Dog Training—Let Reason Prevail Over Force!

Roger Abrantes and Boxer doing retrieve

“Whether you (or I) follow a particular line of morality is not a necessary consequence of any model of social behavior. Moral stances are solely your (or my) decision” (Picture by Lisa J. Bains).

The dog trainers’ dispute about training methods blazes on unabated, with the erroneous and emotive use of terms such as dominance, punishment and leadership only adding fuel to the fire. There is no rational argumentation between the two main factions, one of which advocates a “naturalistic” approach and the other a “moralistic” stance. The term ‘dominance’ generates particular controversy and is often misinterpreted. We can detect, in the line of arguing about this topic, the same fundamental mistakes committed in many other discussions. By taking the controversy over dominant behavior as my example, I shall attempt to put an end to the feud by proving that neither side is right and by presenting a solution to the problem. Plus ratio quam vis—let reason prevail over force!

I shall demonstrate that the dispute is caused by:

(1) Blurring the boundaries between science and ethics. While ethics and morality deal with normative statements, science deals with factual, descriptive statements. Scientific statements are not morally right or wrong, good or bad.

(2) Unclear definitions. We cannot have a rational discussion without clear definitions of the terms used. Both sides in the dispute use unclear, incomplete definitions or none at all.

(3) Logical fallacies. The opposing sides commit either the ‘naturalistic fallacy,’ ‘the moralistic fallacy,’ or both. We cannot glean normative statements from descriptive premises, nor can we deduce facts from norms. The fact that something is does not imply that it ought to be; conversely, just because something ought to be does not mean that it is.

(4) Social conditioning and emotional load. As a result of inevitable social conditioning and emotional load, some terms develop connotations that can affect whether we like or dislike, accept or reject them, independent of their true meaning.

(5) Unclear grammar. The term dominance (an abstract noun) leads us to believe it is a characteristic of certain individuals, not an attribute of behavior. The correct use of the term in the behavioral sciences is as an adjective to describe a behavior, hence dominant behavior.

Bottom line: We need to define terms clearly and use them consistently; otherwise rational discussion is not possible. We must separate descriptive and normative statements, as we cannot derive what is from what ought to be or vice versa. Therefore, we cannot use the scientific concept of dominant behavior (or any descriptive statement) to validate an ethical principle. Our morality, what we think is right or wrong, is a personal choice; what is, or is not, is independent of our beliefs and wishes (we don’t have a choice).

Solution to the problem: The present dispute focuses on whether we believe it is right or wrong to dominate others (as in, totally control, have mastery over, command). It is a discussion of how to achieve a particular goal, about means and ends. It is a moral dispute, not a scientific quest. If both sides have similar goals in training their dogs, one way of settling the dispute is to prove that one strategy is more efficient than the other. If they are equally efficient, the dispute concerns the acceptability of the means. However, if either side has different goals, it is impossible to compare strategies.

My own solution of the problem: I cannot argue with people who believe it is right to dominate others (including non-human animals) as, even though I can illustrate how dominating others does not lead to harmony, I can’t make anyone choose harmony or define it in a particular way. I cannot argue with people who think it acceptable to hurt others in order to achieve their goals because such means are inadmissible to me. I cannot argue with people who deny or affirm a particular matter of fact as a means of justifying their moral conduct, because my mind rejects invalid, unsound arguments. With time, the rational principles that govern my mind and the moral principles that regulate my conduct may prove to be the fittest. Meanwhile, as a result of genetic pre-programming, social conditioning and evolutionary biology, I do enjoy being kind to other animals, respecting them for what they are and interacting with them on equal terms; I don’t believe it is right to subjugate them to my will, to command them, to change them; and I don’t need a rational justification as to why that’s right for me*.

Roger Abrantes and Bulldog

“I do enjoy being kind to other animals, respecting them for what they are and interacting with them on equal premises; I don’t find it right to subjugate them to my will and dispositions, to command them, to change them; and I don’t need a rational justification for why that’s right for me” (Picture by Lisa J. Bain).


1 Science and ethics are not the same

Science is a collection of coherent, useful and probable predictions. All science is reductionist and visionary in a sense, but that does not mean that all reductionism is equally useful or that all visions are equally valuable or that one far-out idea is as acceptable as any other. Greedy reductionism is bound to fail because it attempts to explain too much with too little, classifying processes too crudely, overlooking relevant detail and missing pertinent evidence. Science sets up rational, reasonable, credible, useful and usable explanations based on empirical evidence, which is not connected per se. Any connections are made via our scientific models, ultimately allowing us to make reliable and educated predictions. A scientist needs to have an imaginative mind in order to think the unthinkable, discover the unknown and formulate initially far-fetched but verifiable hypotheses that may provide new and unique insights; as Kierkegaard writes, “This, then, is the ultimate paradox of thought: to want to discover something that thought itself cannot think.”

There are five legitimate criteria when evaluating a scientific theory or model: (1) evidence, (2) logic, (3) compatibility, (4) progression, and (5) flexibility.

(1)  Evidence: a scientific theory or model must be based on credible and objective evidence. If there is credible evidence against it, we dismiss it. It must be testable and falsifiable.

(2)  Logic: If a theory or model is based on logically invalid arguments or its conclusion are logically unsound, e.g. drawing valid conclusions from false premises, we must also dismiss it.

(3)  Compatibility:  If a theory or model shows crucial incompatibility with the whole body of science, then it is probably incorrect. If it is incompatible with another model, then we have a paradox. Paradoxes are not to be discarded, instead worked on and solved (or not solved as the case may be). Since “Paradoxes do not exist in reality, only in our current models of reality,  […] they point the way to flaws in our current models. They therefore also point the way to further research to improve those models, fix errors, or fill in missing pieces.” In short, “scientists love paradoxes,” in the words of Novella.

(4)  Progression: A scientific theory or model must explain everything that has already been explained by earlier theories, whilst adding new information, or explaining it in simpler terms.

(5)  Flexibility: A theory or model must be able to accept new data and be corrected. If it doesn’t, then it is a dogma, not a scientific theory. A dogma is a belief accepted by a group as incontrovertibly true.

Science provides facts and uncovers important relationships between these facts. Science does not tell us how we should behave or what we ought to do. Science is descriptive, not normative. In other words: we decide what is right or wrong, good or bad, not necessarily depending on what science tells us.

Morality and science are two separate disciplines. I may not like the conclusions and implications of some scientific studies, I may even find their application immoral; yet, my job as a scientist is to report my findings objectively. Reporting facts does not oblige me to adopt any particular moral stance. The way I feel about a fact is not constrained by what science tells me. I may be influenced by it but, ultimately, my moral decision is independent of scientific fact. Science tells me men and women are biologically different in some aspects, but it does not tell me whether or not they should be treated equally in the eyes of the law. Science tells me that evolution is based on the algorithm “the survival of the fittest,” not whether or not I should help those that find it difficult to fit into their environment. Science informs me of the pros and cons of eating animal products, but it does not tell me whether it is right or wrong to be a vegetarian.

Ethologists study behavior on a biological and evolutionary basis, define the terms they use, find causal relationships, construct models for the understanding of behavior and report their findings. Ethologists are not concerned with morality. They simply inform us that the function of x behavior is y. They don’t tell us that because animal x does y, then y is right or wrong, good or bad, or that we ought or ought not do y.

The model I present in “Dominance—Making Sense of the Nonsense” is a scientific model that complies with all five of the requirements listed above.

(1)  It is based on overwhelming data, i.e. given my definition of ‘dominant behavior,’ one cannot argue that it does not exist.

(2)  The conclusions are logically consistent with the premises.

(3)  It is consistent with our body of knowledge, particularly in the fields of biology and evolutionary theory.

(4)  It explains what has been explained before and in more carefully defined terms.

(5)  It accepts new data, adjustments and corrections (the current version is an updated version of my original from 1986). The model tells us nothing about morality. No single passage suggests that we should classify any particular relationship with our dogs as morally right or wrong. You will have to decide that for yourself. As an ethologist, I’m not concerned with what ought to be, only with what is. Echoing Satoshi Kanazawa, if I conclude something that is not supported by evidence, I commit a logical fallacy, which I must correct, and that’s my problem, but if my conclusion offends your beliefs, then that’s your problem.

Therefore, whether you (or I) follow a particular line of morality is not a consequence of any model of social behavior. Moral stances are solely your (or my) decision. It is not correct to draw normative judgments from descriptive claims. If you do so, you either commit the ‘naturalistic fallacy,’ the ‘moralistic fallacy’ or both, as I shall explain below (see point 3).

2 Unclear definitions

Having just pointed out the rigors of science, I must concede that the scientific community does bear some responsibility for the present dispute in as much as definitions and use of terms have sometimes been sloppy. Some researchers use particular terms (in this case ‘dominance’) without defining them properly and with slightly different implications from paper to paper.

Wikipedia writes: “Dominance (ethology) can be defined as an ‘individual’s preferential access to resources over another’ (Bland 2002). Dominance in the context of biology and anthropology is the state of having high social status relative to one or more other individuals, who react submissively to dominant individuals. This enables the dominant individual to obtain access to resources such as food or access to potential mates, at the expense of the submissive individual, without active aggression. The opposite of dominance is submissiveness. […] In animal societies, dominance is typically variable across time, […] across space […] or across resources. Even with these factors held constant, perfect dominant hierarchies are rarely found in groups of any size” (Rowell 1974 and Lorenz 1963).

It explains a dominance hierarchy as follows: “Individuals with greater hierarchical status tend to displace those ranked lower from access to space, to food and to mating opportunities. […] These hierarchies are not fixed and depend on any number of changing factors, among them are age, gender, body size, intelligence, and aggressiveness.”

Firstly, defining ‘dominance’ instead of ‘dominant behavior’ seems somewhat imprudent for a science that is intrinsically based on observational facts. It suggests we are dealing with an abstract quality when in fact we are referring to observable behavior (see point 5 below). Secondly, it implicitly equates ‘dominance’ with hierarchy (social status), which is misleading because some hierarchies may be supported by conditions other than dominant behavior. The use of the term ‘dominance hierarchy’ creates a false belief. Clearly, the terms dominance and dominant behavior are attributed with varying meanings, a highly unadvisable practice, particularly in stringently scientific matters.

As John Locke wrote in 1690 (An Essay Concerning Human Understanding),  “The multiplication and obstinacy of disputes, which have so laid waste the intellectual world, is owing to nothing more than to this ill use of words. For though it be generally believed that there is great diversity of opinions in the volumes and variety of controversies the world is distracted with; yet the most I can find that the contending learned men of different parties do, in their arguings one with another, is, that they speak different languages. ”This has contributed […] to perplex the signification of words, more than to discover the knowledge and truth of things.”

To remedy this, I propose in “Dominance—Making Sense of the Nonsense” a set of carefully constructed definitions that are compatible with behavioral science and evolutionary theory, whilst paying special attention to the logical validity and consistency of the arguments. I’m convinced that we would avoid many pointless disputes if all those dealing with the behavioral sciences were to adopt such definitions.

Roughly speaking, there are currently two main schools of thought in dog training. For our present purpose, we shall call them ‘Naturalistic Dog Training’ and ‘Moralistic Dog Training.’ Of course, there are various other stances in between these two extremes, including a significantly large group of bewildered dog owners who do not adhere to any particular ideology, not knowing which way to turn.

Naturalistic Dog Training (aka the old school) claims their training echoes the dog’s natural behavior. They don’t provide a proper definition of dominance, but use it with connotations of ‘leader,’ ‘boss,’ ‘rank,’ implying that dominance is a characteristic of an individual, not of a behavior. In their eyes, some dogs are born dominant, others submissive, but all dogs need to be dominated because their very nature is to dominate or be dominated. They use this belief to justify their training methods that often involve punishment, flooding, coercion, and even shock collars, if deemed necessary, by the more extreme factions. For them, a social hierarchy is based on (assertive) dominance and (calm) submission, the leader being the most dominant. Their willingness to accept the existence of dominant behavior is motivated by their desire to validate their training theories, but their interpretation of the term is far from what ethologists understand by it.

Moralistic Dog Training (aka positive reinforcement training) distances itself from punishment, dominance, and leadership. They don’t define ‘dominance’ properly either, but use it with connotations of ‘punishment,’  ‘aggression,’  ‘coercion,’  ‘imposition.’ They claim dominance does not exist and regard it as a mere construct of philosophers and ethologists aimed at justifying the human tendency to dominate others. Their view is that we should nurture our dogs as if they were part of our family and should not dominate them. Therefore, they also distance themselves from using and condoning the use of terms like ‘alpha,’  ‘leader’ and ‘pack.’ The more extreme factions claim to refrain from using any aversive or signal that might be slightly connected with an aversive (like the word ‘no’) and deny their using of punishers (which, given the consensually accepted scientific definition of punishment, is a logical impossibility). Their refusal to accept the existence of dominant behavior is motivated by their desire to validate their training morality, but their interpretation of the term is again far from what ethologists understand by it.

An ethological definition of ‘dominant behavior’ is (as I suggest in “Dominance—Making Sense of the Nonsense”): “Dominant behavior is a quantitative and quantifiable behavior displayed by an individual with the function of gaining or maintaining temporary access to a particular resource on a particular occasion, versus a particular opponent, without either party incurring injury. If any of the parties incur injury, then the behavior is aggressive and not dominant. Its quantitative characteristics range from slightly self-confident to overtly assertive.”

This is a descriptive statement, a classification of a class of behaviors, so we can distinguish it from other classes of behaviors, based on the observable function of behavior (according to evolutionary theory). It is clearly distinguishable from the statements of both opposing mainstream dog-training groups in that it does not include any normative guidance.

3 Logical fallacies

logical fallacy is unsound reasoning with untrue premises or an illogical conclusion. Logical fallacies are inherent in the logic structure or argumentation strategy and suit irrational desires rather than actual matters of fact.

An argument can be valid or invalid; and valid arguments can be sound or unsound. A deductive argument is valid if, and only if, the conclusion is entailed by the premises (it is a logical consequence of the premises). An argument is sound if, and only if, (1) the argument is valid and (2) all of its premises are true. The pure hypothetical syllogism is only valid if it has the following forms:

If P ⇒ Q and Q ⇒ R, then P ⇒ R

If P ⇒ ~R and ~R ⇒ ~Q, then P ⇒ ~Q

This mixed hypothetical syllogism has two valid forms, affirming the antecedent or “modus ponens” and denying the consequent or “modus tollens”:

If P ⇒ Q and P, then Q (modus ponens)

If P ⇒ Q and ~Q, then ~P (modus tollens)

It has two invalid forms (affirming the consequent and denying the antecedent).

The naturalistic fallacy is the mistake of identifying what is good with a natural property. In this fallacy, something considered natural is usually considered to be good, and something considered unnatural is regarded as bad. The structure of the argument is “P is natural, therefore P is moral” or “P is natural and non-P is unnatural, natural things are moral and unnatural things immoral, therefore P is moral and non-P immoral.” G. E. Moore coined the term naturalistic fallacy in 1903 in “Principia Ethica.” It is related to the ‘is-ought problem,’ also called ‘Hume’s Law’ or ‘Hume’s Guillotine,’ described for the first time by David Hume in 1739 in “A Treatise of Human Nature.” The ‘is-ought fallacy’ consists of deriving an ought conclusion from an is premise. The structure of the argument is “P is, what is ought to be, therefore P ought to be.”

The moralistic fallacy is the reverse of the naturalistic fallacy. It presumes that what ought to be preferable is what is, or what naturally occurs. In other words: what things should be is the way they are. E. C. Moore used the term for the first time in 1957 in “The Moralistic Fallacy.” The structure of the argument is, “P ought to be, therefore P is.”

Roger Abrantes and Shakira

“There is no evidence that dogs attempt to dominate others or that they don’t. On the contrary, all evidence suggests that dogs (as most animals) use different strategies depending on conditions including costs and benefits. Sometimes they display dominant behavior, other times they display submissive behavior, and yet other times they display some other behavior” Picture by (L’Art Au Poil École).

The line of argumentation of Naturalistic Dog Training is: Dogs naturally attempt to dominate others; therefore, we ought to dominate them. We can transcribe this argument in two ways (argument 1a and 1b):

Argument 1a

(A) If the nature of dogs is to attempt to dominate others, then I ought to train dogs according to their nature. (P⇒Q)

(B) It is the nature of dogs to attempt to dominate others. (P)

Therefore: I ought to train dogs by attempting to dominate them. (Q)

Argument 1b

(A) If dogs dominate others, then it’s right to dominate others. (P⇒Q)

(B) If it’s right to dominate others, then I have to do the same to be right. (Q⇒R)

Therefore: If dogs dominate others, then I have to do the same to be right. (P⇒R)

We cannot derive ‘ought’ from ‘is.’ Arguments 1a and 1b commit the ‘naturalistic fallacy.’ Both arguments seem formally valid, except that they derive a norm from a fact. There is no logical contradiction in stating, “I ought not to train dogs according to their nature.” They are also unsound (the conclusions are not correct) because premises P are not true.

There is no evidence that dogs attempt to dominate others or that they don’t. On the contrary, all evidence suggests that dogs (like most animals) use different strategies depending on conditions, which include costs and benefits. Sometimes they display dominant behavior, other times they display submissive behavior, and other times they display other behavior. Even when particular dogs are more prone to use one strategy rather than another, we are not entitled to conclude that this is the nature of dogs.

Conclusion: whether science proves that dogs display or don’t display dominant behavior has nothing to do with whether or not it is morally right for us to dominate our dogs.

The line of argumentation of Moralistic Dog Training is: We ought not to attempt to dominate our dogs; therefore, dogs do not attempt to dominate us. We can transcribe this argument in two ways (argument 2a and 2b):

Argument 2a

(A) Dominance is bad. (P⇒Q)

(B) Dogs are not bad. (R⇒~Q)

Therefore: Dogs do not dominate. (R⇒~P)

Argument 2b

(A) If [dominance exists], it is . (P⇒Q)

(B) If it is , [dogs don’t do it]. (Q⇒R)

Therefore: if [dominance exists], [dogs don’t do it]. (P⇒R)

We cannot derive ‘is’ from ‘ought.’ Arguments 2a and 2b commit the ‘moralistic fallacy.’ Argument 2a is formally invalid even if the premises were true because the conclusion is not entailed in the premises (it is the same as saying red is a color, blood is not a color, so blood is not red). Argument 2b sounds a bit odd (in this form), but it is the only way I have found of formulating a valid argument from the moralistic trainers’ argument. It is formally valid but it is unsound because it commits the moralistic fallacy: in its second line, it derives a fact from a norm. It assumes that nature doesn’t do wrong (or what is good is natural), but there is no contradiction in assuming the opposite.

Conclusion: the fact we believe it is morally wrong to dominate our dogs does not mean that dogs do not display dominant behavior. We are entitled to hold such a view, but it does not change the fact that dogs display dominant behavior.

4 Social conditioning and emotional load

The choice of word by ethologists to coin the behavior in English, i.e. ‘dominant,’ also contributes to the dispute. Curiously enough, the problem does not exist in German where dominant and submissive behaviors are ‘überlegenes verhalten’ and ‘unterlegenes verhalten.’

All words we use have connotations due to accidental social conditioning and emotional load. A scientist knows he** cannot afford his judgment to be clouded by his own accidental social conditioning or emotions. A defined term means that and only that. It’s not good, not bad, not right, not wrong, and the issue of whether he likes it or not does not even enter the equation. As an individual he may have his own personal opinion and moral viewpoint, but he does not allow them to affect his scientific work. As individuals, we all have our own likes and dislikes because we are constantly being conditioned by our environment. Culture, trends, movements, environments, relationships and moods, all bias our attitudes towards particular terms. Nowadays, for reasons I will leave to historians and sociologists to analyze, the words ‘dominance’ and ‘submission’ have negative connotations for many people. When people, all of whom are subject to social conditioning, fail to distinguish between the scientific meaning of the words and their everyday connotations, they repudiate them, which is understandable.

Conclusion: a class of behavior that animals use to solve conflicts without harming one another is what ethologists call dominant and submissive behavior. This behavior, in the way I describe and define, exists (see above). You may not like the terms or indeed the behaviors, but that doesn’t mean they don’t exist. ‘Red’ is a characteristic of an object that provides particular information to our eyes as a result of the way it reflects or emits light. We can argue (and we do) about the definition of ‘red,’ what is red, what is not, when it becomes orange, but we do not deny that red exists. You may object to the name ‘red’ but objects will continue to reflect or emit light in a particular way that produces what we call red (or whatever we choose to call it). A ‘red flower’ (or a display of ‘dominant behavior’) is not an abstract concept, but a real, detectable thing, whilst the concept of ‘redness’ is an abstract notion, as are the concepts of ‘dominance,’  ‘height,’  ‘weight,’  ‘strength,’ etc…

5 Unclear grammar

Another problem is that we use the word dominance as a noun (an abstract noun in contrast to a concrete noun) when in this case it is (or should be) a ‘disguised adjective.’ Adjectives don’t make sense without nouns (except for adjectival nouns). Dominance is an abstract noun, something that by definition does not exist (otherwise it wouldn’t be abstract), except as an abstract notion of ‘showing dominant behavior’ and as in ‘dominant allele,’  ‘dominant trait,’  ‘dominant ideology,’ ‘dominant eye,’ etc. However, the behavior of alleles, traits, ideologies and eyes, which we call dominant or classify as dominant, exists. For example, the question “Do dogs show dominance towards humans?” uses the abstract noun ‘dominance’ as an adjectival noun instead of the more correct ‘dominant behavior’. This can be confusing for some as it suggests that dominance is an intrinsic quality of the individual, not the behavior. Therefore, I suggest that, in the behavioral sciences, we henceforth drop the adjectival noun and only use the term as an adjective to behavior. This is a very important point and a source of many misunderstandings and misconceptions regarding the character of behavior.

Behavior is dynamic and changeable. An individual displays one behavior at one given moment and another a while later. The popular view maintains the notion of a ‘dominant individual’ as the one that always shows dominant behavior and the ‘submissive individual’ as the one that always shows submissive behavior, which is not true. Dominant and submissive (dominance and submission) are characteristics of behavior, not individuals. Individuals may and do change strategies according to a particular set of conditions, although they may exhibit a preference for one strategy rather than another.

It is the ability to adopt the most beneficial strategy in the prevailing conditions that ultimately sorts the fittest from the less fit—moral strategies included.

Have a great day,



* This is my normative judgment and as such no one can contest it.

** The most correct form would be ‘he/she,’ or ‘he or she,’ but since I find it extremely ugly from a linguistic point of view (my normative judgment) to use this expression repeatedly, I chose to write, ‘he’ though not by any means neglecting the invaluable and indisputable contribution of my female colleagues.


  • Abrantes, R. 1986. The Expression of Emotions in Man And Canid. Waltham Symposium, Cambridge, 14th-15th July 1986.
  • Abrantes, R. 1997. The Evolution of Canine Social Behavior. Wakan-Tanka Publishers (2nd ed.  2005).
  • Abrantes, R. 2011. Dominance—Making Sense Of The Nonsense.
  • Ayer, A. J. 1972. Probability and Evidence. Macmillan, London.
  • Bekoff, M. & Parker, J. 2010. Wild Justice: The Moral Lives of Animals. Univ. Of Chicago Press.
  • Bland J. 2002 About Gender: Dominance and Male Behaviour.
  • Copi, I. M. and Cohen, C. 1990. Introduction to Logic (8th ed.). Macmillan.
  • Dennet, D. 1996. Darwin’s Dangerous Idea: Evolution and the Meanings of Life. Simon & Schuster.
  • Dennet, D. 2003. Freedom Evolves. Viking Press 2003.
  • Futuyma, D. J. 1979. Evolutionary Biology. Sinauer Assoc.
  • Galef, J. 2010. Hume’s Guillotine.
  • Hewitt, S. E., Macdonald, D. W., & Dugdale, H. L. 2009. Context-dependent linear dominance hierarchies in social groups of European badgers, Meles melesAnimal Behaviour, 77, 161-169.
  • Hume, D. 1739. A Treatise of Human Nature. Oxford University Press, Oxford, 1967, edition.
  • Locke, J. 1690. An Essay Concerning Human Understanding
  • Kanazawa, S. 2008. Two Logical Fallacies That We Must Avoid.
  • Kierkegaard, S. 1844. Philosophiske Smuler eller En Smule Philosophi (Philosophical Fragments). Samlede Værker, Nordisk Forlag, 1936.
  • Lorenz, K. 1963. Das sogenannte Böse. Zur Naturgeschichte der Aggression. Wien, Borotha-Schoeler Verlag, 1969.
  • Moore, E. C. 1957. The Moralistic Fallacy. The Journal of Philosophy 54 (2).
  • Moore, G. E. 1903. Principia Ethica.
  • Novella, S. 2012. The Paradox Paradox.
  • Pinker, S. How the Mind Works. New York: W. W. Norton & Company, 1997.
  • Popper, K. 1963. Conjectures and Refutations.  Routledge and Kegan Paul, London, UK.
  • Popper, K. Objective Knowledge: An Evolutionary Approach. Oxford University Press.
  • Rachels, J. 1990. Created From Animals. Oxford University Press.
  • Rowell, T. E. 1974. The Concept of Social Dominance. Behavioral Biology, 11, 131-154.
  • Ruse, M. 1986. Taking Darwin seriously: a naturalistic approach to philosophy. Prometheus Books.

Thanks to Anabela Pinto-Poulton (PhD, Biology), Simon Gadbois (PhD, Biology), Stéphane Frevent (PhD, Philosophy), Victor Ross (Graduate Animal Trainer EIC), Parichart Abrantes (MBA), and Anna Holloway (editor) for their suggestions to improve this article. The remaining flaws are mine, not theirs.

Canine Ethogram—Social and Agonistic Behavior

Behavior is the response of the system or organism to various stimuli, whether internal or external, conscious or subconscious, overt or covert, and voluntary or involuntary.

Behavior does not originate as a deliberate and well-thought strategy to control a stimulus. Initially, all behavior is probably just a reflex, a response following a particular anatomical or physiological reaction. Like all phenotypes, it happens by chance and evolves thereafter.

Natural selection favors behaviors that prolong the life of an animal and increase its chance of reproducing; over time, a particularly advantageous behavior spreads throughout the population. The disposition (genotype) to display a behavior is innate (otherwise the phenotype would not be subject to natural selection and evolution), although it requires maturation and/or reinforcement for the organism to be able to apply it successfully. Behavior is, thus, the product of a combination of innate dispositions and environmental factors. Some behaviors require little conditioning from the environment for the animal to display it while other behaviors requires more.


Canine Ethogram Social Agonistic Photos

Pictures illustrating canine social and agonistic behavior. For the classification of the behavior, please see ethogram below. Behavior is dynamic (not static). All interpretations are therefore only approximate and as pictures allow.


An organism can forget a behavior if it does not have the opportunity to display it for a period of a certain length, or the behavior can be extinguished if it is not reinforced for a period.

Evolution favors a systematic bias, which moves behavior away from maximization of utility and towards maximization of fitness.

Social behavior is behavior involving more than one individual with the primary function of establishing, maintaining, or changing a relationship between individuals, or in a group (society).

Most researchers define social behavior as the behavior shown by members of the same species in a given interaction. This excludes behavior such as predation, which involves members of different species. On the other hand, it seems to allow for the inclusion of everything else such as communication behavior, parental behavior, sexual behavior, and even agonistic behavior.

Sociologists insist that behavior is an activity devoid of social meaning or social context, in contrast to social behavior, which has both. However, this definition does not help us much because all above mentioned behaviors do have a social meaning and a context unless ‘social’ means ‘involving the whole group’ (society) or ‘a number of its members.’ In that case, we should ask how many individuals are needed in an interaction to classify it as social. Are three enough? If so, then sexual behavior is not social behavior when practiced by two individuals, but becomes social when three or more are involved, which is not unusual in some species. We can use the same line of arguing for communication behavior, parental behavior, and agonistic behavior. It involves more than one individual and it affects the group (society), the smallest possible consisting of two individuals.

Agonistic behavior includes all forms of intraspecific behavior related to aggression, fear, threat, fight or flight, or interspecific when competing for resources. It explicitly includes behaviors such as dominant behavior, submissive behavior, flight, pacifying, and conciliation, which are functionally and physiologically interrelated with aggressive behavior, yet fall outside the narrow definition of aggressive behavior. It excludes predatory behavior.

Dominant behavior is a quantitative and quantifiable behavior displayed by an individual with the function of gaining or maintaining temporary access to a particular resource on a particular occasion, versus a particular opponent, without either party incurring injury. If any of the parties incur injury, then the behavior is aggressive and not dominant. Its quantitative characteristics range from slightly self-confident to overtly assertive.

Dominant behavior is situational, individual and resource related. One individual displaying dominant behavior in one specific situation does not necessarily show it on another occasion toward another individual, or toward the same individual in another situation.

Dominant behavior is particularly important for social animals that need to cohabit and cooperate to survive. Therefore, a social strategy evolved with the function of dealing with competition among mates, which caused the least disadvantages.

Aggressive behavior is behavior directed toward the elimination of competition while dominance, or social-aggressiveness, is behavior directed toward the elimination of competition from a mate.

Fearful behavior is behavior directed toward the elimination of an incoming threat.

Submissive behavior, or social-fear, is behavior directed toward the elimination of a social-threat from a mate, i.e. losing temporary access to a resource without incurring injury.

Resources are what an organism perceives as life necessities, e.g. food, mating partner, or a patch of territory. What an animal perceives to be its resources depends on both the species and the individual; it is the result of evolutionary processes and the history of the individual.

Mates are two or more animals that live closely together and depend on one another for survival.

Aliens are two or more animals that do not live close together and do not depend on one another for survival.

A threat is everything that may harm, inflict pain or injury, or decrease an individual’s chance of survival. A social-threat is everything that may cause the temporary loss of a resource and may cause submissive behavior or flight, without the submissive individual incurring injury. Animals show submissive behavior by means of various signals, visual, auditory, olfactory and/or tactile.

Canine Ethogram

Pictures illustrating canine social and agonistic behavior. For the classification of the behavior, please see ethogram below. Behavior is dynamic (not static). All interpretations are, therefore, only approximate and as pictures allow.


The diagram does not include a complete list of behaviors.

As always, have a great day!


PS—I apologize if by chance I’ve used one of your pictures without giving you due credit. If this is the case, please e-mail me your name and picture info and I’ll rectify that right away.


Muzzle Grab Behavior in Canids

Muzzle grab in adult wolves (photo by Monty Sloan).

A “Muzzle grab” is a common behavior shown by social canines, e.g. wolves (Canis lupus lupus), dingoes (Canis lupus dingo), and dogs (Canis lupus familiaris)The function of this behavior is to confirm a relationship rather than to settle a dispute. The more self-confident individual will muzzle grab a more insecure opponent and thus assert its social position. The more insecure individual does not resist the muzzle grab. On the contrary, it is often the more insecure individual that shows submissive behavior by literally inviting its opponent to muzzle grab it. Even though we sometimes see this behavior at the end of a dispute, wolves and dogs only use it toward individuals they know well (pack members) almost as way of saying  “You’re still a cub (pup).” The dispute itself does not tend to be serious, just a low-key challenge, normally over access to a particular resource. Youngsters, cubs and pups sometimes solicit adults to muzzle grab them. This behavior appears to be reassuring for them, a means of saying, “I’m still your cub (pup).”

Dog muzzle grab.

Dogs also show the muzzle grab behavior (photo by Marco de Kloet).

The muzzle grab behavior emerges early on. Canine mothers muzzle grab their puppies (sometimes accompanied by a growl) to deter them from suckling during weaning. Cubs and pups also muzzle grab one another during play, typically between six and nine weeks of age. They probably learn through play that the muzzle grab is a good way of stopping an opponent from doing something. Cubs and pups also learn the importance of bite inhibition when showing muzzle grab. If they bite their opponent too hard, they will elicit a fight and will get hurt. A muzzle grab, therefore does not involve biting, just grabbing. This behavior helps develop a relationship of trust between both parties: “We don’t hurt one another.”

When used as a means of settling a dispute, a muzzle grab looks more violent and normally ends with the muzzle-grabbed individual showing passive submissive behavior. However, the participants very seldom get hurt, an occurrence that would counteract the function of the behavior itself.

Wolf cub muzzle grab

Cubs and pups muzzle grab one another during play (photo by Monty Sloan).

A muzzle grab requires self control. Higher ranking wolves and dogs muzzle grab their pack members (team mates) and by doing so confirm their rank and display self control. Lower ranking wolves and dogs invite muzzle grabbing behavior in order to confirm their acceptance of their social position and to reassure themselves that they are still accepted.

The muzzle grab behavior probably originated as both a form of maternal (and later paternal) behavior and as a play behavior amongst cubs. As it proved beneficial to all concerned, it became a factor for natural selection and spread from generation to generation, evolving in the same way as any other trait that increases the fitness of an individual.

In domestic dogs, when the puppies are five to seven weeks old, their mother muzzle grabs them regularly. At first, their mother’s behavior frightens them and they may whimper excessively, even if the mother has not harmed them in any way. Later on, when grabbed by the muzzle, the puppy immediately shows passive submission (lies down with its belly up). Previously, it was assumed that the mother needed to pin the puppy to the ground, but this is not the case as most puppies submit voluntarily. Over time, this behavior pattern assumes variations. Wolf cubs and puppies often invite the alpha male (leader of the pack) as well as other adults to grab them by the muzzle. They solicit a demonstration of their elders’ superiority and self control, whilst at the same time they show their acceptance and submission. This is the most reassuring behavior an adult dog can show a puppy.

Domestic dogs sometimes approach their owners puffing to them gently with their noses. By grabbing them gently around the muzzle, we reaffirm our acceptance of them, our self-control and our control of the environment in general. After being muzzle grabbed for a while, the dog will normally show a nose lick, maybe yawn and then walk calmly away. It’s like the dog saying, “I’m still your puppy” and the owner saying, “I know and I’ll take good care of you.”

The muzzle grab behavior can be difficult to classify. Some researchers classify it as social behavior, others as agonistic behavior and a third group places it in the distinct category of pacifying behavior. Since the function of this behavior is primarily to confirm a relationship between two individuals, this author classifies it as social behavior.

As always, have a great day!


Related articles

Dogs And Children

Dogs And ChildrenDogs And Children

Children and dogs in the same household equals many moments of joy for the whole family, dogs included. There are a few considerations that parents should bear in mind and a few rules that children and dogs must learn. These rules are simple and easy to learn.

This book contains sound advice for parents and dog owners.

This little book was published in Danish by Borgen Publishers in 1986 as one in a series of five booklets that dealt with the most common questions asked by dog owners and the problems they ran into. It was never reprinted after the first edition of 25,000 copies sold out. It became since then a bit of a collector’s item.

“Dogs and Children” was published in Danish, Norwegian and Italian, and never translated into English until now. I have often been asked to write about dogs and children and I have done so occasionally in short articles and blogs, but the advice has never been published as a book, except for this booklet. The other day, whilst dusting off my books, I came across the five booklets and thought it would be a good idea to translate the original “Hund og Barn” into English. I have kept the original photos and layout and it is now available free of charge as a flip-page E-book.

The intention of this little book was to provide dog owners with sound advice that would help them prevent accidents from happening and, as such, I believe that it can still perform the same role today as it did 26 years ago.

Click the full-screen view icon for better reading. Zoom in and out as you please.


Sorry, this book is no longer available here. Please, visit Ethology Institute’s Online Bookstore.


Related articles

Life’s All About Food And Sex

Pandas at the Chiang Mai Zoo in northern Thailand.

To offer food to females in exchange for sex works well for most males in various species, except when one eats too much of it. At the Chiang Mai Zoo in Northern Thailand, the male panda, is apparently too fat to have sex and his partner, the female Lin Hui, has lost interest. Zoo keepers have done everything to spice up their sex life including showing them movies of other pandas having sex!... (photo from Chiang Mai Zoo).

The other day I went to my favorite bar (and yes, of course it’s Irish) to drink a couple of beers, play some pool and have a bit of fun. The regulars, my mates, are an eclectic mix of professions, trades, ages, economic status, ethnic backgrounds and sexual orientations, all with very different interests in life. Mostly, we just have fun, drink beers and martinis, play pool, discuss football and holidays, complain about how everything got to be so expensive and the incompetence of politicians, and we laugh at a good as well as a bad joke. Sometimes, someone throws in a kind of provocation, a more complex question.

“Cheers, mate!” someone shouts to me from across a table. “You’re a biologist, so you may be able to answer my question for the day: what’s life all about?”

Fruit Fly Boozing

Sex-deprived fruit fly males drink alcohol four times more than others (photo from Geekologie).

“Cheers to you too, mate. You can’t ask a thing like that. That’s not a bar question,” I say. “But, no problem there. Life is all about food and sex,” I add, taking a slug of my wonderfully cold beer and confirming my view that the first beer always tastes the best.

“Hey, I’m asking a serious question,” he protests, “gimme a serious answer!”

“I’m giving you a serious answer. Everything living organisms do is to get either food or sex. Food is a great thing. Firstly, it is necessary to survive and you need to survive in order to have sex, because, if you’re dead, you can’t have sex. Secondly, females in particular love food because they need food to survive, so they can have sex, so they can have progeny; and their progeny needs food, lots of food.”

“Maybe for other animals,” he argues. “But for us humans, there’s more to life than food and sex. What about science, for one?”

“Very simple—science is a means to an end. Why do you think there are many more male scientists than female? Because they need to invent easy ways to get food to give to females, because the females then get all crazy about them and they have sex. Then, they get more progeny who need more food, which implies more science, more sex…”

“You’re far out, mate, we’re more than that. We have a soul, we produce great art!”

“Most good art is produced by unhappy males. How often do  you hear of a great, happy artist? Do you know why they are unhappy? Because they don’t get the sex they want. That’s why they produce art. Females like beauty because the more beautiful sons or daughters they have, depending on the species, the more grandchildren they will have. Also, artists are normally safe, they are sensitive and it is unlikely they will kill their progeny. So, males produce all the art they can to impress the females so they have sex with them. Then, they get more progeny, and the progeny needs more food, which…”

“OK, I got it,” he says, “artists are sensitive, but what about power?  I guess you’ll say it’s another way of getting sex.”

“You’re right. You’re a quick learner. Powerful males can in theory provide better for their progeny so females like powerful males. For the males, this is good news because if they don’t have a clue about art, they can always try to become rich or powerful, which are basically the same thing. Power means more sex because progeny that are well provided for survive longer, have more sex and have their own progeny, which means grandchildren. This means they need more food, more science…”

“What if I’m not good at art or at the power game?”

“Then, mate, you are in deep s… in terms of sex, but don’t worry, it happens to most males in most species. You can always bluff. Most males do.”

“Well, that’s maybe why I’m here drinking with my mates…”

“Could be. Fruit fly males deprived of sex drink four times more than their mates that have sex.”

“You’re kidding me!” he exclaims.

“No, I’m not, that’s scientific proven. It’s all a question of maintaining the levels of a neuropeptide in the brain and if you can’t have sex, booze seems to do it—for fruit files, that is. Fruit flies don’t play pool though, so no worries about that,” I say.

“Doesn’t sound fair to me,” he replies, “but who programmed this bloody thing anyway? Don’t tell me it was…”

“Nobody. Genes have only one goal, which is to reproduce, no matter what genes we’re talking about. It’s all about surviving and reproducing, eating and copulating. It’s like an algorithm, a very simple one indeed.”

“Not that I’m complaining, mate, not too much anyway, but it does bother me. It seems like the females control everything.”

“They do. In most species they choose the males. Virtually all females will mate and reproduce. For the males, it’s a lot more difficult. Competition is fierce and females are picky. Many males never get a chance. That’s why they have to trick the females with all their cunning, but food is the best and most direct way. Males try desperately to improve their chances, in some species by means of attractive exteriors, in others by appearing powerful. Basically, it’s all a bluff to impress the females.”

“So, the females are picky so they can get the best progeny and the best progeny of the progeny. Did I get that right?’

“Too right, mate. Males bluff, but females get better and better at calling their bluff because their main concern is to produce good progeny.”

“OK, I understand that and I can see what the females get, but one things beats me: what about the males, what do they get?”


Have a beautiful day!


Should We Reinforce the Effort or the Results?

Dog using its nose to search a target scent (photo from

If you ask, “should we reinforce effort or the results?” you are liable to get as many answers supporting one idea as the other. Supporters of reinforcing effort sustain that reinforcing results creates emotional problems when one doesn’t succeed and decreases the rate of even trying. Supporters of reinforcing results maintain that reinforcing effort encourages sloppiness and cheating.

I shall proceed to argue for and against both theories and prove that it is not a question of either/or, rather of defining our criteria, processes and goals clearly.

I shall compare the learning of some skills in dogs and humans because the principles are the same. The difference between them and us is one “of degree, not of kind,” in the words of Darwin.

I will use SMAF to accurately describe some of the processes whenever I consider it advantageous. If you are not proficient in SMAF, you can read the free SMAF manual at

The main difficulty in some learning processes is reinforcing the right behavior at the right time, which bad teachers, bad parents and bad trainers do not master (here bad means inefficient, it is not a moral judgment).

Much of my personal work with dogs (and rats) is and has been detection work, mainly of narcotics and explosives, but also of people tobacco and other items. One of the first signals I teach the animals is a disguised reinforcer.

With dogs, I use the sound ‘Yes’ (the English word) and with rats a ‘bip….bip….bip’ sound produced on their backpacks and triggered by me.The signal part of this signal/reinforcer means “continue what you’re doing” and the reinforcer part “we’re OK, mate, doing well, keep up.” This is a signal that becomes a reinforcer: Continue,sound(yes) that becomes a “!+sound”(yes).

The difference between the most used “!±sound”(good-job) and “!+sound”(yes) is that the former is associated and maintained with “!-treat”(small food treat) and “!-body(friendly body language) and the latter with a behavior that will eventually produce “!-treat”. The searching behavior does not produce a treat, but continuing searching does, eventually (find or no find). This is why “!+sound”(yes) is a disguised Continue,sound(yes), or the other way around.

Why do I need this interbreeding between a signal and a reinforcer?

Because the signal ‘Search’ (Search,sound) does not mean ‘Find the thing.’ Sometimes (most of the time) there’s nothing to find, which is a relief for all of us (airports and the likes are not that full of drugs and explosives).

So, what does Search,sound mean? What am I reinforcing? The effort?

No, I’m not. We have to be careful because if we focus on reinforcing the effort, we may end up reinforcing the behavior of the animal just strolling around, or any other accidental and/or coincidental behavior.

I am still reinforcing the result. ‘Search’ means, “Go and find out whether there is a thing out there.” ‘Thing’ is everything that I have taught the dog to search and locate for me, e.g. cocaine, hash, TNT, C4.

“Go and find out whether there is a thing out there” leaves us with two options that are equally successful: ‘here’ and ‘clear.’ When there is a thing present, the dog answers ‘here’ by sitting as close to it as possible, or pointing to it (I have taught it those behaviors). When there is no thing, I want the animal to tell me exactly that: the dog answers ‘clear’ by coming back to me (again because I have taught it to do that). We have two signals and two behaviors:

Thing,scent => dog sits (‘here’ behavior).

∅Thing,scent => dog comes back to me (‘clear’ behavior).

The signals are part of the environment, they are not given by me, which does not matter: a signal (SD) is a signal*. An SD is a stimulus associated with a particular behavior and a particular consequence or class of consequences. When we have two of them, we expect two different behaviors and when there is none, we expect no behavior. What fools us here is that in detection work we always have one and only one SD, either a scent or the absence of one. It is not possible to have none. Either we have a scent or we don’t, which means that either we have Thing,scent or we have ∅Thing,scent, each requiring two different behaviors as per usual. The one SD is the absence of the other.

Traditionally, we don’t reinforce a search that doesn’t produce a find. To avoid extinguishing the behavior, we use ‘controlled finds’ (a drug or an explosive, we know it is there because we have placed it there to give the animal a possibility to obtain a reinforcer).

This solution is correct, except that it teaches the dog that the criterion for success is ‘to find’ and not ‘not to find,’ which is not true. ‘Not to find’ (because there is no thing out there) is as good as ‘to find.’ The tricky part is, therefore, to reinforce the ‘clear’ and how to do it to avoid sloppiness (strolling around) and cheating.

Let us analyze the problem systematically.

The following process does not give us any problems:

{Search,sound ⇒ b1(dog searches) => “!+sound”(yes) or Continue,sound(yes) ⇒ b1(dog searches) ⇒ dog finds thing (Thing,scent) ⇒ b2(dog sits=’here’ behavior) => “!+sound” + “!-treat”};

No problem, but what about when there is no thing (∅Thing,scent)? If I don’t reinforce the searching behavior, I might extinguish it. Then, I reinforce the searching with “!+sound”(yes):

{“Search,sound” ⇒ b1(dog searches) => “!+sound”(yes) ⇒ b1(dog searches) => ∅Thing,scent ⇒ b3(dog comes back to me=’clear’ behavior) => “!+sound” + “!-treat”};

It all looks good, but it poses us some compelling questions:

How do I know the dog is searching versus strolling around (sloppiness)?

How do I know I am reinforcing the searching behavior?

If I reinforce the dog coming back to me, then next time I risk the dog having a quick sniff round and coming straight back to me. That’s the problem. I want the dog to come back to me only when it finds nothing (as in it didn’t find anything).


Reinforcing the searching behavior.

Identifying the searching behavior versus strolling around (sloppiness). How can I make sure that the dog always searches and never just strolls around?


Reinforcing the searching behavior with “!+sound”(yes) works. OK.

Remaining problem:

I have to reinforce the ‘clear’ behavior (coming back to me), but how can I ensure the dog always searches and never just strolls around (avoid sloppiness)?

How can I make sure the dog has no interest in being sloppy or cheating me?


To teach the dog that reinforcers are only available if and only if:

1. the dog finds the thing. {Thing,scent ⇒ b2(dog sits) => “!+sound” + “!-treat”};

2. the dog does not ever miss a thing. {∅Thing,scent ⇒ b3(dog comes back to me) => “!+sound” + “!-treat”};


I gradually teach the dog to find things until I reach a predetermined low concentration of scent (my goal). In this phase of training, there is always one thing to find. After 10 consecutive successful finds (my criterion and quality control measure), all producing reinforcers for both the searching (“!+sound”(yes)) and the finding (“!+sound” + “!-treat”), I set up a situation with no thing present (∅Thing,scent). The dog searches and doesn’t find anything. I reinforce the searching and the finding (no-thing) as previously. Next set-up: I make sure there is a thing to find and I reinforce both searching and finding.

I never reinforce not-finding a thing that is there, nor finding a thing that is not there.

Consequence: the only undesirable situations for a dog is

(1) not-finding a thing that is there (the dog did not indicate Thing,scent), or (2) indicating a thing that is not there (the dog indicates ∅Thing,scent).

(1) {Thing,scent ⇒ b3(dog comes back to me=‘clear’ behavior) => [?+sound] + [?-treat]};


(2) {∅Thing,scent ⇒ b2(dog sits=‘here’ behavior) => [?+sound] + [?-treat]};

This is (negatively) inhibiting negligence, but since it proves to increase the intensity of the searching, we cannot qualify it as an inhibitor (earlier punisher). Therefore, we call it a non-reinforcer: “∅+sound”, “∅-treat”.

In the first case: Thing,scent => Dog comes back to me => [?+sound] + [?-treat].


Thing,scent => Dog comes back to me => “∅+sound”, “∅-treat”.


{Thing,scent ⇒ b3(dog comes back to me) => “∅+sound”, “∅-treat” ⇒ b1(dog searches–more intensively) => Thing,scent ⇒ b2(dog sits=‘here’ behavior) => “!+sound” + “!-treat”};

In the second case, I have to be absolutely sure that there is indeed no thing. The training area must be free from any scent remotely similar to the scent we are training (Thing,scent). This is an imperative, especially in the first phases of the training process, and the trainer that misses this point is committing major negligence.

Nevertheless, should the dog, show ‘here’ for ∅Thing,scent, then we can use the same procedure as above:

{∅Thing,scent ⇒ b2(dog shows ‘here’ behavior) => “∅+sound”, “∅-treat” ⇒ b1(dog searches–more intensively) => ∅Thing,scent ⇒ b3(dog comes back to me=‘clear’ behavior) => “!+sound” + “!-treat”}.

What if later the dog doesn’t find a thing that is there in a lower concentration than the one I used for training, or is masked by other scents?

That’s no problem, it’s not the dog’s fault. I didn’t train for it. The dog doesn’t know that it is making a mistake by giving me a wrong ‘clear.’ As far as the dog is concerned, the room is clear: {∅Thing,scent ⇒ b3(dog comes back to me => “!+sound” + “!-treat”}; The dog was not strolling around and is not cheating me.

A human example:

I reinforce the child trying to solve a math problem. ‘Well done, but you got it wrong because…” The solution is wrong, but the method was correct. Therefore, it is all a question of training. The ‘wrong’ will be eliminated with more or better training, or maybe it was caused by an excessive increase in the difficulty curve of the problem (the teacher’s problem). We are not reinforcing trying; we are reinforcing the correct use of a method.

Why reinforce the process?

We must reinforce the process because of its emotional and motivational consequences. The dog and the child must accept the challenge, must want to be challenged, and be able to give their best in solving the problem. The exercise in itself will eventually end up being self-reinforcing.

Are we reinforcing the effort rather than the success?

No, we are not. Reinforcing the effort rather than the result can even lead to false positives. The animal indicates something that it is not there because it associates the reinforcer with the behavior, not the thing. Children give us three-four quick, consecutive, wrong answers if we reinforce the trying, not the process (thinking before answering).

We reinforce result (success) only.

When the dog doesn’t find because there’s nothing to find, that is success. When the dog doesn’t find because the concentration was too low, that is also success because ‘too low’ is here equal to ‘no thing.’ When the child gets it wrong, it is because the exercise exceeded the capacity of the child (he or she has not been taught to that level).

We are still reinforcing success and exactly what we trained the dog and the child to do. We don’t say to the child, “Well, you tried hard enough, good.”

We say, ” Well done, you did everything correctly, you just didn’t get it right because you didn’t know that x=2y-z and there was no way of you knowing.”

Next time, the child gets it right because she now knows it; and if not, it is because x=2y-z exceeds the capacity of that particular child in which case there’s nothing you can do about it.

The same goes for the dog: the dog doesn’t indicate 0.01g of cocaine because I trained it to indicate as low as 0.1g. When I reinforce the dog’s ‘clear,’ I say, “Well done, you did everything correctly, you just didn’t get it right because you didn’t know that 0.01g cocaine is still the thing.”

Now, I train the dog that ‘thing’ means ‘as low as 0.01g cocaine’ and either the dog can do it or it cannot. If it can, good; if it cannot, there’s nothing you or I, the dog or the child can do about it.


We reinforce result, success, not the effort, not the trying. We must identify success, have clear criteria for success, plan a successive approach to our goal and gradually increase difficulty. We must be able to recognize limits and limitations in ourselves, in the species we work with, in the individuals we tutor, in the particular skill we teach. We must know when we cannot improve a skill any further and when an individual cannot give us more than what we are getting; and be satisfied with that.

Have a great day!


Footnote: * Strictly speaking, the scent that the detection dog searches is not a signal, but a cue, because it is not intentional. In this context, however, it is and SD because we have conditioned it to be so and, therefore, we can call it a signal. Please, see “Signal and Cue—What is the Difference?” at

SMAF Manual

SMAF Cover Page

SMAF Cover Page

Mission SMAF—Bringing Scientific Precision Into Animal Training

This is the SMAF manual, latest update. The SMAF manual is in a way a super concentrated course in animal learning. This is only a booklet, but it will take you time to read and digest. Don’t rush thru it.

We update this manual regularly, sometimes even daily. Come back often to check if there are any new updates. Each manual has a version number (v.YYMMDD).

Changes in this version v.120302:

  • Updated and improved syntax with some new symbols.
  • Updated examples of how to transcribe the teaching of skills in SMAF.
  • New Chapter 4 “Quick guide to designing a POA in SMAF.”

Enjoy your reading!


Click the full-screen view icon for better reading. Zoom in and out as you please.

Related articles


Changes in version v.111023:

  • Definition of command with example, page 10: “While a signal is everything that intentionally changes the behavior of the receiver, a cue is everything that unintentionally changes the behavior of the receiver. A command is a signal that changes the behavior of the receiver in a specific way with no variations or only extremely minor variations.”
  • SMAF syntax corrected, page 22: “25.2. Example 1: Sit,sound + Sit,hand + γSit,treat  ⇒ The dog sits ⇒ “!+sound” + “!-food”.” (The second ⇒ was missing).
  • Minor typos corrected.

Changes in version v.111017:

  • Semi-conditioned reinforcers have their own code.
  • Non-SD has its own code.
  • New code for reinforcers and punishers.
  • Signal and cue slightly redefined.
  • A new POA example (POA example 3).
  • New photos.

Signal and Cue—What is the Difference?

Male Lion (Panthera leo) and Cub eating a Cape...

Secondary sexual traits, as the mane of the male lion, are powerful cues (Image via Wikipedia).

In the behavioral sciences, there is some confusion about the meaning of the terms signal and cue (as with so many other terms) and some authors use it interchangeably. To make it even more difficult, communication theory also uses the same terms with slightly different meanings and in the theatre and movies world a ‘cue’ is actually a ‘signal.’

However, in behavioral sciences, the general consensus (see references below) is that signal and cue have the following meanings.

signal is a perceivable behavior or feature that has evolved and has acquired the specific characteristic of conveying information about the signaler or the signaler’s environment. Information (communication) changes the behavior or the beliefs of the receiver.

This definition of signal implies that if a signal changes the behavior of an organism, this change of behavior must be profitable to both sender and receiver more often than not, or otherwise, signalers would cease to send the signal and receivers would cease to respond. This definition distinguishes, in principle, a signal from coercion, although some signals may be coercive, e.g. threats.

In general, signals must be honest and reliable, or otherwise they cease to have any effect (receivers don’t behave appropriately) and they undermine communication (honest senders will not benefit from sending the signals). However, some signals can tolerate a certain degree of dishonesty, all depending on the costs and benefits for all parties. H. W. Bates discovered in 1861 that some (palatable) butterflies had an advantage in mimicking (Batesian mimicry) poisonous butterflies, which is detrimental to the poisonous butterflies inasmuch as it turns their signals of unpalatability less reliable. On the other side, some species use the same signals to convey the same information and they all benefit from it (Mullerian mimicry).

cue is any feature that an organism can use as a guide to display a particular behavior or series of behaviors. The classical example is the mosquito seeking a mammal to bite and flying up wind when it detects CO2. The CO2 is a cue for the mosquito, but it is surely not a signal sent by the mammal, which would prefer to remain undetected and not be bitten. Intentionality is the key element to distinguish signals from cues.

A cue is a regularity, a pattern that either is permanently ‘on,’ or is ‘on’ and ‘off” depending on specific conditions, e.g. a rock, a tree, or the position of the sun in the sky cues us of directions, and dark clouds cues us of impending rain. The rock, the tree, the sun and the clouds are not there to give us information, but they do if we interpret them correctly. A signal is more malleable, more intentional and we can turn it ‘on’ and ‘off’ in response to relevant cues in the environment, e.g. the warning cry that many species (signal) issue in response to the appearance (cue) of a feared predator.

Cues are traits or actions that benefit the receiver exclusively. The sun and the rock do not profit from us getting our bearings. When a mouse by accident makes a rustling sound in the leaves and attracts a predator increasing the risk of being killed, the sound is a cue for the predator about the location of its prey. When an alert animal deliberately gives a warning call to a stalking predator resulting in the predator giving up the hunt, this sound, the alert call, is a signal both for conspecifics and the predator. Different species can, thus, communicate by means of signals which both recognize and behave accordingly.

Secondary sexual traits are features that distinguish the two sexes of a species, but that are not directly part of the reproductive system. They are probably the product of sexual selection for traits, which give an individual an advantage over its rivals in courtship and competitive interactions. Secondary sexual traits are also cues for the opposite sex. They are not directly related to a better production of offspring, but are normally good indicators of better sperm quality or egg production, e,g, manes of male lions (Panthera leo) and long feathers of male peacocks (Pavo cristatus). In humans, visible secondary sexual traits include enlarged breasts of females and facial hair on males.

The study of signals and cues is more complex that it may appear at first sight. Cues can become signals. In 1952, Niko Tinbergen described ritualization as the evolutionary process whereby a cue may be converted into a signal, e.g. the canine behavior of baring teeth. In 1975, Zahavi described the handicap principle where the reliability of some signals is ensured because they advertise greater costs than absolutely necessary, e.g. the exaggerated plumage of the peacock.

We must understand correctly what the intentionality of signals means and not to confound the intentionality of the signal itself with its origin, development and evolution. Signals do not origin by design with a determined purpose. Some features or behaviors just happen at a certain time to be efficient for an organism in generating in another organisms the right behavior at the right time. If they convey an advantage to these organisms in their struggle for survival (and reproduction), they will spread in the population (provided these organisms reproduce). With time, they gain intentionality and become true signals, but their origin was accidental like everything else. This is the reason why I had to modify (some extensively) the definitions I use in this text and I had to create new ones—to make them compatible with the Darwinian theory of evolution.

Applying the principle of simplicity, as always, I suggest the following definitions:

signal is everything that intentionally changes the behavior of the receiver. A cue is everything that unintentionally changes the behavior of the receiver.

These definitions open for the possibility to better distinguish between the intentional signals (proper signals) we send and the unintentional ones (which are cues). For example, many dog owners say “no” to their dogs meaning “stop what you are doing,” but their (unintentional) body language (cue) says “yes.”

In conclusion: signal is the most correct term to denominate what we use when we communicate with our animals; and signals may assume many forms, auditory (the words we use), visual (the hand movements and body language we use), olfactory (in canine detection work), tactile (a touch, very common in horse training) and probably also palatable.

So, enjoy the consequence of your (intentional) signals and be careful with any cues you may be (inadvertently) sending to your favorite animal. Enjoy as well your further studies of this fascinating topic: animal communication.

Keep smiling!



References and further readings

  • Dawkins, M. S., and T. Guilford (1991). The corruption of honest signalling. Animal Behaviour 41:865–873.
  • Donath, J. (2007). Signals, cues and meaning (February draft for Signals, Truth and Design. MIT Press)
  • Hasson, Oren (1997). Towards a general theory of biological signaling. Journal of Theoretical Biology 185: 139-156.
  • Hauser, Marc D. and Mark Konishi, eds. (1999). The design of animal communication. Cambridge: Bradford/MIT Press.
  • Maynard Smith, John and David Harper (1995). Animal signals: Models and terminology. Journal of Theoretical Biology 177: 305-311.
  • Maynard Smith, John and David Harper (2003). Animal signals. Oxford University Press, UK.
  • McFarland, D. (1999). Animal Behaviour. Pearson Education Limited, UK.
  • Otte, D. (1974). Effects and functions in the evolution of signaling systems. Annual Review of Ecology and Systemat- ics 5:385–417.
  • Saleh, N et al. (2007) Distinguishing signals and cues: bumblebees use general footprints to generate adaptive behaviour at flowers and nest. Arthropod-Plant Interactions, 2007, 1:119–127
  • Schaefer, H. M. and  Braun, J. (2009). Reliable cues and signals of fruit quality are contingent on the habitat in black elder (Sambucus nigra). Ecology, 90(6), 2009, pp. 1564–1573.
  • Searcy, W. A., and S. Nowicki (2005). The evolution of animal communication. Princeton University Press, Princeton, New Jersey, USA.
  • Tinbergen, N. (1952). The curious behavior of the stickleback. Scientific American December 1952.
  • Zahavi, A. (1975). Mate selection: a selection for a handicap. Journal of Theoretical Biology 53:204–214.


Related articles

Commands or Signals, Corrections or Punishers, Praise or Reinforcers

An Australian Shepherd doing agility at the Ro...

Do you use signals or commands to communicate with your dog? (Image via Wikipedia)

Commands or signals, corrections or punishers, praise or reinforcers—does it matter what we call them?

If you think it doesn’t matter, there’s no need to read any further. If you think it does matter, please continue reading because I’d like to help you. I noticed some inconsistencies in contemporary dog training terminology  and will proceed to argue that they need correcting.  Trainers use too many terms that either are badly defined, not defined at all or already exist and mean something else.

Why is it important to agree on one single terminology? Because only then can we have a meaningful discussion and avoid falling out with people with whom we might otherwise like to cooperate.

For example, the majority of “positive” dog trainers have no problems using the word command and yet a command means “an authoritative direction or instruction to do something,” or “a line of code written as part of a computer program.” To command means “to exercise authoritative control or power over.” The word has connotations of the military, the police and of authority in general. Of course, we may use the word command but it beats me why we ban the terms dominance (without defining it properly) and punisher (whilst disregarding the correct, technical definition of the term) and use command with no concern whatsoever.

Personally, I have a problem with the use of command in dog training for several reasons. A command implies the obligation to execute a behavior in a very precise way. We give computers commands to execute actions in exactly the same way every single time, no variations allowed (that’s what we want from our computers). Army officers issue commands they want obeyed with no questions asked and disobedience is severely punished.

Is this what we want from our dogs? No, it is not. We want them to perform a behavior within a particular class of behaviors where variations are both inevitable and acceptable. There are many ways to sit correctly, but not many ways to “copy” or “paste”. The authoritarian aspect also bothers me; it implies subjugation. I don’t want my dog so much to obey me as to understand what I want him to do. The essence of communication is to convey information, not to enforce it. When we communicate, we use signals, and signals are understood, not obeyed and not commanded. I can’t say “I command you to understand.” A command is a signal before becoming a command, only we don’t need to issue commands to our dogs if we’ve done our job properly.

Signal seems to me undoubtedly the right term and has much better connotations than command. Your dog is not a computer, nor a soldier (PS—I have nothing against computers nor soldiers).

The same goes for the term punisher. You’ll find “positive” trainers using the word command without blinking, but demonizing you if you dare as much as whisper the word punisher, which doesn’t make any sense at all to me. If we are sensitive about the connotations of one term, it seems that we would also be sensitive about those of the other. Whilst blithely employing the word command, some trainers substitute punisher with correction, which doesn’t make any difference, it still means the same and is interchangeable with punisher in some senses.

Then, there is praisePraise means “an expression of approval and commendation,” “applaud,” “pay tribute to,” “compliment.” It is true that praise can influence learning in humans, but I doubt it very much that it has any value in animal training. Praise and reinforcer are two different things. We use reinforcers in dog training, not praise. “Positive” animal training claims to be a more humane way of training animals (meaning showing compassion or benevolence), to be more scientific than the “old-fashioned” training, and to know all about “classical conditioning” and “operant conditioning.” If this is true, why don’t we show it and educate people accordingly? Why don’t we use the proper scientific terms?

Some claim that the right scientific terms are too difficult. I fail to see what’s more difficult in reinforcer than in reward, in signal than in command, but even if it were true, this appears to me to be one of the situations where the end justifies the means. It would be a small price to pay in order to gain more clarity and avoid misunderstandings. Using technical terms instead of everyday words would also help people fully understand and use the various learning tools correctly. Sometimes, in trying to simplify things, we miss the point completely. Most dog owners don’t know that praise in dog training means “everything that increases the frequency, intensity and/or duration of a behavior when presented simultaneously or immediately after the behavior takes place” (= reinforcer). Dog owners are not more stupid than dog trainers and if the latter can learn it, so can the former. It’s up to us to motivate them. We were all dog owners before we became dog trainers. Did we like condescending dog trainers back then?

As far as I can see, we only have two options: (1) to claim that it doesn’t matter what we call things, in which case nobody should be labelled for using terms such as punisher, dominance, etc., and we can all be “positive” nevertheless; (2) to use, teach, encourage and propagate the use of correct, well-defined terms, starting with ourselves, in order to be consistent with ideology and methodology.

Personally, I am not worried by the terms you use and I will not label you solely on your choice of words. The only concern I have is that (unless I know you) when you say command I’m not so sure you know about the intricacies of signals, and when you say praise I’m not certain that you fully appreciate the function of reinforcers (and punishers). I understand that you don’t like the word punisher because you are a good person, but I’m not sure that this is the right way to manifest it. Changing a term doesn’t change an attitude. Sometimes, quite on the contrary, if you used the word punisher, you’d have an opportunity and a reason to emphasize that it has nothing to do with violence and abuse.

On the other side, I do have worries that we label good, humane, “positive” dog trainers otherwise because of their correct use of the scientific terms; and that we label good, humane, “old-fashioned” trainers abusive due to their ignorance of the terminology that is fashionable nowadays. Before you even think of labeling me on the basis of my comments here, I would like to remind you that my first book on dog training, published in 1984 and entitled “The Dog, Our Friend—Psychology rather than Power,” was a revolution in dog training at the time; it was the first book (as far as I know) to describe exactly how to teach a dog sit, stand, down, come, heel, jump, slalom, treat-on-the-nose, retrieve, etc., without the use of any force at all. I showed even pictures of the clicker (except that we used a whistle) and of the precursor of these so fashionable toys that you fill with treats to stimulate the dog. Since then, many have followed in the same spirit: respect for the dog as a species and as an individual.

The bottom-line is that we should define terminology and implement it consistently. As it stands now, I’m afraid we’ll lose many good people for our cause of “a better world for dogs and dog owners” because of fashionable trends and pettiness.

Think about it.

Keep smiling,


Unveiling the Myth of Reinforcers and Punishers

Cute Dog

Positive and negative reinforcerspositive and negative punishment— these terms are no doubt familiar to you but the definitions are confusing or you may be unsure of how and when to use each. I shall endeavor to explain. As a biologist and an ethologist, I study and explain such topics irrespective of political correctness, commercial interests, or fashion trends.

Except for reflexes, the behavior of all living creatures changes as a result of its consequences; and there are only two ways in which behavior can change: there can be more of it or less of it. Even what we call new behavior is nothing more than an increase in frequency, intensity and/or duration of components of an individual’s behavioral repertoire. New behavior sometimes amounts to the recombination of well-practiced elements. We may alter its frequency, its intensity, its duration and we may associate it with new stimuli, but if that particular element of behavior is not present in the behavioral repertoire of the animal, it will not be displayed.

As most people know, reinforced behavior tends to increase in frequency, intensity and/or duration and punished behavior tends to do exactly the opposite, i.e. it decreases in frequency, intensity and/or duration.

As most people should also know, a reinforcer is not a reward, like the bonus our boss gives us at the end of the year because he earned a lot of money. A reinforcer is anything that somehow increases a certain behavior. It may not increase the behavior of every individual, or every behavior. A reinforcer is thus only a reinforcer in relation to a specific behavior and a particular individual. It may also work, as reinforcers often do, in circumstances other than those originally envisaged, and on a class of individuals, but this is incidental (an added extra), not a requirement.

In contrast, a punisher tends to decrease the frequency, intensity and/or duration of a behavior. Again, punishers are particular to specific behaviors and individuals, and need not operate on various individuals or behaviors. There is a tendency to relate punishers to violence, but a punisher is simply an aversive, i.e. something, one would like to avoid in a specific context, and does not necessarily have anything to do with violence. For example, I immensely dislike mayonnaise, which implies that any restaurant that serves me a sandwich with mayo will decrease the frequency of my visits to that specific restaurant. Unbeknown to him, the chef is actually punishing me (or rather my behavior of visiting his restaurant). When I open a window and am almost blown away by gale force wind, I hasten to shut it again. The natural elements punished me for my opening-the-window-behavior.

In short, reinforcers and punishers are everywhere and we are exposed to them by simply living and interacting in this world. There’s no way of avoiding them completely. You can learn how to control them, by controlling your behavior, so you are reinforced more often than you are punished if that’s what you want, but even experienced people, wolves, bears, wombats, jellyfish and, of course, dogs, sometimes display behaviors, which are instantly and duly punished. Behavior punished and behavior reinforced—that is how we all learn and it’s a fact of life whether we like it or not.

In terms of learning theory, the scientific definitions of reinforcers and punishers are:

reinforcer is everything that increases the frequency, duration, and/or intensity of a particular behavior when presented or removed simultaneously or immediately after that behavior takes place. Reinforcement is the presentation of a positive reinforcer or the withdrawal of a positive punisher (an aversive).

punisher is everything that decreases the frequency, duration, and/or intensity of a particular behavior when presented or removed simultaneously or immediately after that behavior takes place. Punishment is the presentation of a positive punisher or the withdrawal of a positive reinforcer.

Bottom line: in principle reinforcers and punishers are neither good nor bad, they are not things we like or don’t like, they are just stimuli that either increase or decrease the frequency, intensity and/or duration of a behavior. A reinforcer may be a punisher one day and a reinforcer another, whilst a reinforcer for you may be a punisher for me. Consider the following example: your dog is standing in front of you and you hold a treat in front of his eyes. You look at the dog and you say ‘sit’. The dog doesn’t sit; he just plays around and barks at you. You then, you put your serious face on, emit a grunting sound, and remove the treat. The dog sits and looks as innocent as ever. You hasten to say ‘good’, you get rid of your serious face and present the dog with your friendliest expression of the day, and give the dog the treat you were holding in front of his eyes, the one you removed when he was being silly. This is a situation that I’m sure all dog owners and trainers have experienced countless times. Is there anything wrong with it? Not at all, right? Ok, let’s take a closer look at it. You say ‘sit’, the dog doesn’t sit, and you remove the treat and put on your serious face. The technical term for the removal of the treat is negative punishment and the serious face is a positive punisher. The dog then sits; you remove your serious face and give the dog a ‘good’ and the treat. The removal of your serious face is negative reinforcement and the presentation of ‘good’ and the treat are positive reinforcements. In two seconds you’ve used all four tools (correctly).

Reinforcers and punishers must have the right intensity in order to function. This is a key feature of both. A stimulus of too low an intensity will not increase or decrease a behavior. Hence, such a stimulus is not a reinforcer or punisher. Conversely, a reinforcer with too high an intensity may create another behavior. If it does, it is no longer a reinforcer for the behavior you wanted to reinforce. For example, showing the dog a treat increases its sitting behavior, but if the treat is too good (particularly yummy or the dog is very hungry), it may overexcite the dog and produce jumping up behavior. Equally, a punisher of too high an intensity will not decrease the behavior you want to decrease and instead may produce a completely different behavior. If this is the case, what you thought was a punisher for a particular behavior becomes instead a reinforcer for another, undesired behavior. For example, saying, “stop” to your dog with an unpleasant tone of voice and stern face decreases its barking behavior, but if you shout or become violent, you may produce fleeing or aggressive behavior.

Reinforcers and punishers are stimuli that have a determined window of opportunity and sometimes this window is very narrow. You have to adjust them to the individual animal you are working with, the environmental conditions and the behavior in question. Remember that you never ever reinforce or punish the animal, only its behavior. For example, you still love your dog equally, independently of whether the dog displays a behavior to your liking or not. If it does, fine. If it doesn’t, you’ll have to work a bit more on that.

If you don’t like the terms reinforcer and especially punisher, we can change them.  I once suggested calling them increasers and decreasers, positive reinforcers thus becoming add-on increasers and negative reinforcers turning into take-away increasers. What do you think about using add-on decreasers and take-away decreasers? They certainly don’t have the same connotations as punishers, do they? If you’re a good dog trainer, I’m sure you use these techniques. If we substitute the terms reinforcer, punisher, positive and negative with my suggestions, the famous table for the four operant procedures looks like this:

Operant Behavior Table

So, life is all about learning how to control the consequences of our behavior—and this is a very apt description of our job as dog trainers. We must help our dogs to learn how to control the consequences of their behavior, which is not the same as avoiding them. If, as a rule, we either only reinforce or only punish everything they do, we are indeed doing a poor job, and we are certainly not preparing them for real life where both reinforcers and punishers (increasers and decreasers) are a reality, depending on circumstances and one’s behavior. If you like my table with the alternative names, you’re welcome to use it. A warning though: the terminology doesn’t make any difference to the dog. It may make a difference, however for dog owners and dog trainers with no, or only rudimentary, knowledge of learning theory. I believe it is our duty to educate dog owners and dog trainers to distinguish between the various stimuli and teach them how to use them correctly.

Learning is changing behavior according to its consequences, and as simple as it may seem, it proves undeniably more complicated in a practical learning situation. To be a good animal trainer, or teacher, we need to master the science of learning theory and behavior modification, as well as the art of applying it at the right time, in the right dose, for the right reason. We need to be able to exercise reason and manage our emotions.

Enjoy your training session!



Q. Can I train my dog without punishers at all?

A. Yes, you can, (we have tried it) but it is extremely difficult if not unrealistic. Sooner or later something will disturb you and your dog and you’ll at least have to withdraw the positive reinforcer, which amounts to negative punishment.

Q. Can I at least avoid using positive punishers?

A. Yes you can, but you might not get as reliable a behavior. If there is no consequence for not displaying a behavior and the dog is not interested in what you can offer as reinforcers, there is no reason for the dog to display the behavior you want. It’s up to you, but don’t expect something that you have not taught the dog.

Q. Are punishers bad things?

A. No. Punishers just decrease a behavior. A reinforcer may be a punisher another day and vice versa.

Q. Do punishers hurt?

A. No. Punishers and violence and two different things. Violence may decrease a behavior, but may also result in increased intensity of the same behavior. You should never recur to violence in animal training.

Q. Are reinforcers good for the dog?

A. Reinforcers are neither good nor bad, they simply increase the behavior in one way or another. We presume the dog likes them, but that is not the essence of a reinforcer.

Q. What happens if I use too many punishers, can I hurt my dog?

A. Yes, you can. You should always teach your dog the desired behavior as a first option by reinforcing it. A punisher is a last resort, a necessity, not a choice. Life is not fun if you’re punished all the time (even if it has nothing to do with violence). Punishers inhibit your behavior. Reinforcers enhance your repertoire of behaviors.

Q. What about reinforcers, can I also hurt my dog if I reinforce everything?

A. Yes, you can. Life is not a bed of roses. Your dog must also learn to cope with adversity. It’s all a question of balance and for you as an educator to use the right tool for the right job.

Q. How do I use what we know about reinforcers and punishers to plan my training?

A. When planning your training, you should devise ways of motivating your dog to display the behavior you want and reinforce its behavior. Your dog’s motivation to do what you wish is your most efficient tool. Punishment should only be an emergency measure. If you plan your training properly, you may not need to use punishers at all, which is the optimal strategy.

Further reading

Abrantes R A. 1997. Dog Language – An Encyclopedia of Canine Behavior.

Abrantes R A. 2011. Animal Trainers Handbook (not published yet).

Bailey J. S. & Burch M R. 1999. How Dogs Learn.

Catania C.1975. Learning.

Chance P. 1999. Learning and Behavior (4th ed.)

Dickinson A. 1980. Contemporary Animal Learning Theory.

Donaldson J. 1999. The Culture Clash.

Dunbar I. 1998. How to Teach an Old Dog New Tricks.

Holland J G & Skinner B F. 1961. The Analysis of Behavior.

Lindsay S. 2000. Handbook of Applied Dog Behavior and Learning.

Pryor K. 1999. Don’t Shoot the Dog.

Ramirez K. 1999. Animal training: Successful animal management through positive reinforcement (don’t get fooled by the title, just read the foreword).

Reid P. 1996. Excel-Erated Learning.

McFarland D. 1987. Animal Behaviour.

Wilkes G. 1993. On Target!
Related articles