In our Western culture, the relation between humans and dogs is played out in a historical and socio-economic context that fosters the emergence of behavioral dysfunctions in animals (the discrepancy between the imagined dog and reality). Many behavioral problems in dogs arise from a failure to recognize social and environmental constraints during their growth. In this article we shall briefly trace these phases of a dog’s social and behavioral ontogeny and epigenesis. We shall also point out the risk factors that can undermine the harmonious interaction between humans and their dogs.
The main phases in neurological development
Like humans, dogs belong to a species that matures slowly after birth: the new-born is not completely developed and is incapable of surviving on its own. This implies a structured and caring parental environment (caring for the young), reflexes that orient the young puppy to its parents, and the existence of optimal, even crucial, periods in the development of the animal’s nervous system.
- The growth of the nervous system underlies behavioral epigenesis. The immaturity of the nervous system at birth is obvious: Cragg (1975) calculated that in cats the number of synapses per cortical neuron grows from a few hundred to nearly 12,000 in the 10th to 35th day after birth (in Changeux, 1983). Various measurements (volume, weight, percentage of dry matter, oxygen consumption) of the brain show that growth is rapid until the 6-7th week when development suddenly slackens considerably. The number of brain cells and their myelination reaches full adult maturity at 4 weeks. It is worth mentioning that the brain is totally unmyelinated at birth, except for the trigeminal nerve and the non-acoustic part of the auditory nerve which correspond to the new-born’s orientation reflexes (Herman 1958, in Scott & Fuller 1965). The motor cortex is the more developed at birth. The occipital cortex, however, then grows more rapidly than the motor and frontal lobes; it also contains several immature neuroblasts that reach full development only around 3 weeks of age (Fox, 1965).
- Behavioural epigenesis (ethogenesis) is linked to the way neuronal connections are organised (theory of selective stabilization). The development of the neuronal networks is a characteristic process: “the phase of synaptic redundancy followed by a phase of regression in the axonal and dendritic branches is a critical period of development… The redundancy is temporary. Active nerve endings are eliminated all the while the nervous system itself is expanding… This neuronal hecatomb is part of the normal development. … The hypothesis that spontaneous, and later evoked, nervous activity contributes to the development of neural networks and synapses appears to be plausible” (Changeux, 1983).
- Behavioural epigenesis is influenced by
environmental factors, by the
surroundings. Activity regulates
neuronal development. In a now classic
experiment (by Weisel and Hubel from 1963, in
Changeux 1983) with monkeys, noticeable visual
defects were caused when one eye was sewn shut
during the first six weeks of life; the
problems were reversible if the eye was
re-opened after three weeks’ time. The same
experiment on adult monkeys showed no effect
on vision. Similar experiments on cats show
there is a sensitive period for visual
development between 3 and 7 weeks of age, and
incapacity to recover vision after three
months (Weisel and Hubel, in Vastrade, 1987).
“There is a critical period during which the
abnormal functioning of a system causes
irreversible lesions.” (Changeux, 1983).
According to Klosovskii (1963), puppies and
kittens that undergo periods of forced
rotation for several days have vestibular
neurons that are larger than those of animals
that have not received this stimulus (in Fox:
*In rodents, postnatal temporary occlusion of the ears leads to subsequent difficulties to locate sounds in space and to reduction of discrimination of auditive patterns (Caston: 1993). In rodents always, precocious exposition to other species odors eases future interspecific socialization (decrease in aggressions, lowering of corticosteroïds hormones) (Caston: 1993).
* This reflects Cyrulnik’s (1991) remarks that the brain becomes atrophied when [an animal] is raised in sensorial isolation, and it develops more than average in an environment of hyperstimulation in noise, affectivity, odors, tastes, sight, etc….
- Neurobiological studies have showed that prolonged precocious isolation was responsible for long-lasting structural or functional cerebral modifications. Isolation leads to a diminution of the dendritic network in the monkey frontal cortex; it also induces a reversible diminution of the activity of the mesocorticofrontal dopaminergic pathways (with hyperreactivity to stress), associated with a slight increase of the activity of the mesolimbic and nigrostriatal dopaminergic pathways (Verdoux and Bourgeois: 1991).
- Development thus seems to come about in stages; although these stages are possibly nothing more than a “simplified classification system where the classifier traces a straight line through a continuum” (Bateson, 1981). Pampiglione (1963) observed marked changes in EEG patterns at 7-8 days, 5-6 weeks (reaching adult levels), and 4-5 months (Fox, 1965). According to Charles and Fuller (1956) (in Scott and Fuller: 1965) the alpha rhythm that appears at 21 days signals an activation of the sense of sight. Scott (1958, 1962) (in Scott and Fuller: 1965, and Fox: 1965) speaks of several stages of neurological, reflexive and behavioral development that are particularly didactic: neonatal (0-14 days), transitional (14-21 days, starting when the eyes open and ending when the animal starts on hearing a noise), socialization (21-70 days) and juvenile (70 days and older). These periods overlap considerably. Since these stages are still used in the literature, they are worth mentioning. We recommend reading Vastrade (1986), Markwell & Thorne (1987), and Nott (1992) for an overview, or Scott and Fuller (1965) or Fox (1965) for a more in-depth study.
In conclusion: behavioral patterns develop over successive phases, according to internal and external factors that interact in a complex and continuous manner. As Cyrulnik wrote, “the World of each animal is built around the double constraint of genetics and development”.
The concept of sensitive periods
Bateson (1981) has described the developing individual as a train with its windows closed – at a certain point (maturing) the windows open and the traveler is encouraged to study the information passing by outside. Depending on the information presented (learning), he/she either continues (motivation) or stops (habituation, impregnation, or self-limitation) looking out the open window. In other cases the windows close when a new point is reached.
This notion of learning in phases has various names: sensitive period, critical moment, optimal period, vulnerable point, crucial stage, susceptible period, and so on.
A sensitive period is a point in the maturing process when events are susceptible to leaving long-term effects, or a period when learning is easier and knowledge gained is stored in the long-term memory. During the sensitive period, a small number of determining experiences have major effects (or damages) on future behavior. The sensitive period is preceded and followed by periods of lower sensitivity, and the transition is gradual.
The notion of sensitive period is used in the place of critical period because the former extends over a longer period of time. Ducklings become attached to their mother between the 13th and 16th hour of life (Hess 1959, in Cyrulnik 1989), it takes 5 minutes of contact during the first hour after birth for a she-goat to become attached to the odor of her kid (Bateson, 1981), and a ewe needs contact within 4 hours after the birth of her lamb. Without this contact the mother will reject her young in the last two cases (Collias, 1956, in Scott and Fuller, 1965). These very short periods justify the term critical. Since puppies do not have such short periods of facilitated learning, we will use the term sensitive period.
I was one of the people who helped spread this concept in French-speaking countries (Dehasse and De Buyser: 1983, 1989, 1991) by emphasizing on several occasions that the sensitive period in the behavioral epigenesis in puppies extended from 3 weeks to 3 months of age. The duration of this sensitive period had to be verified by delving into the literature on experiments in this realm and clinical review.
The prenatal period
Pregnant rats that have been placed under stress or injected with ACTH or adrenaline give birth to young rats that are emotive and perform less well than a control group. (These young rats are raised by another mother that has not been placed under stress to preclude the possibility of postnatal maternal influence.) (Fox, 1978)
In a similar vein, when a pregnant animal is petted her litter is more docile (Denenberg and Whimbey 1963, in Fox 1978). This effect, called the “gentling”, “petting” or “caress” effect, can be prolonged by caresses to the new-born. According to Fox (1975, in Fox 1978) this activates the parasympathetic system, facilitating relaxation, digestion and emotional attachment, and thus socialization as well. Experiments by Cyrulnik with cats have shown that attachment depends on the cholinergic system; anti-cholinergics block the attachment process. The object of attachment is a being whose presence soothes and whose absence causes distress, who possess the signs of familiarization; a “reference being” (Eibl-Eibesfeldt 1984). This is probably linked to the social species’ innate need for contact.
A dog’s tactile capacities develop before birth, and it is possible that it already becomes used to contact in the uterus, when the mother is petted. Puppies manipulated this way show a greater tolerance to touching than dogs born of a mother who was not petted.
In rats, once again, manipulation (contact, exposure to cold, etc.) at a young age or before birth (manipulation of the pregnant mother) gives greater resistance to stress (cold, hunger) and disease (implanted tumors). This phenotypical effect is transmitted non-genetically for several generations (Denenberg and Rosenberg, in Fox 1978).
These experiments enable us to deduce that when a gestating pet is given a friendly and caring human environment (with affectionate physical contact), the domestication and emotional balance of her offspring is facilitated, as compared with an environment where there is no contact and interaction with people.
The neonatal period
We will only say a few words about this period, which arbitrarily ranges from birth to the opening of the eyes at approximately 13 days.
Superficial, limited observation of the new-born puppy could lead one to believe it did not even belong to the canine species: awkward, dragging itself around, oriented to contact, the mother’s teats and the smell of milk, yapping in distress when isolated, cold, hungry or in pain, and having only a limited capacity to keep itself warm and to learn. The new-born puppy is a completely dependent being, and apparently hardly influenceable psychologically in classical conditioning tests. As such this phase holds only minor interest in our study.
It is possible, however, that the future holds surprising discoveries about the epigenetic importance of this neonatal period, especially as concerns the “manipulation” effect on neuro-hormonal development.
The identification phase
At birth the puppy has not an innate recognition of members of its species; in a way it does not know it is a dog. This must be learned!
Through species identification a puppy is able to recognize its parents (filial imprinting), and develop preferential intraspecific social relations (fraternal imprinting) and the relations (sexual imprinting) which mean the survival of the species (filial and sexual imprinting). An animal that is badly imprinted is lost for the species.
Here are a few examples:
Christy, a female puppy, was raised in complete isolation from other dogs by the “colony” of students at the Jackson Laboratory. At 9 weeks she was introduced to other dogs: the adults growled at her but showed no further signs of aggression and the puppies (litter-mates) began to play-fight and she responded. In 4 days time her behavior was indistinguishable from that of the other puppies (Scott and Fuller, 1965).
Note that, unlike goats and sheep, adult dogs show no parental rejection of their young!
A fox terrier puppy (male) raised in complete isolation and introduced to other dogs at 16 weeks displayed inhibited behavior and was attacked by the other puppies that were normally socialized. He was placed with other dogs also raised in isolation; the dogs lived alongside each other, without aggression but also without interaction (Fisher, 1955 in Scott and Fuller, 1965). Dogs raised in isolation and placed in contact with others of their species at 16 weeks are attacked and rejected. When the experimenters mime play-fights against these same dogs, they are able to recover a positive dog-dog interaction and complete integration in the same pack within a few days (Fuller, 1961 in Scott and Fuller, 1965).
Male chihuahuas raised by cats until 16 weeks of age demonstrate preference for the presence of cats, and submission – or fear – in the presence of dogs (they also show no reaction to their reflection in a mirror). When they are placed with other dogs at 16 weeks, they recover intraspecies socialization in two weeks; they now prefer dogs to cats and react to images of themselves in a mirror (Fox, 1971, in Pieters 1984).
On the other hand, puppies raised in a family from 4 weeks of age (becoming used to dogs, cats and/or children), without renewed contact with the laboratory dogs, show greater familiarity with people than with dogs. An adult sheltie (who had lived with a cat and two children) showed sexual attraction for the cat and attacked all dogs (male and female alike); a beagle became “attached” to a vacuum cleaner bag; a basenji (who lived with a female dog) became a delinquent stray who attacked other dogs (Scott and Fuller, 1965).
Clinical practice shows that when a puppy is acquired at 6 weeks this is already a handicap in developing its adult social and sexual preferences.
We should also mention that the first signs of humping (pre-imitation of future sexual behavior) appear as early as 3 to 4 weeks (Scott and Fuller, 1965).
This behavior is provoked by pressing on the sternum or the stomach. It is possible that this is a factor in sexual imprinting, but it has yet to be proven.
To my knowledge, no statistical studies have been made on dogs raised in isolation, covering a broad range of breeds (for ethical reasons?), which means that crucial experimental data are lacking. Our knowledge is partially extrapolated from ethology studies in birds. Among birds, imprinting lasts throughout parental care and this period is shortened when there is a danger of mixing species. Preference is given to visual and auditory imprinting whose effects last almost a whole lifetime. With mixed imprinting, there is a preference (innate predisposition?) for one’s own species over a neighboring species, and for this species over a more distant one (such as humans).
In conclusion, species identification (filial, fraternal and sexual imprinting) is acquired during a sensitive phase of development, and depends on “play-fighting” among puppies (litter-mates). This begins about the third (3±½) week and ends somewhere between 11 and 17 weeks (12±5), when the dogs loose their ability to play with unfamiliar dogs and become “serious” in defending their group. In the absence of siblings, a puppy establishes identification through care-giving, care-searching and/or playful interaction with its parents or other dogs. This interaction must last until at least, if not beyond, the 6th week. The presence of other species during this period does not hamper identification with one’s own species.
The end of this phase varies depending on factors that are internal (breed, line of descendants, individual) and external (behavior of the mother, other dogs, quality of the surroundings). A stressful environment (feral dog) will close this phase ahead of time (probably around 7 to 9 weeks).
This type of learning presents several characteristics:
- it is stable, rigid and persistent (sometimes for life);
- it is easily acquired;
- sexual imprinting occurs on supra-individual and supra-breed characteristics, which permits species generalization;
- filial imprinting (attachment) seems to be more discriminating and is limited to parents;
- fraternal imprinting is the basis of sociability;
- attachment is an interactive process.
They are similar to those found among birds.
The total absence of other dogs (own species) between 3 and 12±5 weeks fosters identification with another species that is closest (in general humans, but occasionally cats, rabbits, etc.) or an appropriate substitute (stuffed animal, vacuum cleaner bag, etc.). This identification is persistent, occasionally for life. In adults this leads to:
Courting behavior and attempts to copulate with the identification species (despite activation by pheromones of one’s own species), no behavior of this type or else awkward attempts with a sexual partner of the same species,
Social preference for the identification species,
Rejection (flight or fight) of one’s own species (including mirror images).
The relative absence of other dogs between 3 and 12±5 weeks leads to relative, total or no handicaps depending on circumstances:
possible recovery of the dog’s species identity at 9 weeks when it plays with other puppies,
attachment to the identification species and disinterest or aggression towards canines, despite the (almost) normal capacity to reproduce, etc.
The imprinting effects of a mirror placed in the surroundings of a puppy isolated from other puppies have not been studied (to my knowledge). Since no interaction is possible with a mirror image, this seems to be a poor substitute for suitable “imprinting”.
The socialization-domestication phase
A puppy is not programmed to interact socially with another species. Twelve thousand years of domestication, however, has shown that this is possible. The dog’s particular nature – a puppy has to learn to identify its own species – can serve to foster socialization with other species (called domestication when it involves interaction with humans).
Let’s look at a few experimental cases:
Puppies raised in a semi-open environment in (nearly) complete isolation from humans reacted differently towards an active unfamiliar observer depending on their age. Each puppy was taken from the surrounding in which it was raised, placed in contact with humans for one week, and again tested. Fear in the presence of a human that handled him decreased from 3 to 5 weeks, was minimal at 5 weeks, then increased again afterwards. “Recovery” (improvement or disappearance of fear) after a week of interaction-socialization was more efficient at 3 weeks; it was roughly the same at 5, 7 and 9 weeks (Freedman, King, Elliott, 1961, in Scott and Fuller, 1965).
A puppy – raised in the same type of surroundings – was placed 10 minutes a day with a passive observer, calmly sitting in the room and paying no attention to the dog (Scott and Fuller, 1965):
- at 3 to 5 weeks, the puppy investigated the observer openly;
- at 7 weeks, it took 2 days before it investigated (2 10-minute sessions)
- at 9 weeks, 3 days;
- at 14 weeks it no longer investigated the observer.
At 12 weeks a puppy is easily frightened. Confinement and hand feeding enable it to accept contact with its laboratory handler(s) but not with strangers, and it still prefers the presence of dogs to that of humans (Scott and Fuller, 1965).
This fearful reaction has been found in all breeds tested. When put on the defensive a cocker’s bite is “softer” than that of other breeds tested (basenji, terrier, beagle, sheltie).
According to Fuller (1961), puppies raised in isolation in a laboratory develop adequate socialization to humans if they receive two 20-minute periods of human contact per week. This short contact, however, is not enough for basenji puppies; this variability is thus truly linked to breed (genetics) (Scott and Fuller, 1965).
In conclusion, puppies demonstrate an investigation-attraction behavior towards the unfamiliar as soon as they are able to express this attraction (almost adult motor capacity), in other words at 3±½ weeks. This attraction subsides in an almost linear manner after the fifth week until at least 9 weeks. The attraction recedes under the influence of fear of the unknown behavior which grows after 5 weeks; the puppy “recovers” from its initial fearful reaction instantaneously from 3 to 5 weeks (investigation behavior effect), and then it remains wary for longer periods as it grows older. At 12 weeks socialization requires active manipulation (mimicking play-fights), at 14 weeks socialization seems to be impossible.
In birds fear of the unknown is delayed when they are raised in isolation; this phenomenon thus appears to depend on experience rather than maturing of the nervous system (McFarland, 1981) - one must first be able to refer to something “known” before fearing the unknown.
An arbitrary limit can thus be set for spontaneous socialization to another species, during a first encounter, at 12 ± 2 weeks. Nothing, however, enables us to affirm or deny that rapid habituation to close stimuli cannot be achieved after 12 weeks.
Interspecies socialization (attachment) does not
have the same characteristics as species
It is easily acquired but requires permanent reinforcement to avoid de-socialization;
it is not generalized (generalisable) to all individuals of the species concerned, but remains relatively limited to the individual’s characteristics. It is thus infra-species: it is a “type” socialization (human: man, woman, adolescent, child, baby, black, white, with/without beard, hat, white apron, etc.). The capacity to generalize varies from one species to another (dog and wolf, more than coyotes), breed (watchdogs less than other dogs, according to Fox, 1978), the family line and individual (no statistical studies available).
the threshold of socialization (number of interactions) is variable and depends on factors that are internal (breed, individual) or external (mother’s fearful behavior, quality of the surroundings, etc.).
Domestication depends on the presence of humans between 3 and 12 ± 2 weeks in the surroundings in which a puppy develops and this socialization must be continued throughout the animal’s life. The lack of human contact between 3 and 12 ± 2 weeks fosters the development of fear/wariness of humans (feral dog).
The relative absence of human contact leads to relative handicaps, such as fear/wariness/phobia towards a type of human (children, men,…).
The interactive presence of different types of humans between 3 and 12 ± 2 weeks facilitates a puppy’s generalized socialization to humans.
The interactive presence of other animals leads to interspecific socialization and attachment, and it counters predatory behavior.
Interspecific socialization counters predatory behavior towards the type of attachment individual.