Behavioural Processes 76 (2007) 1–13
Review
Equine learning behaviour
Jack Murphy ∗ , Sean Arkins
Department of Life Sciences, University of Limerick, Limerick, Ireland
Received 14 June 2006; accepted 15 June 2006
Abstract
Scientists and equestrians continually seek to achieve a clearer understanding of equine learning behaviour and its implications for training.
Behavioural and learning processes in the horse are likely to influence not only equine athletic success but also the usefulness of the horse as a
domesticated species. However given the status and commercial importance of the animal, equine learning behaviour has received only limited
investigation. Indeed most experimental studies on equine cognitive function to date have addressed behaviour, learning and conceptualisation
processes at a moderately basic cognitive level compared to studies in other species. It is however, likely that the horses with the greatest ability to
learn and form/understand concepts are those, which are better equipped to succeed in terms of the human–horse relationship and the contemporary
training environment. Within equitation generally, interpretation of the behavioural processes and training of the desired responses in the horse
are normally attempted using negative reinforcement strategies. On the other hand, experimental designs to actually induce and/or measure equine
learning rely almost exclusively on primary positive reinforcement regimes. Employing two such different approaches may complicate interpretation
and lead to difficulties in identifying problematic or undesirable behaviours in the horse. The visual system provides the horse with direct access to
immediate environmental stimuli that affect behaviour but vision in the horse is of yet not fully investigated or understood. Further investigations
of the equine visual system will benefit our understanding of equine perception, cognitive function and the subsequent link with learning and
training. More detailed comparative investigations of feral or free-ranging and domestic horses may provide useful evidence of attention, stress and
motivational issues affecting behavioural and learning processes in the horse. The challenge for scientists is, as always, to design and commission
experiments that will investigate and provide insight into these processes in a manner that withstands scientific scrutiny.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Horse; Behaviour; Learning; Processes; Memory
Contents
1.
2.
3.
4.
5.
6.
7.
8.
9.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Learning processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Learning ability and intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.1. Comparative studies of animal intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Memory in the horse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Experimental task learning and behaviour in the horse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.1. Experimental testing of equine learning and behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5.2. Horses experience learning difficulties due to temporal delays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5.3. Maze learning trials in the horse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5.4. Sidedness could influence learning and behaviour in the horse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Social and observational learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Attempts at assessing higher order cognition in horses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Learning and behaviour in the feral horse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Contemporary training schemes and equine learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
∗
Corresponding author. Present address: School of Agriculture, Food Science and Veterinary Medicine, University College Dublin, 206 Veterinary Sciences
Building, Belfield, Dublin 4, Ireland. Tel.: +353 87 284 3070; fax: +353 1 7166104.
E-mail address: Jack.Murphy@ucd.ie (J. Murphy).
0376-6357/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.beproc.2006.06.009
2
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
10.
Cellular and molecular basis of equine learning behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction
The horse has evolved through domestication to adapt to man
and the environment he provides (Price, 1999). Adaptation to
domestication in any of the animal species has been largely
dependant on the degree of developmental plasticity of the animal and the typical behavioural and learning patterns compatible
with the husbandry techniques utilised during the domestication
process (Price, 1999). Neurobiological and behavioural factors
influence several aspects of equine learning and ultimately athletic ability and success in the horse (Visser et al., 2003). It is
likely that the horses with the greatest ability to understand or
conceptualise are those, which are better equipped to deal with
the demands of contemporary and future training schemes. There
is also general consensus among scientists and lay practitioners
that equine training regimes and welfare programmes associated
with the horse should continually strive to match the complexity, levels of comprehension and learning intensity that is innate
to the horse. However, even with the importance attributed to
the role of the horse in human society, it has been reported that
surprisingly little scientific research has addressed the issue of
equine learning and its implications (Nicol, 2002).
Historically, difficulties have arisen with the elucidation of
many equine learning and behavioural processes and some
notable inconsistencies regarding terminology and the interpretation of subsequent equine behaviour have been reported
(Mills, 1998a,b). In an attempt to appraise and categorise equine
behavioural terminology and its meanings, this inconsistency
in terminology issue has been recently addressed and as a
result more helpful ethograms detailing inventories of specific
behaviours in the horse have been completed (McDonnell and
Poulin, 2002). Nevertheless, even following further attempts at
producing unequivocal descriptive terminology and definitive
analysis of equine behaviour and learning processes, some difficulties with interpretation still remain (McGreevy et al., 2005).
This situation may lead to a lack of rigor in attempting to identify
and control problem behaviours in the horse. However, equine
ethology and investigation of equine behaviour under experimental conditions are subject areas of research that are currently
becoming more popular under a number of general categories.
The areas of interest that have been targeted for detailed investigative research in the horse include: learning, training, feral
behaviour, stereotypies, breeding behaviour and temperament
assessment (Houpt and Rudman, 2002).
The primary goal, for those interested in equine behavioural
and learning processes in the horse, and how this affects
human–horse relationships, should be to maximise the potential benefits for both man and animal. One of the very earliest
acknowledged authorities, Xenophon, ca. 400 bc declared that
‘what we need is that the horse should of its ‘own accord’ exhibit
his finest airs and paces at set signals. . . such are the horses on
10
10
10
which gods and heroes ride’ (Rees, 1997). While Xenophon
referred to the outward expression of athleticism in the horse,
one inference is that the horse would also learn and perhaps
more importantly understand the signals involved in requesting
such behavioural demonstrations in its association with man.
Whereas human interaction with the horse and the domestication process have been of enormous benefit to the horse in
terms of veterinary care, protection and survival, some potential
disadvantages and conflicting practices have also developed in
tandem. Because of mans’ often insensitive selection techniques
and modern training regimes, the resulting social isolation and
the restricted breeding opportunities have regularly been at variance with the evolutionary processes of the ancestors of the
modern horse (Goodwin, 1999).
A more detailed understanding of these conflicting practices
would help to promote improved equine management interactions and in so doing would likely maximise man’s appreciation
of behavioural and learning processes in the horse. Humans have
regularly attempted to reinforce dominance strategies on the
horses in their care in an attempt to elicit the desired outcomes
and responses from the animals (Creigier, 1987). This may be
a misguided strategy given that the natural equine response to
dominance is likely to be one of avoidance and it has recently
been shown that training is actually enhanced when the training methods employed exactly match the mental ability of the
horse (McLean and McGreevy, 2004). While their methods
may not always have been based on scientific research, some
informed trainers have highlighted the importance of a better understanding and appreciation of equine behavioural and
learning processes (Roberts, 1996). Given this raised awareness
and apparent benefit, it is likely that learning behaviour and
the horse–human relationship might be aptly modified with the
imposition of a better balanced social interaction between horse
and human (Goodwin, 1999).
Several other conditions affecting equine learning behaviour
have been reported to induce fearfulness in the horse including isolation from conspecifics, exposure to novel objects or
novel conditions and, under certain circumstances, proximity to
humans (Lansade et al., 2004). It has been reported that early
handling has particularly positive behavioural effects in animals
and it has been shown to reduce animals’ fear of humans, while
high levels of fearfulness have certainly been shown to impair
learning ability in the horse (Fiske and Potter, 1979). Although
foal imprint training has been promoted in the equine industry,
there are only limited documented scientific studies available
regarding this form of training or its efficacy.
In one such study Williams et al. (2002) actually concluded
that there was no difference between foals at three months of
age between controls (foals on pasture without training) and
trained foals (following a three month programme) and therefore, imprint training appeared to have limited effect on the foals
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
as a result. In a longitudinal study, with respect to early training
on the jumping technique of horses, Santamaria et al. (2005)
concluded that specific training for jumping at an early age
was unnecessary because effects on both technique and jumping capacity was only temporary in nature. Nonetheless further
scientific evaluation of handling, apparent beneficial effects of
training and interpretation of learning processes in the horse is
necessary and will be welcomed.
2. Learning processes
Within the current human–horse relationship, the horse is primarily involved in a wide variety of sporting and leisure time
activities (Lansade et al., 2004), and is the basis for an industry of considerable commercial importance (Giulotto, 2001).
Curiously however, given the long association between man
and horse, there currently exists only minimal published scientific research with regard to equine learning and comprehension
(Nicol, 2002). Learning has been described as changes in an
animal’s behaviour resulting from experience of some condition or set of circumstances (Tarpey, 1975; Chance, 1993).
However, while learning represents modification to the internal
behavioural organisation of any species, the process depends
on the reinforcing properties or experience of the species’ environment (Domjan and Burkhard, 1986). Additionally, learning
can be described as either an active or a passive occurrence,
and working descriptions of what learning actually is vary only
slightly.
In any event, the processes of both active and passive leaning behaviours always involve experience – the experience of
learning some phenomenon. Mackintosh (1983) reported that
learning could be sub-divided into three broad forms that might
be considered important in relation to general training procedures. The broad forms include: (i) non-associative learning,
typically habituation and sensitisation, (ii) associative learning
or conditioning and (iii) complex learning or insight. It has also
been generally accepted from a psychological perspective, that
learning typically follows a series of incremental stages of (a)
exposure to a stimulus, (b) acquisition of a response behaviour,
(c) fluency, (d) generalisation and (e) subsequent maintenance
of the learned response with sustained reliability even under
various settings.
In terms of learning behaviour, no evidence has as of yet been
produced to suggest that horses actually learn any differently
than do any other species (Mills, 1998a,b). Learning behaviour in
any species is also critically influenced by the timing of exposure
to the stimulus and introduction of the associated reinforcement
strategy. Experienced handlers have been astutely aware of the
necessity to apply reinforcement schedules immediately or as
close as possible to the demonstration of the desired behaviour
in the horse for optimal effect. This concept is also the basis
for Pavlovian (classical) and operant (instrumental) learning
(Bouton and Swartzentruber, 1991). Operant learning or conditioning is a training technique employed within several aspects
of equestrianism (Cooper, 1998). Scientific research in this area
of equine learning is relatively sparse to date, but it is certainly
warranted (Miyashita et al., 2000; McLean, 2001; Williams et
3
al., 2004). Training and subsequent learning in the horse are particularly aggravated by delayed, conflicting or meaningless cues
and reinforcements. Hull (1943) showed that the application of
two intensive stimuli simultaneously would result in ‘blocking’,
where neither correct response would be learned. This is an
important issue regarding learning and training for the horse as
Wiepkema (1987) has indicated that conflict behaviours such as
ambivalent, redirected and displacement behaviour result from
unpredictability in the stimulus–response relationship employed
during animal training.
With regard to the efficacy of training, repetitions, temporal distribution and duration of training schedules and exposure
to the test stimuli have been investigated experimentally in the
horse. In general, the findings of such experimental trials have
concluded that extended sessions of concentrated training schedules produces inappropriate and inefficient learning behaviour
in the horse (Rubin et al., 1980; McCall, 1990; Sappington
et al., 1997). Indeed, following a substantive review, Nicol
(2002) reported that there was poor correlation between learning
behaviour in individual horses and the subsequent performance
of the same horses during different experimental tasks. At best it
appeared that learning behaviour was a function of the individual
horse and any correlation with performance levels in subsequent
experimental trials was very much dependant upon the specific
task involved. This is interesting, particularly when considering
earlier work suggesting that behavioural and learning characteristics observed in foals were not only heritable but that the
subsequent performance features of the animals could be predicted from observations of pre-weaning behaviour of the foals
(Wolff and Hausberger, 1994). The inference from the Wolff
and Hausberger (1994) study was that progeny of certain sires
appeared to exhibit similar behavioural displays in their play patterns and other spatial interactions with their dams. With regard
to dealing with young horses, this might be important when
training similarly bred horses and predicting how they might
learn and react to certain stimuli.
3. Learning ability and intelligence
There is enormous difficulty in assessing intelligence levels
within and between all animals, primarily because of the difficulties in asserting what actually constitutes intelligence per se.
Many of the earliest attempts at comparative psychology have
postulated that intelligence and learning behaviour were intrinsically linked to or based on a ‘scalae naturae’ or so called ladder
of life (Hodos and Campbell, 1969). The ladder system placed
species in a hierarchical order with humans at the top of the ladder in terms of intelligence and learning ability and transcended
downwards through a level including apes, monkeys and dolphins to further groupings of dogs, cats rat, birds, reptiles, fish
and amphibians and finally leading to a basal level of insects
(Linnaeus, 1758).
However, numerous attempts at applying this ‘order format’
to animal learning ability and intelligence assessment across
species have received repeated criticism to such an extent that it
has been labelled as no longer valid (Hodos and Campbell, 1969;
Houpt, 1979; Mackintosh, 1988). The difficulty lies not least in
4
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
Table 1
Hierarchy of learning abilities
Level
Hierarchy of learning abilities (adapted from Thomas, 1986)
(1) Habituation
(2) Classical conditioning
(3) Simple operant conditioning
(4) Chaining operant responses
(5) Concurrent discriminations
(6) Concept learning
(7) Conjunctive, disjunctive and conditional concepts
(8) Biconditional concepts
Description
Learning not to respond to a repeated stimulus that has no consequences
Making reflex responses to a new stimulus that has been repeatedly paired with the original
innate stimulus
Learning to repeat a voluntary response to obtain reinforcement
Learning a connected sequence of operant responses to obtain reinforcement
Learning to make an operant response to only one set of stimuli for more than one set of stimuli
concurrently applied
Discrimination learning based on some common characteristic shared by a number of stimuli
Learning of concept involving a relationship between stimuli of the forms ‘A and B’, ‘A or B’,
and ‘If A then B’, respectively
Learning of concept involving complex logical reasoning, such as ‘A if and only if B’
designing an appropriate experimental situation or trial, to test a
range of species in a manner, whereby the experimental design
does not bias the result in favour of one or more of the species
under investigation. However, Thomas (1986) produced an interesting hierarchy of learning skills ranging from purely basic
habituation to complex logical reasoning. This matrix pointed
towards an index of intelligence levels by determining relative
position within the hierarchy of the learning skills at which an
animal was capable of performing (Table 1).
Thomas (1986) compiled results based on experimental trials demonstrating the ability for concurrent discrimination in
various fish, reptiles, birds, and mammals, including mice, rats,
zebras, donkeys, horses and elephants. The results appeared to
suggest that of all the species tested; only the elephant was
capable of successfully completing as many concurrent discrimination tasks as the horse (level 5; Table 1). Earlier equine
learning research had shown that basic discrimination, memory
and learning behaviour in the horse was very good (Giebel, 1958;
Dixon, 1970; Houpt, 1979). On the other hand Sappington et al.
(1997) concluded that equine learning between simple discrimination tasks was poorly correlated. In addition, it has also been
reported that performance under one set of experimental conditions were not necessarily predictive of similar performance
levels involving different experimental conditions (Sappington
et al., 1997; Nicol, 2002).
3.1. Comparative studies of animal intelligence
Davis and Cheeke (1998) declared intelligence and learning ability in the horse relative to other species was the subject
of some considerable speculation, particularly on an anecdotal level. While it has been reported that horses are less than
intellectual giants among domestic animals, they have been
selected not only for muscle mass and speed but also specifically for trainability (Houpt, 1979). Trainability and intellect
are of considerable importance and are highly influential when
considering the requirement for horses to respond to subtle stimuli in various disciplines of equitation (Visser et al., 2003). One
of the most famous examples of highly intelligent equine responsiveness or perception was undoubtedly the early 20th century
German horse Clever Hans, whose owner actually believed that
the horse was capable of solving various mathematical problems
by numerically tapping out correct answers with a fore-limb.
Rather than possess a proficiency in the calculations, the horse
had very astutely learned to interpret extremely subtle stimuli
from unsuspecting but participating (and anticipating) audiences (Budiansky, 1997). Further cross species comparisons
have indicated that horses were capable of making better discriminations than sheep, zebras or donkeys. Furthermore, while
horses remembered what they learned, though not as well as
cows, they learned to avoid pain by running or jumping and
were faster to achieve this than pigs, but not as quick as dogs
(Houpt, 1979).
In general, horses tend to perform poorly in tests based on
food tasting and food aversion. While rats and pigs, easily
learned to refrain from eating sweet feeds, which had poisonous
consequences, horses did not seem to have this learning ability
(Houpt, 1979). Although their consummatory behaviour is often
based on large infrequent meals, carnivores also have the ability
to form food and/or taste aversions. In one study where illness
was associated with the consumption of a relatively novel feed
(induced by apomorphine administration immediately following
consumption) ponies demonstrated the ability to form an aversion to a novel feed under some conditions (Houpt et al., 1990).
However, during trials where apomorphine administration was
delayed for 30 min following ingestion of the feed, the ponies
were unable to associate the illness with the consumption of
discrete feeds (Houpt et al., 1990). As a consequence horses are
likely to be exposed to poisonous and toxic challenge because
of the inability to learn specific feed aversions in situations
where illness is delayed. This is particularly the case following
consumption of long-acting toxins associated with some plant
species such as Senecio and Equistrium (Oehme, 1987). This
apparent learning difficulty could be due to differences in feeding behaviour in the species. Horses are natural trickle feeders
and as such, may find it more difficult than some other species to
distinguish discrete feeds, which could harbour or be associated
with detrimental effects following ingestion. This scenario highlights another issue that warrants further investigation in terms
of learning and intelligence in the horse.
In terms of differences among species, it was also previously
suggested, that the ratio of the brain to body weight was the best
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
measure of intelligence (Jerison, 1973). However, it has been
subsequently demonstrated that this method also falls short as
a critical index or measurement technique for intelligence during comparisons of animals of very different size (Houpt, 1979).
Relative species intelligence ranking may prove at best questionable, even though horse owners in particular, often speculate how
intelligent horses are in relation to other animals (McDonnell,
2001). Moreover, idiosyncrasies in animal behaviour probably
reflect a range of sensations actually experienced by the animal species under investigation and this issue needs to be better
understood in order to definitively evaluate animal intelligence.
Nonetheless McLean (2001) suggested that current experimental methods should be extended to a greater diversity of species,
and that innovative experiments should also be designed to test
for specific cognitive function in all groups of animals including
the horse.
4. Memory in the horse
While it is generally accepted that memory is the function
of encoded neural connections, there is as yet no universally
agreed model of how memory specifically works (Schacter,
1996). Wolff and Hausberger (1996) reported that equine learning and memorisation did not appear to be linked following
simple experimental discrimination and spatial discrimination
tasks and concluded that learning and memory in the horse, at
least, may involve different processes. Nicol (2002) reported that
many examples of excellent memory and recall ability have been
documented in the horse. It does appear however, that the extent
of memory can be easily taken for granted and that assumptions
about memory in the horse are often made without knowing the
basis for such assumptions. Indeed there is scarcely an animal
behaviour that is not affected by memory to some degree. In
one study, horses proved themselves capable of remembering
and repeating a learned response after an interval of 1 week
under experimental maze conditions (Marinier and Alexander,
1994). In another study based on a spatial task requirement, foals
learned and remembered very well when they were exposed to
identical wall mounted compartments to locate food (Mal et al.,
1993). The achievement of successful learning and memory outcomes is likely to be extremely important in the human–horse
relationship and contemporary training programmes.
Given the reports of excellent memory in the horse, it is perhaps somewhat surprising that recent equine short-term spatial
memory research has suggested that horses may have limitations
in recall ability. Specifically, horses may not have a prospective
type of memory, particularly in relation to temporal delays during exposure to stimuli (McLean, 2004a). This may be of crucial
importance in equitation and all other forms of training associated with the horse and it is an issue that will no doubt invite
further investigation. Notwithstanding the difficulty that any
form of temporal delay may present for the horse, it would appear
that the more a horse’s brain is stimulated in terms of memory recall, the quicker the learning of new experiences occurs
(Hanggi, 1999). It is equally likely, that the optimal method for
keeping the horse’s brain actively involved in any learning or
memorisation task is to provide variation in the animal’s envi-
5
ronment and activities. There is certainly anecdotal evidence to
suggest that horses require active learning behaviours to maximise learning potential and memorisation and that horses only
learn and remember very poorly under passive conditions (the
acquisition of learning without the motivation to do so). This is
an area, which requires further objective assessment in terms of
memory limitations and the implications this has for contemporary training schemes. Objective measurement of memory
capability in the horse, and, comparisons of performance of
learning and remembering under both active and passive systems, would provide useful data in terms of equine learning and
behavioural processes.
5. Experimental task learning and behaviour in the
horse
Some of the earliest experimental work designed to investigate equine learning demonstrated that horses could discriminate
between a regular feed box and a feed box that was covered with
a black cloth (Fig. 1), and younger horses exhibited fewer fear
responses and more interest in new stimuli (Gardner, 1937a).
This result has potential implications for contemporary training schemes suggesting the possibility of identifying a most
advantageous time for learning in the horse. Such a clearer
understanding of, and at what point in terms of age or perhaps
neural and behavioural development, horses learn optimally
would also have welfare implications for dealing with equines.
Further evidence of differences in age related learning followed
during a subsequent study using 62 equine subjects, where the
cloth was alternately placed either above or below the feed box
(Gardner, 1937b). In this second study the number of errors
made by the horses increased. It should be remembered however,
that there is an adaptive value for the horse in perceiving and
responding, usually by avoidance, to small changes in a familiar
environment. The second study also indicated that although the
learning curve for all horses showed a rapid initial descent, the
younger horses still learned the correct behaviour more readily
than older horses, even though the differences were not significant (Gardner, 1937b).
While no sex differences in learning behaviour were actually reported during the Gardner studies, in total contrast to the
female horses, 32 of the 37 male horses removed the black cloth
from the feed boxes during the trials to access the feed boxes
Fig. 1. An artist’s impression of the Gardner feed box. From one of the earliest
experimental trials for horses (from Gardner, 1937a,b).
6
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
(Gardner, 1937a,b). The placement of the black cloths in different locations added to the difficulty of the task and perhaps
challenged the spatial awareness of the horses. In an experimental study of 62 horses’ aptitude for discrimination processes
and learning behaviour, Murphy et al. (2004) reported sex differences in learning skills and visuo-spatial ability in the horse. The
findings indicated that male horses appeared to have benefited
from superior visuo-spatial ability, as has also been reported in
other species including humans (Masters and Sanders, 1993),
meadow voles (Kavaliers et al., 1998) and rodents (Roof and
Stein, 1999). There may be subtle sex differences in elements of learning behaviour or perceptual ability in all species
and how such a phenomenon affects the horse could have
implications within equitation and the contemporary training
environment.
5.1. Experimental testing of equine learning and behaviour
Reversal learning is the ability to adjust responses when the
reinforcement values of stimuli are changed. As an experimental technique, reversal learning has also received some attention
in equine behavioural studies. Where horses were required to
discriminate between a black feed box and a white feed box, the
horses were not only successful in the discrimination task but
they were also successful in learning a daily reversal as to which
box contained the feed (Warren and Warren, 1962). Experimental designs have also simultaneously encompassed both visual
reversal discrimination (different colour feed buckets) and spatial reversal trials (left or right positional placement of the target
feed bucket) in the horse. In general, findings from such studies have demonstrated that horses were successful in learning
both types of discriminations, but that the spatial reversals were
more easily learned than the visual reversal problems. There are
however, conflicting reports of exactly what colours, degree of
luminescence and perhaps shades of colour horses can successfully learn to discriminate (Macuda and Timney, 1999; Smith
and Goldman, 1999; Saslow, 1999; Geisbauer et al., 2004; Hall
and Cassaday, 2006). It may be that the location of the food
source or other stimulus is a more salient cue than colour cues,
particularly within reversal trial designs. For example, Mal et al.
(1993) reported that foals generalised the location of food after
only one trial within a 40 compartment apparatus. Interestingly
however, foals appeared to have very short attention spans in that
they exhibited almost total extinction of the desired response
within 2–3 min of the commencement of the experimental condition. However, Mal et al. (1993) concluded that a one-trial
appetitive conditioning protocol may have useful application
for learning research in the horse.
5.2. Horses experience learning difficulties due to temporal
delays
Food reinforcement is widely used as a positive stimulus in
equine learning and behavioural experimental trials, although
freedom from aversive stimuli has been adjudged as more reinforcing than food provision in the horse (McGreevy, 2004).
Studies of equine pattern discrimination (using a food reinforcer)
have shown that horses successfully selected the correct choice
option of pairs of simultaneously presented cue cards (McCall,
1990). However the introduction of a temporal delay (of only
10 s) into experimental trials, where horses were attempting to
access the correct option of two spatially diverse feed buckets,
following provision of the eliciting stimulus appears to cause a
significant degree of difficulty for the horse (McLean, 2004a).
This level of learning would appear to demand proficiency on the
part of the horse at level 5 or even level 6 from the Hierarchies
of Learning Skills as listed in Table 1. Why the horse should
experience difficulties with temporal delays of such small magnitude is certainly interesting if not totally understood, and poses
a significant challenge to designing training programmes for the
horse. Perhaps motivational issues, attention behaviour, memory shortcomings or the inability to ‘chain’ or interpret on the
part of the horse are critical factors influencing temporal delay
based experimental trials.
Several species of birds, primate and dolphins apparently
demonstrate the capacity to deal with temporal delay trials, at
least, under some experimental conditions and pigeons in particular have demonstrated this ability (Hope and Santi, 2004).
Goats have been trained successfully to discriminate visual stimuli in delayed response tasks (Soltysik and Baldwin, 1972;
Baldwin, 1979). It is unclear whether learning difficulties under
such conditions challenge memory or intelligence status or both
in the horse. Perhaps efficient learning of this nature requires
an element of ‘un-learning’ of a previous episode before the
new learning behaviour will become effective in the horse.
Sappington et al. (1997) have suggested that horses may have
difficulty in replacing ‘old learning’ with ‘new learning’ and this
might account for lack of progress in some experimental studies
and training regimes. In this regard, it may well be that the law
of primacy (first learned is best learned) is far more critical to
equine training programmes than was originally acknowledged
(Atkinson and Shiffrin, 1971). If the law of primacy influences
learning behaviour in the horse to the extent of causing difficulties with re-learning, this may have major implications for
equine training programmes.
5.3. Maze learning trials in the horse
Various types of maze, often with more than two choice
options, have been employed to investigate learning behaviour in
several rodent species. The maze method has also been utilised
with larger animals including sheep (Liddell, 1925), pigs (Koba
and Tanida, 2001), cattle (Arave et al., 1992) and horses (Kratzer
et al., 1977; Haag et al., 1980). McCall et al. (1981) conducted
exploratory research using a Hebb–Williams closed maze field
to investigate the extent and degree of equine maze learning abilities. The horses were presented with a different maze problem
every day for 12 days. The findings indicated that the horses were
capable of learning the new maze problems. This approach also
permitted the researchers to apply a rating of the horses’ learning
ability by ranking the horses based on the order and magnitude
of their maze learning ability under test conditions. It would be
interesting to re-visit this area of equine research and examine
the repeatability of the rankings assigned to individual horses
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
and to assess performance based on rankings across a range of
different maze problems.
The use of maze testing in equine memory and learning
behaviour has also been reported as having had important beneficial results. In one study, the authors declared that the horses
were found to not only learn and understand the problem, which
was presented, but Marinier and Alexander (1994) also concluded that the horses understood the principles behind the
problem. However, it appears that there is still little evidence
suggesting a correlation between rankings on different types of
experimental trials in the horse. This may be purely a reflection of the horses’ ability or inability to perform under certain
conditions or equally perhaps, inappropriate experimental trial
design on the part of researchers. Nonetheless, it would appear
that some horses are better at learning certain individual and specific tasks, while other horses perform better at different tasks.
In particular, a lack of correlation in the findings for the same
horses between the outcomes of spatial (following a route) and
instrumental (opening a wooden box for food) tasks would seem
to suggest that both tasks might actually require different learning processes and ability in the horse (Wolff and Hausberger,
1996).
5.4. Sidedness could influence learning and behaviour in
the horse
There is increased interest in the topic of sidedness or laterality among ethologists and evolutionary biologists (Ventolini
et al., 2005). While the study of laterality was originally more
associated with neurology and neuropsychology, it has become
apparent that this phenomenon is not just unique to man but
widespread among other species also (Vallortigara et al., 1999).
Sidedness or lateralised motor behaviour in the horse may have
the potential to influence maze-testing outcomes and laterality has been shown to convey a negative influence on athletic
performance in the horse (Dalin et al., 1985). When designing
experimental trials involving maze or choice tests for horses,
care should be taken to consider the possibility of inherent sidedness associated with the subjects. Previous experimental choice
trials in cattle have shown that cattle were reluctant to change
from one side of a maze apparatus to the other (Grandin et al.,
1994). More recent work in dairy heifers has shown that the
animals learned to choose a preferential side of a Y maze in an
attempt to avoid an aversive stimulus such as noise (Arnold et
al., 2004).
Kratzer et al. (1977) set out to measure the learning ability of
horses in a situation relatively free of human interactions during testing where the animals had to choose a correct escape
route. The authors had assumed that learning would have been
reflected in the simple maze by a decrease in the number of
errors, a decrease in latency of escape and the tendency to choose
the correct escape route. The results from this study showed that
preferences for left and right choices varied among the horses,
and, curiously, taller and thinner horses tended to opt for the
left choice alternative. As is the case, in many species, male
horses tend to be taller than their female counterparts and the
results may have been due to a sex affect. On the other hand, it
7
may have occurred due to breed differences in reactivity such as
those reported in many recent temperament and personality trials
(Le Scolan et al., 1997; Momozawa et al., 2003). Nonetheless,
in other studies, individual sidedness choices of horses have
revealed patterns significantly different from random choice
expectations where individual animals exhibited pronounced
laterality. Grzimek (1968) reported observations of 53 horses
in which 77% of horses observed displayed a preference for the
foot used to paw, 67% had a preference for a foot to initiate walk
and 23% of the horses had a foot preference for galloping, and
therefore indicated significant degrees of sidedness in the test
subjects.
Although significant right handed bias has been reported
in humans (Rife, 1940; McManus, 1985; Klar, 1996; Annett,
2003) there is a higher incidence of left-sidedness observed in
human males compared to human females (Gladue and Bailey,
1995). Similar trends have been reported in rodents (Waters
and Denenberg, 1994), domestic dogs (Wells, 2003) and more
recently in the horse (Murphy et al., 2005; McGreevy and
Rogers, 2005). As a consequence, experimental trials involving food preference choices in the horse could be influenced
by sidedness issues and it would be important to control for
sidedness when conducting this type of experimental work in
equine diet preference type trials. McGreevy and Rogers (2005)
have highlighted that the convention of handling horses from the
left side could possibly influence side bias with regard to motor
behaviour even when the horses were not being handled. Studies
of feral, unhandled or young naı̈ve horses (foals) may provide
more accurate data outlining the extent and impact of laterality
in the horse.
6. Social and observational learning
True observational learning is a complex higher mental ability predicated on reasoning and insight to allow exact imitation
(Nicol, 1996). Horse handlers have regularly trained younger
animals to follow older more experienced horses over jumping obstacles, to travel together in transporters and stand at
ease beside conspecifics for procedures such as clipping and
shoeing. Whether or not this constitutes absolutely true observational learning, it is generally accepted that young/naı̈ve horse
learn something at least from older/more experienced conspecifics (Kiley-Worthington, 1987). There are some reports
albeit anecdotal, suggesting that horses may develop undesirable stereotypic behaviours as a direct result of observational
learning (Kiley-Worthington, 1983; McGreevy et al., 1995) but
the majority of scientific data to date does not appear to support
this thesis (Cooper and Nicol, 1994).
It does seem however, that at least some animals actively
learn to imitate stereotypic behaviour by observing conspecifics
and voles exposed to such activity have acquired the learned
behaviour earlier and perform the stereotypic behaviour for
much longer periods (Cooper and Nicol, 1994). Nicol (1995)
reviewed studies on the ability of a number of species of domestic animals to acquire information and skills by observation
of conspecifics. There was some evidence of varying levels of
social learning in pigs, hens, dogs and cats. Hens and cats demon-
8
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
strated impressive social learning skills; pigs and dogs to a lesser
degree, while both cattle and horses performed less well in this
regard. No significant evidence of true observational learning
has, as of yet, been produced in the horse following several
studies (Baer et al., 1983; Baker and Crawford, 1986; Clarke et
al., 1996; Lindberg et al., 1999). However, this lack of enhanced
social learning reported in the horse may be an artefact of the
experimental conditions and trial designs employed with equine
subjects to date. Further work in this area is warranted as it does
seem likely that there could be some real advantage associated
with enhanced social learning skills (if indeed it exists) in the
horse.
It may be that actually designing an appropriate experimental
set of conditions to investigate social and observational learning
in the horse is a more complex task than has been originally
thought. More recent attempts have been made to address this
issue but the methodology although modified somewhat was
basically similar to that employed in earlier work (McLean,
2004b). Clarke et al. (1996) revisited the issue of observational
learning in horses and broadly based their study on the earlier methods used by Baer et al. (1983) with some adjustments.
They tested fourteen horses of mixed age and breed (seven controls and seven observers) where the observers were exposed to
correct performances of a trained demonstrator (an unfamiliar
horse) for 20 trials over 2 days. The controls, although handled
and subjected to the similar placement procedures, were not
exposed to the demonstrator conducting the task. As had been
the case with the earlier (Baer et al., 1983) work, Clarke et al.
(1996) found no significant effects of prior observation on discrimination accuracy in the feed related task in the horses. There
were however, strong significant effects of prior observation on
latency to approach the goal area on the first trial. Might there be
scope to develop training based on ‘chaining processes’ whereby
the horse might learn some desired behaviour if this latency to
approach the target area were harnessed? It would seem that the
horses were at least better motivated to participate immediately
following exposure to the demonstrator.
One other point worthy of note in the Clarke et al. (1996)
study was the fact that the demonstrator was an animal not known
to or familiar with the test subjects prior to the experimental
conditions. The familiarisation protocol only required that the
demonstrator be stabled ‘next door’ for a period of 18 h for social
customisation. Was this sufficient time and did the conditions
imposed allow for adequate socialisation among the horses, or
could this mechanically controlled social contact have had any
adverse affect on the learning outcomes of the experimental animals? Perhaps observational learning may be influenced by a
dominance or alpha type factor whereby subjects might have
more interest in or possibly be more motivated by the actions of
a ‘respected’ conspecific. It may be important to employ a dominant type animal as the use of the acknowledged alpha leader
of a group of test subjects might be more influential in terms of
interest and attention on the part of the test subjects. Whatever
the case may be, it appears that horses pastured outdoors in small
groups learn to complete trials and training programmes more
quickly than horses housed singly in stalls (Rivera et al., 2002;
Sondergaard and Ladewig, 2004). It has also been proposed that
horses kept in group situations realise their motivation for social
behaviour more easily and, as a consequence, such interaction
allows the horses to understand the signals better from a trainer
or handler (Sondergaard and Ladewig, 2004). Certainly, issues
such as motivation and attention would appear to be a very
important concern for consideration in such circumstances. A
better understanding is required of how these qualities impact
on the behavioural and learning processes in the horse.
7. Attempts at assessing higher order cognition in
horses
Intrinsic conceptualisation capability or the ability to form
concepts based on some common characteristic among different stimuli involves greater mental ability and higher cognitive
function on the part of an organism or species (Table 1). There
had been no known research prior to 1994 detailing investigation
of concept learning in horses. Sappington and Goldman (1994)
designed a study to test the ability of horses to perform at this
level of the hierarchy of learning skills as per Thomas (1986).
Perhaps one problem with attempting to assess concepts theory in another species is that humans have previously decided
what the common characteristics between stimuli actually are.
However, as we often realise to our peril, even between humans,
perceptions can and do vary enormously, with different individuals often deducing a different meaning or concept from identical
stimuli. Sappington and Goldman (1994) presented a series of
two choice discrimination problems on stimulus panels that
could open to allow access to food bowls in an attempt to explore
concept formation ability in the horse. The results demonstrated
complex pattern discrimination ability in horses, and suggested
that they may be able to solve higher order problems using
concept formation in some problem solving scenarios, which
equates to at least level 6 as per Thomas (1986). The conceptually based discrimination task such as that used in the Sappington
and Goldman (1994) study is a much more challenging problem
than a simple discrimination task for any species including the
horse. This is because it requires that the subjects recognise or
perhaps more accurately realise and understand that different
stimuli share a common characteristic and the characteristic is
in essence the focus of the desired goal directed behaviour.
There have been more recent attempts to assess higher order
cognitive ability in the horse. In the standard identity matchingto-sample format (IDMS), experimental subjects are presented
with different coloured lights, shapes, sounds, or static versus
moving stimuli. Initially a sample stimulus is presented in the
centre of a stimulus array and response to the stimulus by the subject prompts the presentation of two comparison stimuli. One of
these stimuli is physically identical to the sample and the second
acts as a distraction. The subject is rewarded when responding to the comparison stimulus that is the same as the original
sample stimulus. Flannery (1997) used a similar conceptually
based or conditional discrimination procedure (an adaptation of
an IDMS task) to test if the subjects realised that there was a
relationship between stimuli. That study used cards on a wall
background to test the acquisition of relational discrimination
ability in the horse. The results of the study demonstrated that:
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
(a) horses could discriminate matching shapes; (b) the ability to
discriminate was not adversely affected by an intermittent primary reinforcer; (c) altering the manner in which the stimuli are
presented (on a novel background with greater distance between
stimuli) did not produce a significant decrement in accuracy.
Flannery (1997) concluded that the performances throughout
this study showed the ability of horses to engage in higherorder discriminations, and that horses successfully learned the
relational discriminations and demonstrated their ability to generalise this learning under several different conditions.
There have been reports of further work, which provides
compelling evidence of horses with the ability to demonstrate
conceptualisation under experimental conditions. Horses have
successfully participated in experimental trials, which have
specifically investigated the study of relative size among objects
and the use of two and three-dimensional objects with open or
closed centres (Hanggi, 1999, 2003). Concept formation, based
on the criterion of differentiating between hard and soft materials, has also been successfully demonstrated where ponies
acted as the experimental group (Watt and McDonnell, 2001).
How learning or perhaps degrees of learning ability is likely
to influence participation in future training is an important
issue when dealing with the horse and this has considerable
implications for training. The relationship between learning
ability and training ability has been addressed in a number
of equine studies. Fiske and Potter (1979) reported a positive
correlation between the test performance of young horses on
a serial reversal learning task and subsequent training for riding. On the other hand Marinier and Alexander (1994) failed
to find any positive correlation between handling or training
behaviour and learning ability during a maze test. However,
positive correlations between ratings of learning ability and individual behavioural reactions of the horses during experimental
trials at several different riding schools have been reported (Le
Scolan et al., 1997). There is an obvious difficulty in this type of
study, whereby factoring in the influence of different handlers
is at best almost impossible and needs much attention in future
investigations.
8. Learning and behaviour in the feral horse
Feralisation has been described as the opposite or the reverse
of domestication and therefore the feralisation process per se
cannot occur in individuals, rather it is restricted to populations of animals (Daniels and Bekoff, 1989). Previous studies
have reported that free-ranging or feral horses have learned to
occupy a home range and generally will attempt to return to
this broadly defined area of the home range following relocation through human interference (Goodwin, 2002). Behavioural
studies of feral populations of breeding mares have provided
some interesting findings whereby after foaling, mares with foals
were reported to have separated off into distinct subgroups (van
Dierendonck et al., 2004) It appears that feral mares attempt to
keep the foals at a safe distance and separated from the more
energetic geldings and sub-adults in the feral population. What
is not absolutely clear however is if this behavioural activity is
totally at the behest of the mares or if it also results at least to
9
some degree from increased mutual attraction between the foals.
The remaining barren or control mares within a herd situation
actually tend to increase interaction with the rest of the herd.
Klingel (1975) proposed a bonding theory whereby pregnant
mares separated themselves from other members of the herd for
a period of several to many hours prior to and around time of
parturition. It seems logical that this behaviour is important in
allowing the neonatal foal sufficient time to recognise, initiate
the bonding process and/or perhaps ‘imprint’ on the dam.
In both feral and domestic populations of horses, foals are
precocious developers and, unlike calves or fawns, which tend
to lie in undergrowth, can gallop with their dams within a few
hours of birth (Goodwin, 2002). Foals generally ‘learn’ to stay
within the immediate proximity of the dam during the first weeks
post-partum and begin to engage in exploratory trips away from
the dam with other foals between one and two months of age.
While foals bond to and appear to learn from their dams, they
also learn to recognise individuals within their own species.
Although the likelihood is that there is a natural instinct present
to do so, the foal also appears to learn about feeding and ingestive behaviour and perhaps sheltering behaviour by participation
with and imitation of its mother in early life. Learning behaviour
in the foal results from developing and practising locomotory
skills and other playing behaviours between foals at this early
stage (Carson and Wood-Gush, 1983; 1996). The play activity
is seen as critically important in order that foals learn to interact
with one another and equally in the social establishment of pair
bonds (McGreevy, 2004). Perhaps interestingly, in terms of the
mare-foal dyad, Houpt et al. (1982) concluded that dams did not
appear to teach their foals with regard to learning a spatial task.
However others have suggested that there may be a genetic influence whereby the ability to deal with spatial tasks in the horse is
more likely to be inherited from the sire (Wolff and Hausberger,
1996).
Studies of equine reproductive behaviours in feral horses
might be useful when designing training programmes that propose to optimise learning in the horse. Some findings to date
indicate that in feral or free-ranging situations, younger colts
appear to learn something of sexual behaviours from observations of the stallion copulating with the harem mares—the
so called Fraser Darling effect (Wilson, 1975). It also appears
that reproductive success is much enhanced by the development
of stable relationships between mares and a single stallion in
free-ranging groups and such activity also has the effect of reducing aggression between the individual animals (Linklater et al.,
1999). Furthermore it seems likely that all horses are capable of
easily learning (understanding) the hierarchical ranking system
in a group or herd, whether it be linear or even more complex
(Houpt et al., 1978). There is still much to be gained from observations of feral horses and their environmental interactions with
regard to behavioural ecology and learning processes, which will
lead to greater understanding of the equine cognitive processes
from the human point of view. In this vein, Houpt and OgilvieGraham (2004) have recently stressed the importance of the
providing appropriate conditions for the domestic horse based
on the range of behavioural and learning activity demonstrated
in feral or free-ranging equine groups.
10
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
9. Contemporary training schemes and equine learning
Modern companion and performance horses are increasingly
required to perform tasks unlikely to emanate from or indeed be
represented in the natural or feral situation. Some of the current
competition and equine husbandry systems present conditions
where horses have to deal with unnatural obstacles and other features that feral horses would naturally or could otherwise avoid.
This contrasts with evolutionary adaptive behaviour for horses,
e.g. jumping, negotiating or manoeuvring around simple or more
elaborate series of impediments or negotiating entry into dark
or narrow areas such as stocks, starting stalls or trailers. Several
of these tasks require the horse to suppress many of its natural
instincts and also either to have or acquire the ability to discriminate and respond to a wide variety of different stimuli (McCall,
1990). Many learning and behaviour studies have routinely challenged the horse to perform trials in a context not particularly
common in practical horse training (Sondergaard and Ladewig,
2004). However, the ability to learn and perform or respond to
the different stimuli influences not only the economic value but
also the status (intellect) of the individual horse to an owner or
trainer.
Curiously within the disciplines of equitation, learning of
the desired behavioural processes is most often attempted using
negative reinforcement strategies (McCall, 1989). Negative reinforcement strategies are premised on the removal of a stimulus
(typically an aversive stimulus) to obtain the desired behaviour
(Chance, 1993). Yet on the other hand experimental designs
to actually measure, assess or induce learning and desired
responses in the horse rely almost exclusively on primary positive reinforcement regimes (Nicol, 2002). In contrast, the basis
for positive reinforcement is the addition of a desirable stimulus
following exhibition of a desirable behaviour (McLean, 2004b).
Furthermore, horse trainers are isolated from advances in animal
training and are largely unaware that they are using negative reinforcement in training (McLean, 2004a). Contemporary training
schemes and the innate intelligence of the horse might be more
harmoniously employed following clearer definition and better
usage of the intrinsic behavioural and learning processes in the
horse.
10. Cellular and molecular basis of equine learning
behaviour
While interest in equine learning, behaviour and welfare
issues is growing, much more research in these areas needs to
be undertaken so as to continue to improve our understanding of
equine ethology and ultimately benefit the horse-human relationship. As the status and value of the horse continues to appreciate
in terms of a companion animal, investigating issues such as
natural balance, sidedness and idiosyncratic biomechanical gait
preferences in performance horses is likely to yield useful data
with regard to improving performance and contemporary training schemes. Motor laterality has been previously studied as an
indicator of cerebral hemispheric asymmetry and various learning abilities have been associated with the different sides or
cerebral structures of the human brain (Coren, 1992; Hellige,
1993). Ventolini et al. (2005) have highlighted the fact that
there is a growing understanding that laterality affects subjects,
particularly animals with laterally placed eyes, not only under
controlled experimental conditions but in more natural conditions also. The use of magnetic resonance imaging (MRI) and
positron emission tomography (PET) scanning techniques have
greatly advanced the understanding of brain activity in terms
of learning behaviour in humans. The development of appropriate experimental trials is complicated by particular logistical
and management difficulties for this technology in dealing with
other species like the horse. But this technology may prove
both very applicable and useful in the future and provide hitherto unavailable data regarding learning and behaviour in the
horse.
There has been remarkable progress in terms of molecular and cellular approaches to biology and the workings of
the nervous systems are now being unravelled by the neurosciences. It may soon be possible to have a clearer understanding
of a range of different equine behaviours in terms of the
underlying processes and mechanisms. Investigative laboratory
techniques including neuronal histology (albeit post mortem)
of the pre-frontal and parietal cortex and other brain structures may yet yield further fascinating insights into equine
behaviour and learning processes. Comparative investigations
between feral or free-ranging and domestic horses, in the
areas of spatio-temporal behavioural patterns of male and
female foals prior to weaning, and the effects of stress on
the mental processes in young horses, may elucidate more
fully, the learning, attention and motivational attributes of the
horse.
Investigations of equine visual systems, perceptual ability and particularly attention studies may have much more to
yield. Male and female monkeys have been reported to select
different human targets as the focus of their aggression and
possibly view humans as agonistic competitors (Hosey, 2005)
and similar studies in the horse might prove equally enlightening. Technological advances have made enormous strides
in terms of understanding human behavioural and learning
processes based on neurobiological and psychological assessment from MRI and PET techniques. Detailed investigations
of the equine visual system in the form of eye-tracking studies might yet provide useful data to aid our understanding of
the extent of perceptual abilities and the link between training and learning ability in the horse. Research of this nature,
as soon as it is feasible, would do much to enhance the
understanding of the behavioural and learning processes in the
horse.
Acknowledgement
The authors would like to thank Debbie Goodwin for advice
and helpful suggestions.
References
Annett, M., 2003. Cerebral asymmetry in twins: predictions of the right shift
theory. Neuropsychologia 41, 469–479.
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
Arave, C.W., Lamb, R.C., Arambel, M.J., Purcell, D., Walters, J.L., 1992.
Behaviour and maze learning ability in dairy calves as influenced by housing,
sex and sire. Appl. Anim. Behav. Sci. 33, 149–163.
Arnold, A., Hemsworth, P., Ng, K., Jongman, E., 2004. Assessment of tape
recorded milking shed noise as an aversive stimulus for dairy heifers, using
a Y maze choice test. In: Proceedings of the 38th International Congress of
the ISAE. ISAE, Helsinki, Finland, p. 85.
Atkinson, R.C., Shiffrin, R.M., 1971. The Control of Short Term Memory. W.H.
Freeman, San Francisco, CA.
Baer, K.L., Potter, G.D., Friend, T.H., Beaver, B.V., 1983. Observation effects
on learning in horses. Appl. Anim. Ethol. 11, 123–129.
Baker, A.E.M., Crawford, B.H., 1986. Observational learning in horses. Appl.
Anim. Behav. Sci. 15, 7–13.
Baldwin, B.A., 1979. Operant studies on shape discrimination in goats. Physiol.
Behav. 23, 455–459.
Bouton, M.E., Swartzentruber, D., 1991. Sources of relapse after extinction in
Pavlovian and instrumental learning*1. Clin. Psychol. Rev. 11, 123–140.
Budiansky, S., 1997. The Nature of Horses. Their Evolution, Intelligence and
Behaviour. Weidenfeld and Nicolson, London.
Carson, K., Wood-Gush, D.G.M., 1983. Equine behaviour. I. A review of the literature on social and dam—Foal behaviour. Appl. Anim. Ethol. 10, 165–178.
Chance, P., 1993. Learning and Behaviour. Brooks and Cole, Belmont, USA.
Clarke, J.V., Nicol, C.J., Jones, R., McGreevy, P.D., 1996. Effects of observational learning on food selection in horses. Appl. Anim. Behav. Sci. 50,
177–184.
Cooper, J.J., 1998. Comparative learning theory and its application in the training
of horses. Equine Vet. J. Suppl. (27), 39–43.
Cooper, J.J., Nicol, C.J., 1994. Neighbour effects on the development of locomotor stereotypies in bank voles, Clethrionomys glareolus. Anim. Behav.
47, 214–216.
Coren, S., 1992. Handedness, traffic crashes and defensive reflexes. Am. J. Public
Health 82, 1176–1177.
Creigier, S.E., 1987. Trailer problems and solutions. In: Current Therapy in
Equine Medicine. Saunders, Philadelphia, USA, pp. 135–138.
Dalin, G., Magnusson, L.E., Thafvelin, B.C., 1985. Retrospective study of
hindquarter asymmetry in Standardbred trotters and its correlation with
performance. Equine Vet. J. 17, 292–296.
Daniels, T.J., Bekoff, M., 1989. Feralization: The making of wild domestic
animals. Behav. Process. 19, 79–94.
Davis, S.L., Cheeke, P.R., 1998. Do domestic animals have minds and the ability
to think? A provisional sample of opinions on the question. J. Anim. Sci. 76,
2072–2079.
Dixon, J., 1970. The horse: A dumb animal? Neigh! Thoroughbred Rec. 192,
1654.
Domjan, M., Burkhard, B., 1986. The Principles of Learning and Behaviour.
Brokes Cole, Pacific Cove, California.
Fiske, J.C., Potter, G.D., 1979. Discrimination reversal learning in yearling
horses. J. Anim. Sci., 49.
Flannery, B., 1997. Relational discrimination learning in horses. Appl. Anim.
Behav. Sci. 54, 267–280.
Gardner, L.P., 1937a. The responses of horses in a discrimination problem. J.
Comp. Physiol. Psychol. 23, 305–333.
Gardner, L.P., 1937b. Responses of horses to the same signal in different positions. J. Comp. Physiol. Psychol. 23, 203–332.
Geisbauer, G., Griebel, U., Schmid, A., Timney, B., 2004. Brightness discrimination and neutral point testing in the horse. Can. J. Zool. 82, 660–670.
Giebel, H.D., 1958. Visuelles lernvermogen bei einhufern Zool. In: Waring, G.H.
(Ed.), Horse Behavior, 1983. Noyes Publishing, Park Ridge, New Jersey, pp.
229–231.
Giulotto, E., 2001. Will horse genetics create better champions? Trends Genet.
17, 166.
Gladue, B.A., Bailey, J.M., 1995. Spatial ability, handedness, and human sexual
orientation. Psychoneuroendocrinology 20, 487–497.
Goodwin, D., 1999. The importance of ethology in understanding the behaviour
of the horse. Equine Vet. J. Suppl. 28, 15–19.
Goodwin, D., 2002. Horse behaviour: evolution, domestication and feralisation.
In: Waran, N. (Ed.), The Welfare of Horses. Kluver Academic Publishers,
Dordrecht/London.
11
Grandin, T., Odde, K.G., Schutz, D.N., Behrns, L.M., 1994. The reluctance of
cattle to change a learned choice may confound preference tests. Appl. Anim.
Behav. Sci. 39, 21–28.
Grzimek, B., 1968. On the psychology of the horse. In: Friedrich, H. (Ed.), Man
and Animal Studies in Behaviour. St. Martins Press Inc., New York.
Haag, E.L., Rudman, R., Houpt, K.A., 1980. Avoidance, maze learning and
social dominance in ponies. J. Anim. Sci. 50, 329–335.
Hall, C.A., Cassaday, H.J., 2006. An investigation into the effect of floor colour
on the behaviour of the horse. Appl. Anim. Behav. Sci. 99, 301–314.
Hanggi, E.B., 1999. Categorization learning in horses (Equus caballus). J. Comp.
Psychol. 113, 243–252.
Hanggi, E.B., 2003. Discrimination learning based on relative size concepts in
horses (Equus caballus). Appl. Anim. Behav. Sci. 83, 201–213.
Hellige, J.B., 1993. Hemispheric Asymmetry. Harvard University Press, Cambridge, MA.
Hodos, W., Campbell, C.B.G., 1969. Scala naturae: why there is no theory in
comparative psychology. Psychol. Rev. 76, 337–350.
Hope, C., Santi, A., 2004. Pigeons’ memory for number of events: effects of
intertrial interval and delay interval illumination. Learn. Motiv. 35, 348–370.
Hosey, G.R., 2005. How does the zoo environment affect the behaviour of captive
primates? Appl. Anim. Behav. Sci. 90, 107–129.
Houpt, K.A., 1979. Intelligence of the horse. Equine Pract. 1, 20–26.
Houpt, K.A., Law, K., Martinisi, V., 1978. Dominance hierarchies in domestic
horses. Appl. Anim. Ethol. 4, 273–283.
Houpt, K.A., Ogilvie-Graham, T.S., 2004. Comfortable Quarters for Horses in
Research Institutions. http://www.awionline.org/pubs/cq02/Cq-horse.html.
Houpt, K.A., Parsons, M.S., Hintz, H.F., 1982. Learning ability of orphan foals,
of normal foals and of their mothers. J. Anim. Sci. 55, 1027–1032.
Houpt, K.A., Rudman, R., 2002. Foreword to special issue on equine behavior.
Appl. Anim. Behav. Sci. 78, 83–85.
Houpt, K.A., Zahorik, D.M., Swartzman-Andert, J.A., 1990. Taste aversion
learning in horses. J. Anim. Sci. 68, 2340–2344.
Hull, C.J., 1943. Principles of Behaviour. D. Appleton Century, New Haven,
USA.
Jerison, H.J., 1973. Evolution of the Brain and Intelligence. Academic Press
Inc., London.
Kavaliers, M., Ossenkopp, K.P., Galea, L.A., Kolb, B., 1998. Sex differences in
spatial learning and prefrontal and parietal cortical dendritic morphology in
the meadow vole, Microtus pennsylvanicus. Brain Res. 810, 41–47.
Kiley-Worthington, M., 1983. Stereotypes in horses. Equine Pract. 5, 34–40.
Kiley-Worthington, M., 1987. The Behaviour of Horses. J.A. Allen, London,
UK.
Klar, A.J.S., 1996. A single locus, RGHT, specifies preference for hand utilisation in humans. Cold Spring Harbour Symp. Quant. Biol 61, 59–65.
Klingel, H., 1975. Social organization and reproduction in equids. J. Reprod.
Fertil. Suppl. 23, 7–11.
Koba, Y., Tanida, H., 2001. How do miniature pigs discriminate between people?
Discrimination between people wearing coveralls of the same colour. Appl.
Anim. Behav. Sci. 73, 45–58.
Kratzer, D.D., Netherland, W.M., Pulse, R.E., Baker, J.P., 1977. Maze learning
in quarter horses. J. Anim. Sci. 46, 896–902.
Lansade, L., Bertrand, M., Boivin, X., Bouissou, M.-F., 2004. Effects of handling
at weaning on manageability and reactivity of foals. Appl. Anim. Behav. Sci.
87, 131–149.
Le Scolan, N., Hausberger, M., Wolff, A., 1997. Stability over situations in
temperamental traits of horses as revealed by experimental and scoring
approaches. Behav. Processes 41, 257–266.
Liddell, H.S., 1925. The behaviour of sheep and goats in learning a simple maze.
Am. J. Psychol. 36, 544–552.
Lindberg, A.C., Kelland, A., Nicol, C.J., 1999. Effects of observational learning
on acquisition of an operant response in horses. Appl. Anim. Behav. Sci. 61,
187–199.
Linklater, W.L., Cameron, E.Z., Minot, E.O., Stafford, K.J., 1999. Stallion
harassment and the mating system of horses. Anim. Behav. 58, 295–
306.
Linnaeus, C. 1758. Systema Naturae, Edito Decima, Tomus I., Impensis Direct.
Laurentii Salvii. i-ii + 1–824.
Mackintosh, N.J., 1983. General Principles of Learning. Blackwells, Oxford.
12
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
Mackintosh, N.J., 1988. Approaches to the study of animal intelligence. Br. J.
Psychol. 79, 509–525.
Macuda, T., Timney, B., 1999. Luminance and chromatic discrimination in the
horse (Equus caballus). Behav. Processes 44, 301–307.
Mal, M.E., McCall, C.A., Newland, C., Cummins, K.A., 1993. Evaluation of a
one-trial learning apparatus to test the learning ability in weanling horses.
Appl. Anim. Behav. Sci. 35, 305–311.
Marinier, S.L., Alexander, A.J., 1994. The use of a maze in testing learning and
memory in horses. Appl. Anim. Behav. Sci. 39, 177–182.
Masters, M.S., Sanders, B., 1993. Is the gender difference in mental rotation
disappearing? Behav. Genet. 23, 337–342.
McCall, C.A., 1989. The effect of body condition of horses on discrimination
learning abilities*1. Appl. Anim. Behav. Sci. 22, 327–334.
McCall, C.A., 1990. A review of learning behavior in horses and its application
in horse training. J. Anim. Sci. 68, 75–81.
McCall, C.A., Potter, G.D., Friend, T.H., Ingram, R.S., 1981. Learning abilities
in yearling horses using the Hebb–Williams closed field maze. J. Anim. Sci.
53, 928.
McDonnell, S., 2001. How smart is he? The thinking horse. In: The Horse.
Article # 3181, Dec.
McDonnell, S.M., Poulin, A., 2002. Equid play ethogram. Appl. Anim. Behav.
Sci. 78, 263–290.
McGreevy, P., 2004. Equine Behavior: A Guide for Veterinarians and Equine.
W.B. Saunders, Edinburgh.
McGreevy, P.D., French, N.P., Nicol, C.J., 1995. The prevalence of abnormal
behaviours in dressage, eventing and endurance horses in relation to stabling.
Vet. Rec. 137, 36–37.
McGreevy, P.D., McLean, A.N., Warren-Smith, A.K., Waran, N., Goodwin, D.,
2005. Defining the terms and processes associated with equitation. In: Proceedings of the First International Equitation Science Symposium. Sydney
University, AEBC, Melbourne, Australia.
McGreevy, P.D., Rogers, L.J., 2005. Motor and sensory laterality in thoroughbred horses. Appl. Anim. Behav. Sci. 92, 337–352.
McLean, A.N., 2001. Cognitive abilities—the result of selective pressures on
food acquisition? Appl. Anim. Behav. Sci. 71, 241–258.
McLean, A.N., 2004a. Short-term spatial memory in the domestic horse. Appl.
Anim. Behav. Sci. 85, 93–105.
McLean, A.N., 2004b. The mental processes of the horse and their consequences
for training. In: Animal Welfare Science Centre. University of Melbourne,
Melbourne, Australia.
McLean, A.N., McGreevy, P.D., 2004. Training. In: Equine Behaviour. A Guide
of Veterinarians and Equine Scientists. W.B. Saunders, Edinburgh.
McManus, I. C. 1985, Handedness, language dominance and aphasia: a
genetic model. Psychological Medicine, Monograph Supplement No.8,
pp 1–40.
Mills, D.S., 1998a. Applying learning theory to the management of the horse:
the difference between getting it right and getting it wrong. Equine Vet. J.
Suppl. 27, 10–13.
Mills, D.S., 1998b. Personality and individual differences in the horse, their
significance, use and measurement. Equine Vet. J. Suppl., 10–13.
Miyashita, Y., Nakajima, S., Imada, H., 2000. Differential outcome effect in the
horse. J. Exp. Anal. Behav. 74, 245–253.
Momozawa, Y., Ono, T., Sato, F., Kikusui, T., Takeuchi, Y., Mori, Y., Kusunose,
R., 2003. Assessment of equine temperament by a questionnaire survey to
caretakers and evaluation of its reliability by simultaneous behavior test.
Appl. Anim. Behav. Sci. 84, 127–138.
Murphy, J., Sutherland, A., Arkins, S., 2005. Idiosyncratic motor laterality in
the horse. Appl. Anim. Behav. Sci. 91, 297–310.
Murphy, J., Waldmann, T., Arkins, S., 2004. Sex differences in equine learning
skills and visuo-spatial ability. Appl. Anim. Behav. Sci. 87, 119–130.
Nicol, C.J., 1995. The social transmission of information and behaviour. Appl.
Anim. Behav. Sci. 44, 79–98.
Nicol, C.J., 1996. Farm animal cognition. J. Anim. Sci. 62, 375–391.
Nicol, C.J., 2002. Equine learning: progress and suggestions for future research.
Appl. Anim. Behav. Sci. 78, 193–208.
Oehme, F.W., 1987. Plant toxicities. In: Robinson, N.E. (Ed.), Current therapy in Equine Medicine, vol. 2. W.B. Saunders, Philadelphia, pp. 672–
682.
Price, E.O., 1999. Behavioral development in animals undergoing domestication. Appl. Anim. Behav. Sci. 65, 245–271.
Rees, L., 1997. The Horse’s Mind. Ebury Press, London.
Rife, D.C., 1940. Handedness with special reference to twins. Genetics 25,
178–186.
Rivera, E., Benjamin, S., Nielsen, B., Shelle, J., Zanella, A.J., 2002. Behavioral
and physiological responses of horses to initial training: the comparison between pastured versus stalled horses. Appl. Anim. Behav. Sci. 78,
235–252.
Roberts, M., 1996. The Man who Listens to Horses. Hutchinson, London.
Roof, R.L., Stein, D.G., 1999. Gender differences in morris water maze performance depend on task parameters. Physiol. Behav. 68, 81–86.
Rubin, L., Oppegard, C., Hintz, H.F., 1980. The effect of varying the temporal
distribution of conditioning trials on equine learning behaviour. J. Anim.
Sci. 50, 1184–1187.
Santamaria, S., Bobbert, M.F., Back, W., Barneveld, A., van Weeren, P.R., 2005.
Effect of early training on the jumping technique of horses. Am. J. Vet. Res.
66, 353–552.
Sappington, B.F., Goldman, L., 1994. Discrimination learning and concept formation in the Arabian Horse. J. Anim. Sci. 72, 3080–3087.
Sappington, B.K.F., McCall, C.A., Coleman, D.A., Kuhlers, D.L., Lishak, R.S.,
1997. A preliminary study of the relationship between discrimination reversal learning and performance tasks in yearling and 2-year-old horses. Appl.
Anim. Behav. Sci. 53, 157–166.
Saslow, C.A., 1999. Factors affecting stimulus visibility for horses. Appl. Anim.
Behav. Sci. 61, 273–284.
Schacter, D., 1996. Searching for Memory. The Brain, the Mind, and the Past.
Basic Books, New York.
Smith, S, Goldman, L., 1999. Color discrimination in horses. Appl. Anim. Behav.
Sci. 62, 13–25.
Soltysik, S., Baldwin, B.A., 1972. The performance of goats in triple choice
delayed response tasks. Acta Neurobiol. Exp. 32, 73–86.
Sondergaard, E., Ladewig, J., 2004. Group housing exerts a positive effect on
the behaviour of young horses during training. Appl. Anim. Behav. Sci. 87,
105–118.
Tarpey, R.M., 1975. Basic Principles of Learning. Scott Foresman, Glenview,
IL, p. 261.
Thomas, R.K., 1986. Vertebrate intelligence: a review of the laboratory research.
In: Hoage, R.J., Goldman, L. (Eds.), Animal Intelligence: Insights into the
Animal Mind. Smithsonian Institution Press, Washington, DC.
Vallortigara, G., Rogers, L.J., Bisazza, A., 1999. Possible evolutionary origins
of cognitive brain lateralization. Brain Res. Rev. 30, 164–175.
van Dierendonck, M.C., Sigurjonsdottir, H., Colenbrander, B., Thorhallsdottir,
A.G., 2004. Differences in social behaviour between late pregnant, postpartum and barren mares in a herd of Icelandic horses. Appl. Anim. Behav.
Sci. 89, 283–297.
Ventolini, N., Ferrero, E.A., Sponza, S., Della Chiesa, A., Zucca, P., Vallortigara, G., 2005. Laterality in the wild: preferential hemifield use during
predatory and sexual behaviour in the black-winged stilt. Anim. Behav. 69,
1077.
Visser, E.K., van Reenen, C.G., Schilder, M.B.H., Barneveld, A., Blokhuis, H.J.,
2003. Learning performances in young horses using two different learning
tests. Appl. Anim. Behav. Sci. 80, 311–326.
Warren, J.M., Warren, H.B., 1962. Reversal learning by horse and raccoon. J.
Genet. Psychol. 100, 215–220.
Waters, N.S., Denenberg, V.H., 1994. Analysis of two measures of paw preference in a large population of inbred mice. Behav. Brain Res. 63, 195–204.
Watt, L.M., McDonnell, S.M., 2001. Demonstration of Concept Formation in
the Horse. Equine Behavior Laboratory, University of Pennsylvania School
of Veterinary Medicine, August 2001 Interim Report.
Wells, D.L., 2003. Lateralised behaviour in the domestic dog, Canis familiaris.
Behav. Processes 61, 27–35.
Wiepkema, R., 1987. Behavioural aspects of stress. In: Wiepkema, P.R., van
Adrichem, P.W.M. (Eds.), Biology of Stress of Farm Animals: An Interactive
Approach. Martinus Nijhoff, Dordrecht.
Williams, J.L., Friend, T.H., Nevill, C.H., Archer, G., 2004. The efficacy of a
secondary reinforcer (clicker) during acquisition and extinction of an operant
task in horses. Appl. Anim. Behav. Sci. 88, 331–341.
J. Murphy, S. Arkins / Behavioural Processes 76 (2007) 1–13
Williams, J.L., Friend, T.H., Toscano, M.J., Collins, M.N., Sisto-Burt, A., Nevill,
C.H., 2002. The effects of early training sessions on the reactions of foals at
1, 2, and 3 months of age. Appl. Anim. Behav. Sci. 77, 105–114.
Wilson, E.O., 1975. Sociobiology: The New Synthesis. Harvard University
Press, Cambridge.
13
Wolff, A., Hausberger, M., 1994. Behaviour of foals before weaning may have
some genetic basis. Ethology 96, 1–10.
Wolff, A., Hausberger, M., 1996. Learning and memorisation of two different
tasks in horses: the effects of age, sex and sire. Appl. Anim. Behav. Sci. 46,
137–143.