Search
  • Canine Consultant - Emma

Positive Reinforcement - The Oldest 'Modern Fad' Ever?


It's pretty common that those who haven't been aware of positive reinforcement or training that avoids the use of aversives, who perhaps are defending what they do or what they used to do, will claim that positive reinforcement based training* is a 'new fad'... To be clear, I define what I do as 'positive reinforcement based' or 'avoiding aversives' or 'being both kind and science based'.. a whole bunch of stuff. Each time someone comes up with a snappy name to try to cover what we do/don't do of course, like 'force free' or 'LIMA' (Least Invasive Minimally Aversive), someone else will either try to twist it's meaning or pick it apart. For example 'oh, you are force free are you, so that means if your dog was running towards a road, you wouldn't grab his collar or pull his lead tight'... or a million and one other dire emergency situations that are patently NOT training related, or nitpicking 'well using a harness can be aversive, even food could be aversive if the dog feels sick that day'... yadda yadda - it gets boring folk, it also makes you look defensive as hell, but I digress.

Back to the point - there is an idea, typically among those new to training, those who have previously trained using aversives, those who want to carry on training using aversives, that positive reinforcement is somehow new. I'd like to counter that idea with a touch of evidence. Not only is the idea of using food rewards, the opportunity to play, the opportunity to perform inherently reinforcing behaviours, physical contact and fuss NOT new, the idea that punishments might be damaging and carry risks is also NOT new! (Keep in mind of course, these authors had very little idea of the science behind the methods they used and some of the methods they used were indeed aversive, and potentially very damaging - we know better now, there is no excuse not to do better, but they genuinely did not know.) Have at these exerpts, I've been trundling through Project Gutenberg which is awesome... https://www.gutenberg.org/ The following are all from: Dog Breaking The Most Expeditious, Certain, and Easy Method, Whether Great Excellence or Only Mediocrity Be Required, With Odds and Ends for Those Who Love the Dog and Gun

MAJOR-GENERAL W. N. HUTCHINSON,

LATE COLONEL GRENADIER GUARDS. First published 1847. On the qualities needed of a trainer: "REQUISITES IN AN INSTRUCTOR.

6. The chief requisites in a breaker are:—Firstly, command of temper, that he may never be betrayed into giving one unnecessary blow, for, with dogs as with horses, no work is so well done as that which is done cheerfully; secondly, consistency, that in the exhilaration [4] of his spirits, or in his eagerness to secure a bird, he may not permit a fault to pass unreproved (I do not say unpunished) which at a less exciting moment he would have noticed—and that, on the other hand, he may not correct a dog the more harshly, because the shot has been missed, or the game lost; and lastly, the exercise of a little reflection, to enable him to judge what meaning an unreasoning animal is likely to attach to every word and sign, nay to every look."

On punishment (the initiatory lessons the author mentions here means teaching the dogs what the commands mean via the use of food and play, in a quiet non-distracting environment! How very modern..)

"134. Common sense shows that you ought not to correct your dog for disobedience, unless you are certain that he knows his fault. Now you will see that the initiatory lessons I recommend, must give him that knowledge, for they explain to him the meaning of almost all the signs and words of command you will have to employ when shooting. That knowledge, too, is imparted by a system of rewards, not punishments. Your object is not to break his spirit, but his self-will. With his obedience you gain his affection."


On spiked (prong) collars:

300. I have made no mention of the spiked-collar, because it is a brutal instrument, which none but the most ignorant or unthinking would employ. It is a leather collar into which nails, much longer than the thickness of the collar have been driven, with their points projecting inwards. The French spike-collar is nearly as severe. It is formed of a series of wooden balls,—larger than marbles,—linked (about two and a half inches apart) into a chain by stiff wires bent into the form of hooks. The sharp pointed hooks punish cruelly when the checkcord is jerked.

On the risks of poor punishment or injudicious punishment:

325. If a man cannot readily get hold of any dog under his tuition whom he desires to rate or punish, you may be certain that he fails either in temper or judgment; perhaps in both. He may be an excellent man, but he cannot be a good dog-breaker. There are men who get quite enraged at a dog’s not coming instantly to “heel” on being called. When at length the poor brute does come within reach, he gets a blow, perhaps a licking—a blow or licking, he has the sense to see he should have longer avoided had he stayed longer away. Thus the punishment increases instead of remedying the evil.

On the 'chastisement' of dogs: 347. Obedience and intelligence are, as I have already remarked, best secured by judicious ratings and encouragements,—scoldings for bad conduct,—praise, caresses, and rewards for good. Never forget, therefore, to have some delicacy in your pocket to give the youngster whenever he may deserve it.

To be absolutely crystal clear here, this military man did use punishment and he did use aversives. He shows a remarkable awareness of the harm that can be done using what we would now call positive punishment, awareness of the timing required when delivering a punishment, awareness that punishment can damage the relationship between man and dog - and remember, W N Hutchinson is writing this some 50+ years before Pavlov won his Nobel Prize, some 100+ years before B F Skinner conducted HIS experiments that taught us much about learning theory. Throughout the book he repeatedly discusses the use of food rewards, of using scent to engage the dogs brain, of using toys (gloves, rabbit skin toys, dummies) to play at training before the dog is exposed to the real thing. He discusses setting the dog up to succeed by training at first indoors, away from distractions, by teaching the dog what the commands mean before ever expecting the dog to understand, he cautions against the futility of shouting commands the dog does not know, and then punishing the dog for not knowing. Incredibly, he describes desensitizing the dog to gun-shot and to water, using food to counter-condition in a way that looks every bit like a modern, positive reinforcement trainer would.


Maybe this is just a one off... I looked at: HANEY’S ART OF TRAINING ANIMALS. A PRACTICAL GUIDE FOR AMATEUR OR PROFESSIONAL TRAINERS. GIVING FULL INSTRUCTIONS FOR Breaking, Taming and Teaching all kinds of Animals INCLUDING AN IMPROVED METHOD OF HORSE BREAKING, MANAGEMENT OF FARM ANIMALS, TRAINING OF SPORTING DOGS; SERPENT CHARMING, CARE AND TUITION OF TALKING, SINGING AND PERFORMING BIRDS; AND DETAILED INSTRUCTIONS FOR TEACHING ALL CIRCUS TRICKS, AND MANY OTHER WONDERFUL FEATS. (Jesse Haney & Co, Publishers, New York, 1869)


(Titles were really NOT snappy in those days!)....


TO TEACH HIM HIS NAME.

If a dog has any particular name by which he is usually addressed, he will in time learn to answer to it. With a little system, however, he will learn much sooner than otherwise, and where there are several dogs it is a good plan to make each know his own distinctive title thoroughly, and to respond promptly to it. This will render your intercourse with, and your management of them, both easier and pleasanter. It may be accomplished by a very simple process. When you feed them, call each one by name to his food. If any of the others come forward, send them back. By dividing the food into small morsels and calling each dog in turn to receive his piece, always insisting that he and no other shall receive it, considerable of a lesson may be derived from each meal. When convenient take 80them out to walk, being careful to provide yourself with a few crackers or a piece of bread. Allow the dogs to ramble about at their pleasure, and whenever you choose call some particular one by his name; when he comes to you reward him with a piece of cracker. By-and-by call some other one, and continue the plan at your discretion. At the end of ten or a dozen lessons they will have pretty well learned their names, and come at your call. If you have only one dog, the plan would be about the same.


Haney (or whoever Haney actually had write this) describes the training of many tricks by many animals and whilst quite a lot of these did involve force, we also see a lot of food lure and food rewards used to get the behaviours required and to capture them.

This understanding of positive reinforcement, desensitization and the harm that aversive experiences could do goes back a heck of a lot further than the 1800's however. Xenophon the Athenian soldier and historian was writing in 355 BC ... Here's Xenophon (translated) on Horsemanship: "[14] Moreover, when the horse is shy of anything and will not come near it, you should teach him that there is nothing to be afraid of, either with the help of a plucky horse—which is the surest way—or else by touching the object that looks alarming yourself, and gently leading the horse up to it.

[15] To force him with blows only increases his terror; for when horses feel pain in such a predicament, they think that this too is caused by the thing at which they shy. "


Just for a bit of a giggle, Xenophon not only cautions on the use of comfortable equipment, warns against the use of punishment and advocates for positive reinforcement with both hounds and horses.. he advises on the naming of hounds. So here is Xenophons list, that you may ensure your hounds are named correctly, with short sounds that may not be mistaken for other commands. "Give the hounds short names, so as to be able to call to them easily. The following are the right sort: Psyche, Thymus, Porpax, Styrax, Lonchê, Lochus, Phrura, Phylax, Taxis, Xiphon, Phonax, Phlegon, Alcê, Teuchon, Hyleus, Medas, Porthon, Sperchon, Orgê, Bremon, Hybris, Thallon, Rhomê, Antheus, Hebe, Getheus, Chara, Leusson, Augo, Polys, Bia, Stichon, Spudê, Bryas, Oenas, Sterrus, Craugê, Caenon, Tyrbas, Sthenon, Aether, Actis, Aechmê, Noes, Gnomê, Stibon, Hormê.1"

[Xenophon. Xenophon in Seven Volumes, 7. E. C. Marchant, G. W. Bowersock, tr. Constitution of the Athenians. Harvard University Press, Cambridge, MA; William Heinemann, Ltd., London. 1925.]

[1 The names are significant of the colour, strength, spirit, sagacity or behaviour of the hounds. Hebe and Psyche are still in the list of bitches' names, and modern equivalents of several of the other names are in use, e.g., Lance (Lonche/), Sentinel (Phylax), Ecstasy (Chara), Blueskin (Oenas), Crafty (Medas), Hasty (Sperchon), Vigorous (Thallon), Impetus (Horme/), Counsellor (Noes), Bustler (dog) or Hasty (bitch); cf. Sperchon. For Πολύς we should probably read Πολεύς, “Rover.”

Xenophon. Xenophon in Seven Volumes, 7. E. C. Marchant, G. W. Bowersock, tr. Constitution of the Athenians. Harvard University Press, Cambridge, MA; William Heinemann, Ltd., London. 1925. ]



I'd like to think that today, Xenophon, whoever Haney had writing for him, the good Major-General W. N. Hutchinson and all their like minded fellows, would delight in understanding just how dogs learn, and how the harsh methods they did use were simply not necessary, and in fact were hampering what they did rather than improving it. But for their times, these people did a remarkable job - and now you know, none of this dog training lark is new, and certainly not a modern fad!


73 views0 comments

Recent Posts

See All