy lifting required for conscious thought. ("I would prefer not to.") That said, it is much more active at an unconscious level than we generally give it credit for being.


Image by Jolyon.

A no-brainer?

What can brown do for you?

As I've been getting ready to formally launch my people/dog training business (and to boldly go where no human or canine has gone before, split infinitives be damned), I've been seeking out advice from many corners, trying to sort and synthesize the ideas that seem most salient to my own particular enterprise. The ideal of "synergy" has become clichéd to the point of comedy, but it's worth pursuing nonetheless, particularly in this context. Given that dog training is primarily about bringing diverse desires into harmonious alignment, it makes perfect sense to extrapolate that goal out to one's business relations. So trainers ally themselves with sitters and vets and shelters and other trainers, and so we all sustain the flow of good things and the trust that follows in their wake.

Well, the muse of synergy tapped me on the shoulder the other day and whispered naughtily in my ear. Why not partner my dog training business with an escort service? Preferably one that employs men as well as women, and preferably one with an emphasis on role-playing and a warehouse full of costumes. How wonderful it would be if -- in my work with a reactive dog -- I could make one call and summon a physically imposing man in a UPS uniform, or a woman in scrubs to play vet tech? One day I might say, "Bring a dozen hats and a long, yellow raincoat." On another day, "Hey, do you still have that Dick Cheney mask? Excellent. A briefcase would be good, too."

I'm pretty sure no one has done this yet, and I think I'd better wait until my business is well enough established that such a partnership would be advantageous to the escorts, too. But anyone who wants to run with the idea has my blessing!

An open letter to Buck Brannaman

In the interests of greater harmony...

Like many who dwell outside the rarefied world of horses and horsepeople, I only recently became acquainted with Buck Brannaman's life and work through the beautiful documentary film Buck that was released last year and is now available on DVD. Buck gave a Q&A at the showing I attended in Portland early in the summer, and made good in person on the charisma so evident on film. It struck me immediately how much his training approach had in common with that of the clicker/marker trainers I most admired, and despite my great ignorance about horses I knew it would be worth my while to attend one of his clinics as a spectator. In late October, I traveled up to Spanaway, Washington with a firm cushion and a warm blanket and planted myself in the bleachers of the arena where Buck taught three separate horsemanship classes every day for four days. (He followed with two more for three days; his dedication and stamina are remarkable.)

I spent most of three days happily lapping up just about everything he had to say and to show about training horses, though I couldn't help remarking that he was somewhat less effective as a trainer of people. But at the end of his second session on that third day, one of his students asked what he thought of clicker training, and he could not have been more contemptuous or less measured in his response. He said he found it worthless at best, exploitative at worst. Good for nothing more than tricks. He recounted a recent encounter with a dangerously spooked steer and joked that a clicker trainer "couldn't click fast enough" to handle such a situation.

Well, that got me riled. And when I'm riled I write. A few days after returning home from Buck's clinic, I sent him an eight-page letter detailing all the reasons I was convinced that a) he was already a "clicker trainer" and b) he could be a better one. I would probably not post it here if I had heard back from him, and I am hesitant to do it now, but I don't know whether he's still trashing the people with whom he should be making common cause, and I'd love to jump start the dialogue that might bring us closer to mutual understanding. As I think I make clear in the letter, I admire Buck a great deal, but I think in this instance he's using his influence to real potential harm. I also realized that this letter represented my own most focused attempt to articulate the power and promise of clicker/marker training. (I regret that my summary of its history contained a couple of significant inaccuracies. I have let them stand here in the interests of fair representation of my own fallibility, but apologies are due to the memory of Keller Breland.) Anyway, here it is:

November 4, 2011

Dear Buck,

First and foremost, I want to thank you. I attended one of your recent clinics in Spanaway as a first-time spectator. Even from that remove, I learned more than I could have hoped, and I left powerfully inspired to put that learning into practice. I should say that I am not a horsewoman in either the casual or the proper sense of that term. I came to your clinic because I have a passion for clear communication between individual creatures who may not be of the same species, a passion I have so far exercised primarily as a writer and as a teacher of humans and dogs. I’ve spent about fifteen years teaching the first (high school, college, and adult students), only about two teaching the second (that is to say, only two with focused intent and the least little bit of efficacy). I guessed that I could learn a great deal from you in spite of the gap in our immediate interests, and I did.

One of the things that impressed me most during the clinic (and contributed immeasurably to your credibility) was your frequent reference to the limits of your own knowledge, your insistence that you still have and will always have more to learn. On a few occasions you expressed your well-founded disgust for people who get ahead of themselves, people who speak in tones of false authority on subjects about which they know next to nothing. (In my experience, next to nothing is often more dangerous than nothing at all when it comes to degrees of ignorance.) I would not have taken you for such a person, when you have generally been so careful to build your authority on a solid foundation from the ground up. So I was sorely disappointed and more than a little angered by your casual and insulting dismissal of clicker training in response to a student question on the third afternoon of the clinic. You made it clear from your comments that you know next to nothing about it, and yet you felt entitled to use the authority you have earned in other ways to trash the devoted work of people who might otherwise be your natural allies. You know only a caricature of clicker training, only the crudest sketch, and that’s the picture that may now persist indelibly in the minds of some of your students because you momentarily and uncharacteristically abdicated your responsibility as a teacher to know whereof you speak.

Imagine that someone who’d seen the film The Horse Whisperer considered himself competent to judge your methods and principles, to get on his mike and tell an arena full of people, “Oh that Buck Brannaman, what a load of mumbo jumbo. If you want to whisper to your horse, you go right ahead, but if you actually want to get something done...” Hell, you probably don’t have to imagine it. I’d bet you’ve heard it many a time, and I’d bet it pissed you off every time. I’d further bet that you’d hate to expose yourself for the same kind of fool, so it pains me to be the one to tell you that your pants were on the ground the other afternoon. But I’m hoping that this is what we both might call a teachable moment. I hope I can teach you enough in a few pages about clicker training that the next time someone asks you a similar question you don’t get yourself caught in a cranial-anal inversion but maybe pause long enough to say, “You know, I need to learn more about that before I can really judge whether there might be something to it.”

The telegraphic leash

Keep a float in your line...

One of the Clicker Expo presentations that I found most interesting and valuable was given by Michele Pouliot on "The Right Touch." Michele is determined (hooray!) to reclaim the leash as a tool for training and communication rather than simply for management, and she's demonstrating more generally that there are ways to employ contact artfully, informatively, and positively. As I told her after her talk, I'd been quite literally feeling some of this stuff out for myself over the last six months or so, inspired in the main by Buck Brannaman's work with horses, by his emphasis on finding a "soft feel" and leaving a "float" in the rein, and by his further emphasis on the importance of developing sensitivity in the horse and the rider so as to make the rein a conduit of information in both directions. I thought there was no reason that a leash couldn't function similarly, and I'd found through trial and error that it very much could. (Of course, the idea that collar pressure -- like bit pressure -- can be communicative is hardly a new one, but the messages people have sent by leash have typically been blunt and unpleasant. The idea that light pressure might be converted from an aversive to a conditioned reinforcer is, I think, novel.)

Michele has been much less clumsy in her efforts, and she gave all of us at her talk a simple, clear, and efficient method for flipping our dogs' conception of pressure (and our own), from oppositional force to welcome invitation.** As she mentioned, there had been some trepidation on the part of the Expo organizers around her presentation of her process, given that it relies on negative reinforcement to get rolling, but I can say with absolute conviction that her method could have saved my dogs a great deal of annoyance if I'd been acquainted with it earlier. And even having muddled my way to a rough approximation of what she's doing with the leash, I am better able now to refine my techniques intelligently (and to expand them into similar work with hand to body contact). I can more easily move forward thanks not only to the clarity of her approach but also to the intellectual and moral affirmation I took from noting its overlap with my own nascent ideas. Out of respect for her care in presenting the specifics of her method, I'll wait to describe them here until I've had a chance to review her notes, but I think they should be disseminated widely, as I'm convinced that they have the potential to reduce the use of negative reinforcement significantly. As long as we use leashes primarily to contain rather than to communicate, and as long as we labor under the misconception that the signals we send each other across the line must necessarily be aversive, we miss a great opportunity to get in better touch with our dogs.

** It's probably no coincidence that Michele is a champion "freestyler," i.e., she dances (beautifully) with her dogs. Anyone who's done much partnered dancing can readily understand how this mode of training is analogous to "giving good weight," and can also guess how seamlessly it might integrate with other vital forms of kinesthetic awareness and communication.

Photo by George Grall.

I love you, dammit!!

Arktomorphism?

I just spent the weekend in the company of a few hundred trainers and a smattering of scientists at the annual west coast Clicker Expo, organized by Karen Pryor and her skilled cohorts. My exhaustion last night spoke to the quality of the program and the liveliness of the other attendees -- it's good to be reminded in a training context of how much energy the brain consumes when it's fully engaged!

There were many ideas and provocations I encountered in the ballrooms and hallways of the Doubletree Hotel that I want to return to, things I'll need to gnaw on for a long time before I can digest them. What looks most temptingly chewy this morning, however, is a question that was posed to me by a fellow trainer yesterday morning. I had volunteered with a dozen or so other KPA graduates to offer a little coaching to interested parties (two sessions of twenty minutes apiece), with donated proceeds going a local charity. We worked in pairs, and it was unfortunately toward the end of our first session that our "client," whose dog was not with her, described how he would often growl when he lay on his bed and she approached and pet him. What should she do?

I really regret that our next client was waiting and we weren't able to give the question the attention it deserves, because it's loaded. Personally, emotionally, and theoretically. I didn't get any further than remarking that the dog was telling her something that she'd be wise to respect, which might have been a fine response if I'd had time to elaborate it, but was surely too brusque given the circumstances. My partner did better, noting that the dog was a terrier, asking whether the dog followed her hand when she withdrew (yes), and suggesting that the dog might be experiencing a conflict of intent: to roughhouse or to cuddle? But we had to leave it at that.

On the theoretical side, this presents as a relatively straightforward matter of strategic reinforcement, and I hope the woman with the terrier found her way later that morning to Ken Ramirez's excellent lecture, wherein he explored the promise and perils of working with secondary reinforcers, those things (not always tangible, sometimes experiential) that accrue value only by their association to other things that satisfy an animal's strong intrinsic needs (i.e. primary reinforcers). Is gentle touch a primary reinforcer? Considered broadly, for slow-developing, social mammals, it does appear to satisfy an intrinsic need, especially early in life. (Harry Harlow's poor rhesus macaques demonstrated this most tragically and persuasively.) But touch is critical at that early stage in part because it is instructive: a mother's or other's tactile tenderness teaches us what kind of touch is safe, and when. Squirming, jostling littermates and human carers contribute significantly to that education in the case of most dogs. Physical intimacy is double-edged for all of us: it has the simultaneous potential to be terribly harmful or deeply rewarding. So each of us necessarily becomes a connoisseur of touch, highly idiosyncratic in our taste for different varieties of contact.

As Ken noted, in the practical life of a trainer or pet owner, the need to draw any distinction between primary and secondary reinforcers is not nearly so pressing as the question of whether something is reinforcing at all. The question for the woman with the terrier is not whether her dog has a primal desire for touch, but whether he wants to be touched by her, in that way, in that place, at that moment. His growling suggests that he does not. Which does not mean that her desire to touch her dog in such a way under such circumstances must remain forever frustrated, only that she needs to teach her dog to enjoy it. Or risk getting bit.

There are many people who see these (sometimes irresistible) urges to kiss, hug, and cuddle our pets as yet another dangerous form of anthropomorphism. This is true to the extent that our species-typical touch repertoires do not everywhere overlap, and we need to be attentive to the places where they typically diverge. But when we're talking about an individual human and an individual dog (or cat or monkey or whale or other human), knowledge of what is typical may not only be immaterial, it may also be distorting. There are quite a few of us humans who find hugs from most people in most contexts highly aversive. Some find them aversive from all people in all contexts. Can we be shaped to enjoy them? Most of us, probably. But the more often we get hugged when we do not want to be hugged, by people who just want to show us how much they love us, the less a hug will communicate that professed love, and the more likely we'll be to interpret it as invasive and aggressive. As someone who should really know better, I am sorry to say that I think I inflicted an unwanted hug on someone this weekend, and the sincerity of my affection had no bearing on the question of whether it was rewarding for the victim. I "anthropomorphized" her, insofar as I made the narcissistic assumption that my desire to hug her was mirrored by her desire to be hugged.

Animals do this to us, too, as we'd be wise to remember the next time we get leapt on, slobbered over, or humped. My husband has a pair of black running tights that we've taken to calling his "sexy pants," because they drive our boy Pazzo into an amorous frenzy. (Perhaps not coincidentally, Pete vaguely resembles Pazzo from the waist down when he wears them.) Pazzo is clearly sincere in his passion for Pete, but the very force of that passion makes him insensitive to the question of how best to express it.

Calm before assertive

I recently picked up Cesar's Rules, Cesar Millan's newest book (co-written by Melissa Jo Peltier), and dug in with interest and a little trepidation. Like many who shelter under the "positive trainer" label, I have strong misgivings about Millan's methods and influence, misgivings inspired partly by the handful of his shows that I've seen and the smattering of his articles that I've read, but in the main by the multitude of people I've encountered who claim him as an authority in their efforts to become "leaders of the pack." In Cesar's Rules, Millan laments that many of his critics lambaste him without taking the trouble to understand his teachings, but I think he should worry at least as much about those who lionize him without taking the trouble to understand his teachings, or more importantly, without mastering the knowledge and skills that would enable them to employ his methods with relative safety. And I remain deeply skeptical that certain of his methods can be used by anyone with absolute safety.

Watching The Dog Whisperer makes it clear even to this critical observer that a) Millan is a highly skilled communicator and listener, b) he has amassed a great deal of informal knowledge about dogs, and c) he genuinely desires the best for all the animals (human and otherwise) with whom he works. Artful editing may give a fairy tale glow to each abbreviated narrative, but I don't think it can mask Millan's essential character, and I have never sensed that his use of strong aversives was an expression either of sadism or of an ego run amok. I bought Cesar's Rules for much the same reason that he says was inspired to write it: I think it is vital that people who share similar core values and aims find a way to air their differences and to learn from them. Many of the book's chapters center on other trainers, including a couple of heroes from the positive training world, Bob Bailey and Ian Dunbar. Elsewhere, too, Millan goes out of his way to express respect, and at moments even deference, toward views that conflict with his own.

Beyond that, I have been impressed while reading Cesar's Rules by what may or may not be a new emphasis on patience. Not having read Millan's other books, I don't know whether they supply a similar corrective to the implicit promise his show makes that "calm and assertive" leadership will produce near-immediate results. But it is striking here in his anecdotes and instruction how often he stresses the importance of baby steps. He has generously (and to his own benefit, of course) highlighted and disseminated the wisdom of other trainers, but I've also found myself nodding along with many of his personal insights. His stated desire to honor each individual animal is at the heart of my own philosophy of training (again, whether we're talking about canine or human animals).

Which brings us back to the questions surrounding "assertive" leadership and the use of strong aversives (or "positive" punishment). To Millan's credit, he acknowledges and makes a good faith effort to hear and faithfully represent the criticisms leveled against this aspect of his teaching. As I noted above, I think Millan's message is dangerous primarily for the encouragement it gives to those who lack his skills and understanding. When such people "assert" themselves with their dogs, they often become abusive and/or teach their dogs to be similarly "assertive," and the results can be disastrous. Millan quotes Ian Dunbar on this risk: "I teach mostly noncontact techniques, and there's a good reason for that. Most human hands can't be trusted... It's one thing if you're an experienced animal handler like you or me, 'cause you know which animals you dare touch and how you can touch them. The training methods that I would prescribe have nothing to do with the way I would train a dog or you would train a dog. It has to do with the fact that this is a family and there's two children in it. They're not necessarily going to have the observational skills that we have, or the speed or the timing, and certainly not the dog savvy. But they still have to learn to live happily with their dog."

I think Dunbar lets Millan off the hook a little too easily here, as I don't imagine there exists any great gap between how either man teaches others to train and how he trains dogs himself. And that's really the point. In his defense of e-collars, Millan enlists Temple Grandin to support his point that a strong aversive may be the most effective way to interrupt the prey drive at its highest intensity, and indeed it may be the only efficient way to do it. I'm grudgingly open to the arguments he makes for their careful use in special circumstances where a dog's safety is at stake and other solutions elusive. But I think (I hope) that, once Cesar opens the door to exclusively positive modes of training, he'll have an increasingly difficult time resorting to (and justifying) the regular use of leash "pops" and other physical corrections. Once you recognize that they're unnecessary to pretty much any training (and rehabilitation!) task you might face, and once you see the enormous uptick in mutual trust that results from a strong commitment to a "no harm, no force" ethic, you may need time to shake your old habits, but the logic of love says you must.

If dogs wore shoes...

Really??

...we'd more easily walk a mile in them. But most don't, and those who do don't look happy about it. Our relationships with our pets are wonderfully peculiar, as the ties that bind us braid together intimacy and alienation. This is true to a degree of all relationships (between dogs, between people, and certainly between cats), but when we extend our interest and care beyond the bounds of our own species, we seem sometimes to find more direct access to each other's emotions than we ever enjoy with our nearer kin. At those very moments, however, we may also be struck by the other's unfathomable otherness.

I think we need to sustain and not to collapse the tension between these simultaneous truths -- "we understand each other perfectly" and "we don't understand each other at all" -- if we want to flourish together. More, I think we should celebrate it. In the history of our relations with other animals, and particularly in the history of our domestication of other animals (and their domestication of us!), views have tended to swing from one pole to the other, from the conviction that other animals exist only as extensions of human need (or of human fear, as in the case of the benighted wolf) to the conviction that they exist utterly apart from us. Wittgenstein's oft-quoted aphorism captures the latter belief nicely: "If a lion could speak, we couldn't understand him." Likewise, Thomas Nagel's famous (and to his mind impossible) question -- what is it like to be a bat? -- encourages an all-or-nothing judgment on the possibility of shared experience. But otherness is always radical, and subjective feelings of connection are always an objective illusion. (Or rather, the connection itself is illusory, though the feeling would probably show up on a brain scan.) We have no direct means of access to any other being's perception of the world, no matter the species, so unless we wish to retreat into lonely solipsism, we have to make do with indirect means and earnest approximations.

With that limitation in mind, we have good reason (founded on objective evidence) to suppose that, in many ways, our pets' and other animals' emotional and cognitive experience strongly resembles our own. The speaker whose talk I'm most excited to hear at the upcoming Clicker Expo is Jaak Panksepp, the neuroscientist who holds the Baily Endowed Chair of Animal Well-Being Science at Washington State University (wonderful that there should be a chair so endowed). He is one of a small (but happily increasing) number of scientists who dare to emphasize the obvious homologies (common structures with common origins) among diverse animal minds, especially among the minds of humans and other social mammals. (I use "minds" advisedly, as Panksepp is interested in subjective as well as scientific modes of inquiry and description.) He is also leading research into homologies that are not so obvious, teasing out the physiology and chemistry that underlie those brain processes that we hold in common (and variations that we do not). His book Affective Neuroscience is a marvel. Densely technical in places, it nonetheless serves both as an excellent overview of contemporary research into the dynamics of primary emotions (including their influence on cognition and learning) and as an eloquent, richly speculative description of the big questions that remain.

Marker training basics III

what next?

Strengthening the marker's power.

Once you've established an association for your dog (or other animal) between the sound of your chosen marker and some valued reward, you can begin to use the marker to identify and encourage any and all behavior you like. If you have a totally untrained young pup, you might mark and reward a moment's quiet in an outburst of barking, then the next moment, and the next, until you find that the moments accumulate into longer spells of sweet silence.* Moments of eye contact are another great place to start with puppies and many full-grown dogs: everything you do in training will be built on a foundation of focused attention, so make it as wide and deep as you can. Don't worry at first about attaching cues to these behaviors. Saying "quiet" before your dog has arrived at a solid physical understanding of what quiet is will only create needless frustration for you both. And saying "QUIET!" will probably convince him that whatever he's barking at is even more threatening or exciting than he thought, since you're suddenly barking too.

If your dog already performs one or more behaviors pretty reliably when you ask for them, you could begin simply by marking correct responses to your cue. What's correct? For now, whatever it has been in the past. If your dog habitually responds to "sit" by backing up a couple of steps and settling lazily onto one haunch while sticking his other back leg out to the side, you know that's his definition of the word. Yours might be different, but for the moment you can set aside the task of bringing the two definitions together. Mark and reward every sit that follows your "sit," no matter how slow, no matter how sloppy. (You may find the sits get quicker and straighter in spite of your absence of effort.)

A few things to remember:

1. Give the cue only once. If your dog fails to respond, wait at least twenty seconds (and until you have his full attention) before giving it again, or "Sit. Sit. Sit!" may become your cue. Treat all words like empty vessels, and fill them deliberately with meaning.**
2. The mark is always followed by a reward. You don't have to mark every repetition of the behavior (I'll talk later about effective "schedules" for marking), but when you do mark, you're making a promise on which you need to deliver.
3. Work on one new behavior at a time. There's a significant exception to this rule that I'll talk about later, but this helps avoid confusion for the animal and accelerates learning.
4. Work in short sessions. Very short! Ten to fifteen repetitions between breaks. At the first sign of fatigue or fading interest, stop.
5. End on a high note. If possible, end with the new behavior you've been training, but if necessary end with a behavior the dog already knows well. Success breeds success.

What's most important in the early going (and ever after) is that you and your animal enjoy yourselves. I won't go into an elaborate defense of positive training methods here, but they follow from the (scientifically sound) premise that the training of voluntary behaviors proceeds most effectively and predictably from a state of eager but contained anticipation (especially in the case of a predatory animal). But even when training's fun, it's taxing. Again, the moment you notice your dog's interest flagging, or your own impatience rising, stop -- always, if you can help it, with a fresh success, however small. That way you'll continually create positive associations (for you and your animal) with training itself.

*On the other hand, I do not recommend that you begin by marking and rewarding a bark, especially if you're using a more powerful marker like a clicker. One thing to keep in mind is that the first few behaviors you effectively marker train will become the animal's default behaviors in future training. Quietly attentive behaviors (like eye contact or sit) are your best choices at the start.

**The trick with words (and other cues and markers) as vessels of meaning is that they might already be topped up. Old meanings can be difficult to dislodge, especially if they're loaded with pain or fear. Thus it's a very good idea (though sometimes difficult in practice) to avoid saying your pet's name in anger. If you want a truly empty word, try something rare, silly, or foreign. I lived in Bologna more than twenty years ago, and pretty much the only time I get to knock the rust off my Italian these days is when I'm cueing Pazzo with "fusilli!" (left spin) or "bombolone!" (right spin).

Marker training basics II

Sounds like...

...steak!!

Making the marker meaningful.

The principles and methods I'm describing here work across species; they've been used effectively in the training of countless animals, from rabbits to rhinoceroses, from grizzlies to gerenuks. (I highly recommend that all skeptics check out this brief article by Karen Pryor and the embedded video of a marker-trained rhino.) Before her death last year, our lovely calico Hops had learned through marker training to target my hand with her nose and to give me a head butt on the cue "Zidane" (which will only make sense if you follow European football). Most of my experience, however, has been with dogs, and for ease of reference I'll focus my discussion on them.

As I noted in the previous post, it's important that the stimulus you choose as a marker should initially possess little or no meaning for the animal you want to train. This will give you full freedom to endow it with the meaning you want: "Well done! Good things coming!" Here's yet another reason that words can be problematic as markers -- unless you're training a young puppy and you can commit in a disciplined way to reserving your marker word(s) exclusively for training, there's a high likelihood that their meanings will become muddied with unintended associations.

I'm strongly sympathetic to the visceral distaste that many people feel for clickers and other mechanical soundmaking doo-hickeys. They're cold and fussy and seem to require a third hand that we haven't got. Still more annoying, they interpose a barrier of artifice between trainer and trainee. However, for consistency, precision, and repeatability, they're really hard to beat. And paradoxical as it may seem, the little gap they introduce in our "natural" communication with other animals actually improves our mutual understanding immeasurably. No, I take that back -- the improvement can be measured, has been measured, and it's big.

So I recommend a clicker despite the possibility that it may not be totally neutral for you or your dog. Contrary to popular belief, you won't need to use it forever: once you've trained a specific behavior and put it reliably on cue, word markers will usually suffice to maintain the training. But you'll get to that maintenance stage much more quickly with a clicker, and once you're there I can pretty much guarantee that you'll notice a significant uptick in your dog's enthusiasm and attention anytime you pick that silly gadget back up. Likewise your cat's -- Hops would immediately start purring whenever I started a new training session.

Of course, that wasn't her first response to the clicker. She was mostly indifferent to it, maybe a little affronted. The sound it makes is short, sharp, and attention grabbing, qualities that make it effective for training but also make it rude in other contexts. So you'll want to introduce it carefully, from a distance, or muffled in a pocket. (It's best to avoid pointing the clicker like a t.v. remote at your dog's face.) Gauge your dog's response. Curiosity and/or indifference are a fine place to start, but if you see him/her shrink back, you'll need to take things more slowly.

If you decide to use another marker, just do what you can to keep it precise and consistent. A single syllable like "yes" generally works better to this end than a longer word.

Your first and most important task is to persuade your dog that, from this moment on, the following equation holds absolutely true:

"click" (or "yes") = wonderfulness

In order to establish that equation, you need to know what your dog already considers wonderful. Ideally you can identify among the many things your dog loves some thing(s) that are easily doled out in small bits. Food is the obvious candidate for a predatory species, and the one I rely on most heavily (probably too heavily, made complacent by my dogs' eager appetites). For early training, when the strength of a good impression takes precedence over perfect nutrition, I like (because my dogs like) hot dogs and smoked mozzarella, which are relatively inexpensive, have great intensity of flavor, and are easily divvied into tiny (1/4" by 1/4") cubes. Red Barn and Natural Balance also make meaty but dry food rolls that are as mouthwatering (to most dogs) as they are nourishing; they're my treat of choice when I work with dogs at the Oregon Humane Society. But it's good and sometimes necessary to think beyond food rewards; depending on the animal and the situation, they may be impractical and/or unrewarding. I've been working with Pazzo recently on his ability to keep the leash loose when we walk through our local squirrel-infested park (where he may be totally indifferent to food that isn't on the move). When he pulls the leash taut, I simply stop. The moment he gives me slack, I click and move with him in the direction of the squirrel. To Pazzo's delight, the squirrels will often double-down on my reward by staying put, and we've thus become a great slow-motion stalking team. Likewise, agility trainers will often reward their dogs with quick games of tug, and trainers of impassioned herders carefully control their access to sheep.

But for clarity's sake I'll assume that you've got a clicker, a hungry dog, and a stash of small, tasty food treats. Here's what you need to do:

1. Click.
2. Treat.
3. Repeat.

That's it. There are only a couple of competing provisos: try not to move your "treat hand" until after you've clicked, but deliver the treat as quickly as possible (within a second of the click). Finding a rhythm that keeps those two events close but distinct will make the click most meaningful to your dog and help unlock his/her exclusive focus on the treats.

Marker training basics I

Worth a thousand words?

On the definition and choice of a marker.

Even (or especially) among experienced animal trainers and savvy pet owners, I often encounter a strong prejudice against "clicker training." The phrase itself is a turnoff to many, which is one reason I've come to prefer the more or less interchangeable terms "marker training" or "bridge training." These latter phrases describe the approach more accurately and inclusively -- people naturally get confused when you tell them that "clicker training" doesn't necessarily involve a clicker. Another reason I favor the second two terms is that they have so far escaped strong commercial association and appropriation.**

"Marker training" is the most literal and straightforward of the three, so it's the one I'll use from now on. My aim here and in future "basics" posts is to lay out the foundational principles, reasoning, and tools of this approach so that you'll be free to adapt them to your own needs and ends. I'll also try to anticipate some of the challenges you might encounter when starting out, and to offer possible solutions. But one of the foremost advantages of marker training is its flexibility: once you have a good command of the core ideas, you and the animal you're training have infinite creative license in putting them to use!

Marker training falls under the larger umbrella of positive training methods; indeed, it's something we all practice whenever we say "good dog!" But a solid understanding of how and why it works can help us practice it much more deliberately and effectively.

As with the "good dog!" example, a marker is simply a stimulus chosen by the trainer to signal two things and establish a vital connection between them:

1. I like that behavior.
2. You will be rewarded.

An effective marker satisfies a few important criteria:

1. It is specific.
2. It is easily reproduced by the trainer.
3. It is easily perceived by the trainee.
4. It is initially neutral, meaning that it has little or no intrinsic meaning to the trainee.

I don't know of any successful use of taste markers, given that few tastes are truly neutral to any animal; smells can be tricky for the same reason but are sometimes used as markers, most obviously and often in tracking work. Sight markers are tops in neutrality, but not always easily reproduced or perceived. Like touch markers, they may be most useful with animals who have lost use of one or more of their senses, or in situations where sound markers are impractical or forbidden. For most trainers and trainees, in most situations, sound markers tend to be most adaptable and workable. That said, they require more care in their choice and use than you might expect.

Why is that? Part of the problem ironically arises from our great facility in producing varied and complex sounds, our gift of the gab. The general human reliance on words to convey meaning makes many of us sloppy with tone (unless we speak a tonal language), volume, enunciation, inflection, and emphasis. In other words, we take the least care with precisely those variables that other animals are most likely to find intelligible. We toss flurries of meaningless syllables their way like so many snowballs -- and instead of congratulating them for catching a few on the fly, we berate them for being stubborn and slow. Our carelessness in expression is mirrored by our bluntness in perception. Few of us can reliably hear the difference between one "good dog!" and another (less enthusiastic, slower in tempo, higher in pitch, etc.) but a dog can. "Sit, Stormy. Sit! Get down, Stormy! No, Stormy! Sit!" may be roughly translated as: "My poor owner is working herself unnecessarily into a lather." The more loquacious we are, the more faith we place in language, the less likely it is that our pets will understand us.

So choosing and using sound markers effectively requires that we get humble; we need to begin from a recognition of our limitations. Most of us just don't possess the emotional and vocal control we need to produce sounds that are highly specific, consistent, and intrinsically neutral. Which is not to say that words cannot work as markers, only that their ease of use masks (and even contributes to) their inefficacy relative to other, more precise sound markers. Like clickers, yes, but also like whistles, chimes, and bells. Training by whoopee cushion, anyone?

**I am an ardent fan of Karen Pryor (trainer and educator extraordinaire, founder and CEO of Karen Pryor Clickertraining), forever grateful to her for her insight, her dedication to reality over "common knowledge" (e.g. dominance theory), and her tireless advocacy of positive training methods. The six months I spent under the instruction of Helix Fairweather with the Karen Pryor Academy for Animal Training and Behavior were tremendously illuminating and rewarding. I am proud to be certified by KPA as a trainer, and I plan to attend ClickerExpo here in Portland at the end of the month. I do, however, think there's a downside to Karen's mostly laudable efforts to establish common professional standards and gather like-minded trainers into one big tent, particularly when there are fees collected at many of the tent's entrances. I had a mostly friendly tussle with KPCT's president, Aaron Clayton, when I graduated from KPA and learned that my promised year of free access to the alumni message boards was contingent upon my entering a marketing agreement that would require me to display the KPA logo on my website and anywhere else I advertised my services as a trainer. In this instance and a few others, I found that independence of thought and self-presentation ran somewhat at odds with the commercial imperatives of KPCT (as they would with those of most for-profit ventures).

Befriending the unconscious mind II

In my last post I made a distinction between analytical and associative logic as one way of separating out the primary modes of thought favored by the conscious and unconscious mind. (Depending on your level of comfort ascribing "thought" to the unconscious mind, you might substitute "modes of response" in that sentence, but my own definition is pretty expansive.) However, the distinction between analysis and association is not absolute, and it gets particularly fuzzy when we contrast classical conditioning and operant conditioning. The question of whether the unconscious mind "analyzes" a given situation (and how effectively it does so relative to the conscious mind) here elbows its way to the fore.

A much simplified review: classical conditioning is Pavlov, and operant conditioning is Skinner. In the first case, an initially neutral (i.e. affectively meaningless) stimulus is paired closely with an "unconditioned" (i.e. intrinsically meaningful) stimulus often and consistently enough that it becomes meaningful even in isolation. Unless a dog is temperamentally nervous, she is unlikely to have any strong primary response to the sound of a bell. Unless she is sick, full, or finicky, however, she will almost invariably respond to the presence of food, by salivating, pricking her ears, widening her eyes, etc. As Pavlov discovered, if the sound of a bell is repeatedly paired with the arrival of food, it will soon provoke many of the same reflexive responses that food does, even in food's absence. (These responses will often extinguish if the association is not periodically maintained -- though threatening associations are more resilient than positive ones -- but there are interesting and somewhat counterintuitive laws governing the effectual timing of that maintenance. More on that another day.)

Operant conditioning involves willed (or, if you won't go so far, voluntary) behavior. In this case, some specific action by the animal repeatedly and consistently provokes a change in her environment; if that change is meaningful to the animal, she will alter her behavior accordingly. Thus the Skinner box: rat presses lever, food pellet arrives, rat presses lever again with same result, and rat soon becomes a lever-pressing fiend. A fat lever-pressing fiend. Time to add a complication, in the form of a "green" light: only when the light is on will food pellets be available at press of lever. When the light is off, the rat can press for all she's worth but press in vain. The rat soon stops pressing the lever in the absence of light.

At what point (if any) in that sequence does analysis enter in? At what point does a rat or dog or human begin to perceive "coincidence" (the predictable proximity of two previously unrelated things or events) as a relationship of cause and effect? And is that perception most potent (most behavior-altering) at a conscious or unconscious level?

Even in the case of the planet's Great Brains (i.e. humans), it appears that the unconscious gets there first and most decisively. "Gut feelings" whisper to the frontal cortex the conclusions that older, deeper structures have already drawn -- and in many cases already prompted our bodies to act upon. Gerd Gigerenzer has done some incisive research into this dynamic, and his book Gut Feelings: The Intelligence of the Unconscious is one of the best introductions I've found. I'd also highly recommend Timothy Wilson's Strangers to Ourselves: Discovering the Adaptive Unconscious, which beautifully assimilates contemporary research with earlier descriptions of the relationship between conscious and unconscious thought. But among books written for the lay reader, Antonio Damasio's Descartes' Error: Emotions, Reason, and the Human Brain remains the most coherent (if necessarily speculative) revision of the "top-down" model of human decision making that I've read, and it's a good place to go if you're ready to dig into the physiological bases of cognition (insofar as these are intelligible to us, which isn't yet very far). His "somatic marker" hypothesis turns the idea that the consciously reasoning mind is in command of the lowly body pretty much literally on its head.

Image by SubVerse Clothing

Befriending the unconscious mind I

What does it mean in practice to treat the unconscious mind with greater respect? It means setting aside many of the strategies that the conscious mind tends to favor (e.g. reasoning, browbeating, and harassment) when it bumps up against inconvenient and recalcitrant desires. It also means setting aside most physical forms of coercion, while embracing strategies of containment. Most crucially, it means forging an alliance with "animal" vitality, whether your own or your dog's. Hunger (in its broadest sense) is the mainspring of life -- if you can harness its power, you'll flourish and so will your dog.

The unconscious mind works by associative rather than analytic logic; it constructs links between things and events that are spatially or temporally close. The more often two things coincide or appear in proximity (one right next to or one right after the other), the stronger the link between them generally becomes. However, as will become important to a later discussion of punishment, there are circumstances that can exaggerate the strength of a link even if it is made only once.

The unconscious mind has stronger ties to the past than to the future, but its first allegiance is to the present. It's very difficult to fob it off with promises, no matter how sincere. "I will gladly pay you Tuesday for a little self-control today" tends not to be persuasive when there's a juicy hamburger (or whatever floats your Pavlovian boat) sitting there under your nose. If, however, you have managed to make self-control itself intrinsically rewarding -- if you have associated it often enough and in a variety of circumstances with strong and immediate payoffs -- you have some leverage. There are marker-trained dogs who will fetch whole hot dogs and deliver them unmolested to their owners' hands... in return for a 1/4-inch cube of hot dog. This takes some work.

Image by Marc Greisinger.

Force of habit

At its most effective, training is primarily about the creation of new habits, habits of association and habits of behavior. In some cases, we're trying to create a small piece of order where disorder currently reigns; in others, we're unhappy with established order and would like to build something more pleasing in its place. The second instance is the "old dogs, new tricks" challenge (though it could apply to a five-month-old puppy or a five-year-old child). For obvious reasons, this invariably requires more time and focused attention than if we start from a state of relative innocence and pliability. (I'm setting aside for now the question of native limits, only looking to discern the behavioral laws that govern the "free ground" where we can play and change.) If ingrained habits were easily dismantled, they wouldn't have so much potential to support us.

Habit enlists the power of inertia, for good or ill. However much the fact may offend our vanity as reasoning creatures, the vast majority of our habits "are created" without our conscious intent. As for humans, so for other animals. The unconscious mind is tenacious and indefatigable (sleep is for sissies!) in its quest to make just enough comparative sense of incoming stimuli to determine how we should act in order to get what we want and avoid what we don't. (How is this like or unlike a situation I've seen five or five thousand times before?) What it lacks in nuance it more than makes up in speed and confidence. If the unconscious mind has lit upon a strategy it really likes -- and likes more every time it repeats it, familiarity in this case breeding affection -- the conscious mind is generally left to mop up after the fact. (Oh, I totally meant to do that, and here are twenty reasons why...) Or to boast about its superior refinement and sophistication. Indeed, in the case of humans, the conscious mind sometimes seems as tireless in the task of self-glorification as the unconscious mind is in the humbler but more critical task of self-maintenance.

We might have a great deal more success in creating new habits and dismantling old ones if we had more respect for the unconscious mind, if we treated it with the courtesy and forbearance that elders should always command from the young. That upstart frontal cortex flatters itself that it knows what's best, but it can't get anything done on its own. It needs the collaboration of older and more resilient structures. It can only lead -- if it leads at all -- by encouraging consensus. Consciously adopted habits are one form that consensus can take.

What it is

In the nine months -- count 'em! -- since I last posted here, my attention has been absorbed by other projects, including the drafting of a memoir that's now more or less at rest. I'm ready to recommit myself to the questions I've been using this blog to explore, but I now think they might best be split. I'd like to establish a clearer focus here on training, and dig only as far into the science and philosophy of cognition as seems immediately useful. I'm putting together a second blog that I'll use as an arena for broader exploration of the overlap between human and non-human minds, also for wild tangents and miscellany.

Forty posts in, I'm finally able to state the premise of this blog: animal training requires self-training. No matter the species I'm working with, if I want to communicate clearly with an animal and persuade him to ally his will with my own, I will need to become more self-aware and self-controlled, more skilled in the signals I send and more attuned to those I receive. By the same token, I can only ask as much of another animal as I'm willing to put in myself. So "as good as I wanna be" makes reference to all the ways that my desire to improve my own behavior might be constrained: laziness, fear, conflicting desires, sheer cussedness. There might be other constraints on my ability to improve my behavior, some of them absolute, others elastic. But I won't discover those constraints except by testing and maybe redrawing the limits of my desire. Just as I won't discover the outer limits of Barley's or Kili's or Pazzo's abilities except by expanding my knowledge of what drives each of them and inventing new ways to channel it productively (i.e., in mutually agreeable directions).

So I start from the assumption that we're all only and always as good as we wanna be. I want to see what's possible from there.

In scientific circles

I think the mirror test demonstrates, at the point where its popularity as a measure of self-consciousness intersects with its inadequacy, the tendency of scientific investigation to wander into tautology when it treats the phenomena of sentience. It requires great care and imagination to conceive an experiment that will yield some verifiable external measure of an internal process, and when someone succeeds as elegantly as the originator of the mirror test, there's a strong temptation among those who credit the significance of the results to "move forward," to avoid any needless backtracking (e.g. to the definitional boundaries of the phenomenon under scrutiny).
     Thus the question of whether an animal possesses self-awareness elides irresistibly with the question of whether he can, with the help of a well-placed polka dot, make a connection between his kinesthetic or proprioceptive sense and an alien image that (bizarrely) coordinates with it; in the absence of any similarly compelling measure, the mirror test becomes definitive for hundreds of scientists who go on to paint scores of unsuspecting animals in their sleep. Will a parrot pass or fail? A tamarin? A zebra? As Frans de Waal observes, "for better or worse, this test has remained the gold standard of self-identity."
     Even the test's critics seem to accept its foundational terms: if, they say, an orangutan who touches a spot on his forehead really understood the image in the mirror as a representation of his own body, then the conditions for self-awareness would be met. But, they argue, he probably just likes to poke at his face. Or he learns to do so because it makes humans grimace in that weird way that means more dates and sunflower seeds.
     Again, methodological limitations lead us to chase our tails: the mirror test measures the capacity for self-consciousness because... we don't have a better test. Or a more complete one. Hell, I don't know what it means to be self-conscious! Do you?
     My feeling, one I'd like to develop into a well-reasoned conviction (so goes the trajectory of my mental life), is that there ought to be a kind of intellectual affirmative action in the direction of granting non-human animals manifold intelligence and complex consciousness. We ought to assume they're endowed with great riches of thought and feeling until they prove otherwise, though we ought not to assume that their thoughts and feelings trace the same patterns as ours. I have some sympathy for Marc Hauser, despite his faults and all the damage he's done to the cause of anthropomorphology, because a bias in favor of non-human intelligence remains so rare, while the bias against almost defines "respectable" research.

Self-consciousness without mirrors

There's a question I want to dig into a little further before I arrange another rendezvous between Hamlet and Burrhus (Frederic Skinner), a question regarding self-consciousness. This is one of innumerable capacities ascribed until recently only to humans. Various experiments with mirrors and paint have widened the circle of self-conscious creatures just a little bit, to include apes and magpies (!) among a few others, but I think the assumption that underlies the research may be too restrictive to allow a full description of the phenomenon. While it may be difficult or impossible to demonstrate under scientific controls (this is clearly a case in which observation itself may distort the nature and behavior of what we observe), informal study argues strongly for the emergence of "personality" in many social species who (pronoun used advisedly) fail the mirror test. Canis familiaris, to take one salient example.
     What if one accumulates (even if unwittingly) a distinct and precious identity, an identity one is motivated to defend (even if reflexively)? Mightn't this constitute a kind of self-consciousness, whether or not the self is pinched off from consciousness and set out as an object for one's contemplation and deliberate manipulation? I think anyone who has ever observed the wounding of a dog's pride or a cat's dignity must admit the possibility.
     The counterexample of the octopus also supports a more expansive definition of self-consciousness. Experiments performed using HDTV suggest that, however intelligent, an octopus has no personality: that is, it demonstrates the patterns of behavior that we generally attribute to personality, but these patterns are extremely short-lived. An octopus that is extroverted and aggressive one day may be terribly timid the next. (Wonderful that the subject of this research was Octopus tetricus: vulgarly, the "gloomy" octopus.)
     If the range of an animal's behavior (and the probability of any specific response to a stimulus) were determined simply by a passive stockpiling of experience and not by any active sense of internal coherence - of individual integrity - one would not expect to see such wild variations in the robustness of behavioral patterns among species.
     **There's another experiment, performed back in 2008, that hints at a canine capacity for self-consciousness. Austrian researchers trained a pair of border collies to sit and shake on cue, then measured the time it took for the behaviors to extinguish when they received no reinforcement. The salient data came from a comparison between "control" trials, wherein one of the dogs worked alone, and trials wherein the two dogs worked side by side but only one received reinforcement. Behaviors extinguished significantly more quickly in the second case (and the unrewarded dog showed many more visible signs of frustration). Discussion of the research has focused primarily on the question of whether this demonstrates that dogs have a sense of "fairness," but it certainly suggests that they have a vigorous sense of "me" distinct from "him," a protective self-regard that might amount to a form of ego.