Content uploaded by Lynda Birke
Author content
All content in this area was uploaded by Lynda Birke on Mar 26, 2018
Content may be subject to copyright.
© , , | ./-
() -
brill.com/soan
Journeys Together: Horses and Humans
in Partnership
Lynda Birke
University of Chester, United Kingdom
lb.thepublisher@care4free.net
Joanna Hockenhull
University of Bristol, United Kingdom
Abstract
Many living with companion animals hope for “good relationships” based on trust,
mutuality, and cooperation. Relationships develop from mutual actions, yet research
often overlooks nonhumans as mindful actors within relationships. This is a study of
horse/human dyads, using multimethod approaches intended to include horses as
participants. We ask: can “good relationships” be observed, especially when the pair
know each other well? We studied familiar/unfamiliar pairs, negotiating simple obsta-
cles, to explore qualities of cooperation between pairs. Interviews with human partici-
pants elicited perceptions of horses’ “personalities” and reactions. We analyzed video
recordings of interactions and also showed them to external observers. We identied
diferences in attention, tension and coordination: familiar pairs were more coordi-
nated, mutually attentive, and less tense, and they showed less resistance. That is,
some relationships displayed discernible qualities of “working together.” We cannot
know nonhuman animals’ experiences, but knowing how they behave says something
about their agency within interspecies relationships.
Keywords
human-animal relationship – horses – qualitative methodology – mutuality
It’s about true communication and understanding of your horse. [It
implies] a brilliant relationship—happiness...it’s such an emotive
SOAN 23.1_81-100_Birke.indd 81 12/15/2014 5:55:48 PM
82
() -
journey and on that journey you take along your best friend and they
experience it too.”
Interviewee; ,
Many of us who live closely with companion animals speak of having a “good
relationship” with a particular individual: the dog or cat who understands our
every move and mood, the horse who trusts and works alongside us. This bond
is something we hope for, or work at; it entails trust, mutual understanding,
cooperation, coordination, predictability. It is, as the speaker quoted above
described, an “emotive journey” on which we—and (we hope) our compan-
ion animals of whatever species—experience togetherness. To engage in such
interspecies relationships, we must all learn to read the body language, the
behavior, of the other, to begin to anticipate their actions (and they, ours).
These are crucial elements of what people may seek with a companion animal
(or vice versa)—an apparent working together.
Working together, in the sense of being able to understand at least some of
the other species’ behavior is part of our dealings with many kinds of nonhu-
man animals. Both we and they must learn when to approach, when to receive
a stroke/lick, or when to keep well away from the potential blow/bite. Working
together can take many forms. When the animal partner is, for example, a large
animal like a horse, it can be the embodied collaboration involved in riding,
but it can also be mutual understanding on the ground, moving around each
other in the barn, or when the human asks the horse to do something (move
around an obstacle, say). A good relationship means knowing the other, but
it also means understanding that species in general, as horse people often
acknowledge (Brandt, 2004; Birke, 2007).
However, despite burgeoning scholarly interest in human relationships
with other animals and an expressed desire to bring animals in to the social
sciences, research on actual animals acting in relation to us seems remark-
ably sparse—not least because of the predominant use of verbal language in
research (e.g., interviewing: see Taylor, 2012). Indeed, all too often our studies
seem to be about animals rather than including them as participants, while
nonhuman animals seem at times to be seen as mere instincts, rather than
mindful actors or selves (Irvine, 2007). As a result, we know rather little about
how one-to-one relationships between people and other animals are built,
how they work together in their daily lives.
To be sure, there are some studies that focus on the detailed one-to-one relationship between
humans and animals—for example, Irvine’s (2004) study of dogs in which she points out
SOAN 23.1_81-100_Birke.indd 82 12/15/2014 5:55:48 PM
83 :
() -
Yet what is involved in focusing on relationships, while trying to recognize
animals as actors? It is not easy to include nonhumans in investigations in
ways that parallel human subjects—you can’t interview them in any obvious
way, for example. What you can do, however, is careful observation in ways that
include both humans and animals and how they engage with each other. Thus,
there have been several calls recently to develop a multispecies ethnography
(e.g., Kirksey & Helmreich, 2010), based on research practices that acknowl-
edge the mutuality of diferent species in producing sociality, and calls to
include ethological approaches that draw on understanding animal behavior
and cognitive abilities alongside, and integrated with, other methodologies
(Birke & Hockenhull, 2012).
Studying specic human-animal relationships requires integrating multiple
methods, drawing on sociology and ethology (Franklin, Emmison, Haraway, &
Travers, 2007) and bringing diverse approaches into the same study. By using
overt observations of animals’ behaviors, we can at least partly “bring in the
animal viewpoint.” Integration is a dicult task, however, as it involves draw-
ing on diferent disciplinary backgrounds, theoretical assumptions, and ways
of going about asking questions; furthermore, such research can never treat
humans and animals completely symmetrically, for many reasons (see Birke &
Hockenhull, 2012). But despite those caveats, researchers can strive to include
animals more fully. To look at human-animal relationships at a micro level,
one-to-one, means focusing less on impacts of one on the other and more on
how both animals and humans contribute to their engagement, producing
something more than the sum of the parts. Only through such multi-layered
work over time, suggested Franklin et al. (2007), can we begin to understand
the roles played by both partners and come to understand how and why rela-
tionships may grow, how they produce an apparent cooperation.
Such cooperative mutuality is not only apparent to humans while engaging
with animals, but it can often be seen by other people. Close experience with
another—whether the animal is of the same or a diferent species—can pro-
duce a familiarity, an ease which outsiders may recognize (as several studies
have noted: Brandt, 2004; Irvine, 2004). Such mutuality cannot, however, easily
be broken down into constituent components; an observer cannot readily say
that A did this so B did that. Rather, the partners in close relationships seem
to co-produce actions, they become mutually responsive to/with each other.
how both the person and dog share in the production of meaning within the relationship.
Such focused studies are, nevertheless, relatively rare.
In a series of studies of human mother-infant pairs, Kochanska et al. (2008) described, for
example, the emergence over time of what they called mutually responsive orientation, that
SOAN 23.1_81-100_Birke.indd 83 12/15/2014 5:55:48 PM
84
() -
Writing about communication between great apes, King (2004) described this
ne-tuning as a coproduction of meaningful gestures between individuals—
what she refers to as a dynamic dance.
Yet what does such mutuality entail? If it can be seen by outsiders, then
what is going on that makes an observer recognize an easy relationship? Is
mutuality recognizable? In the horse worlds of which we are part, it is com-
monplace to refer to such mutually supportive relationships, to acknowledge
that a good relationship entails a high degree of trust, and horse people wax
lyrical about horses with whom they have had especially close bonds—a feel-
ing of “being one with the horse” (Keaveney, 2008). There is, they would assure
us, a special quality to these well-established relationships, a mutuality appar-
ent whether the horse is ridden or just interacting on the ground—something
which can at times produce a kind of dance.
In the study reported here, we sought to explore this idea: is there indeed
something observable, a quality that can be discerned in the interaction
between horse and person? Equestrian observers may believe that they can
tell a close relationship when they see it, but what is it that they (may) be
seeing? Our aim in this investigation was to consider this specic claim that
there is something discernible that marks established relationships: if inter-
viewees in previous work on horse people could talk about seeing good relation-
ships, is this empirically observable? We wanted, furthermore, to look at details
of horse-human relationships along the multilevel lines suggested by Franklin
et al. (2007), to include direct observation of the horse/human dyad, so that
the horses themselves become participants, at least to some extent.
Dening what constitutes a good relationship may be dicult, but perhaps
there is some special quality to well-established ones. So, for this study, we
focused on established, familiar pairings, and contrasted them with unfamiliar
ones. If established bonds do produce coordination and mutual understand-
ing, built over time, then we may expect subtle diferences between pairings
that are well established and those that are not—taken-for-granted ways of
is, patterns in which both mother and infant are responding to and working with each other
cohesively.
We have both lived and worked with horses throughout our lives, and have long been
immersed in several diferent kinds of equestrian cultures within the United Kingdom.
They are participants in the sense that their behaviors afect how the humans behave and
vice versa. We acknowledge disparities of power when horses are led around or ridden by
people, but believe, along with Cudworth (2010) that companion animals can, at times, expe-
rience relationships with people positively, despite that.
SOAN 23.1_81-100_Birke.indd 84 12/15/2014 5:55:48 PM
85 :
() -
working together. In practice, familiar partnerships usually involve the ani-
mals’ guardians, who spend time with them each day.
Familiar partners have a history together. This can, of course, include nega-
tive moments, disagreements, or attempts by one or the other to “win an argu-
ment” (even within an otherwise good relationship). But on the whole, most
people who keep a horse for a time seek a peaceful coexistence as, presum-
ably, do horses who are required to live in human-dened contexts. So, while
familiarity may famously breed contempt, it can also breed ways of adjusting
to each other, which may be absent or less developed if the two individuals
have never met before, irrespective of how experienced the person is with
horses. By observing these sometimes subtle diferences we aimed to describe
what we might term “relationship quality”—that hard-to-dene something to
which horse people so often aspire (Birke, 2007).
To study how horse-human pairs moved together, we asked them to under-
take a simple task, with the person leading the horse around some simple
obstacles. This was broadly similar to the technique employed by Chamove,
Crawley-Hartrick, and Staford (2002) in a study of the efects of human con-
dence on horse responses, which allowed detailed comparisons of horses being
handled by unfamiliar and familiar people. Filming the interactions allowed us
to evaluate individual horse-human pairs in detail.
Our primary emphasis in this study was thus qualitative/descriptive, and
we drew in part on the approach to observing animals pioneered by Françoise
Wemelsfelder (2007, 2012). In her work, she sought to prioritize the whole ani-
mal as subjective being, as behaver. Her work showed that there are qualities
to the whole animal, which can be observed and quantied systematically, and
which provide assessments of the animal’s subjective state. Here, we extended
this idea to include qualitative assessments of horse-human relationships:
what are the overall qualities a particular pairing show when they are working
together? Evaluations of quality, however, should not rely solely on research-
ers’ judgements; we also wanted to see if there was any consensus among
external observers regarding such qualities. This step involved showing lm
sequences to outside panels of observers (university students) for evaluation
(a step that was rather loosely modeled on the approach used by Wemelsfelder
et al., 2001, 2007).
Wemelsfelder et al. (2001) drew on Free Choice Proling, based on rating scales derived from
observers’ prior generation of verbal descriptors. In , these scales then form the basis of
detailed quantitative analysis (generalized Procrustes analysis), which assesses reliability of
observers’ scores. In our study, however, we did not seek to further quantify the rating scores,
SOAN 23.1_81-100_Birke.indd 85 12/15/2014 5:55:49 PM
86
() -
Our study was thus multi-strategy, with our ethological observations
of horses and people supported by evaluations from other observers, and
interviews with participants; ndings from the qualitative core were further
complemented by some selected quantitative measurements, to allow fur-
ther conrmation. Taken together, the thematic analysis from eld notes/
interviews and assessment by external observers, together with the quanti-
tative component, provide a form of triangulation, each validating the other
(Bryman, 2004).
Materials and Methods
Subjects
We sought (human) volunteers at four venues: one was a private yard, housing
two horses, while the other three were livery yards ofering stabling for large
numbers of animals. Twenty-one participants agreed to take part with their
horses (ve at each of the three venues, and six at the other). All testing took
place in enclosed arenas (approximately 20 m wide × 40 or 60 m long) between
spring and autumn; three of these were outdoors and was one indoors. Filming
was done using a camcorder set up along the long side of the arena.
Human volunteers’ ages ranged from 14 to late 50s; all but two were female,
reecting the predominance of women among “leisure riders” (i.e., they were
amateurs who rode or worked with the horse in their spare time). Although
some took part in equestrian competitions, this was at an amateur/novice
level. They came from diverse occupational backgrounds, including univer-
sity teachers, students, a veterinarian, and people employed in the equestrian
world (though not as professional riders). Unfamiliar handlers were recruited
from among the people present at the venue at the time of testing (e.g., care-
givers of other horses seeing their own animals that day).
All human participants were told that the research involved a study of
human-horse relationships, and that they would be asked to navigate a simple
route while leading their horses with headcollars and ropes. These movements
would be lmed, and the human handler subsequently interviewed. All people
involved, regardless of whether they were familiar with a particular horse, had
but used them simply to provide comparative descriptions of each dyad, as a means of
corroborating our qualitative observations.
See Birke and Brandt (2009) and Hockenhull and Creighton (2013) regarding the marked
gender imbalance among “leisure” riders. (This is much less noticeable among competition
riders.)
SOAN 23.1_81-100_Birke.indd 86 12/15/2014 5:55:49 PM
87 :
() -
several years of experience with horses and dened themselves as reasonably
condent in horse handling.
Equine participants were between 2 and 24 years old, of varied breeds
(including thoroughbreds, native ponies, and cobs); seventeen of them were
housed at the relevant testing venues, while the remaining four were trans-
ported a short distance to the venue (the private yard). Each had been known
to the familiar person for a minimum of ve months. In most cases, the famil-
iar person was the guardian, but for six horses, he or she was someone else
who shared taking care of, or working with, the animal on a day-to-day basis.
Six horses were mares, the rest were geldings. Horses were used for various
purposes: some for simple hacking or trail riding, others for competitions
including novice dressage and show jumping.
The Task
We asked handlers to lead the horses around a series of obstacles on the
ground. After an initial circle around the arena, they rst entered a chicane
(a double L of poles on the ground), then walked around to a “gate” marked by
two trac cones. Here, they were asked to halt and move the horse backward
a few steps. They then proceeded through a slalom of cones, followed by three
more “gates,” involving movements to both left and right. Finally, they were
asked to run a few steps so that the horses could trot alongside before return-
ing to the start point. The various obstacles required the pair to make several
turns, and adjustments to each other. We began video recording prior to the
pair moving of initially and stopped on their return to start.
Each horse was led by someone familiar and someone unfamiliar (11 were
led rst by the familiar person, 10 by the unfamiliar), with a break for a few
minutes in between during which they stood quietly. In all cases, both handlers
were present throughout, and they were able to observe. Once all the horses at
that venue had done the course, we conducted short interviews with handlers;
handlers had watched all the horses being led, and had seen the horses they
led also handled by someone else. We asked them how they perceived the
individual horse’s personality and reactions, what they noticed and how they
themselves experienced the task.
Video Analysis
The video recordings were broken down into lm clips for each horse/human
pair. We then analysed them in detail, and also showed copies to external
observers, as follows:
This was done to control for possible order efects.
SOAN 23.1_81-100_Birke.indd 87 12/15/2014 5:55:49 PM
88
() -
Assessment by external observers. The clips were rst coded, and then shown
to a panel of observers (a group of 36 students in either general animal studies
or specialist equine degree programs). Before the viewing, students were told
only that the study was concerned with human-horse interaction; further
details of how pairs were compared were given afterward. Each clip of a horse
with a familiar person was shown alongside the clip for the unfamiliar per-
son (left-right positions were varied randomly), and observers were asked to
describe in their own words how they saw these paired interactions between
horses and humans, comparing one to the other. They were asked to focus not
on the individual animal’s demeanor or specic behaviors, but on the overall
qualities of each interspecies interaction, using whatever words came to mind.
Although observers were free to choose any words, there were consistent
patterns in the resulting “word map.” We then used the four most frequent
words/meanings (e.g., tense/not tense, cooperative/not cooperative) to gen-
erate simple rating scales to be used by a second panel of observers (N = 20).
These observers were asked to compare each pair and give them a score on a
scale of 1 to 5 for each of the four categories (ratings of tension, cooperative-
ness, trust, and attention). This resulted in a mean rating for each pair (famil-
iar and unfamiliar), which could then be compared. The purpose here was to
provide a check on the consistency of the terms observers used, rather than to
quantify responses in detail.
Assessment of themes from video. Alongside the external assessments, we
carried out detailed analysis of lm clips, using “Kinovea” sports analysis soft-
ware, with replay speeds of approximately one third of the original recording,
combined with frame-by-frame analysis. This enabled us to produce detailed
description (eld notes) of how each human and horse engaged as a pair. We
then analyzed these narrative descriptions, alongside interview transcripts, for
specic recurring themes.
Results
Qualitative Results
External observers: generation of descriptors. While the student observers inev-
itably used many diferent words to describe horse-human pairings, several
of the descriptors had broadly similar meanings (cooperative and working
together, for example). We were therefore able to trace certain themes, four of
See www.kinovea.org for details.
SOAN 23.1_81-100_Birke.indd 88 12/15/2014 5:55:49 PM
89 :
() -
which occurred consistently (that is, in at least two thirds of responses). The
themes were:
Trust/condence: observers often referred to an apparent mutual trust or
condence between horse and handler, or the lack of it.
Cooperation/working together: observers mentioned that some pairings
appeared cooperative, seeming to work together, while others apparently
lacked such cooperation.
Listening/paying attention to the other: observers perceived some pairs as
paying mutual attention, while other pairs seemed to be less attentive,
even rather oblivious.
Relaxed/tense: observers thought that some pairings were relaxed with
each other, while others were apparently tense.
The second set of observers then watched the lm clips and were asked to
score each pair of horse-human dyads on the four traits identied above (e.g.,
a score of 1 = no trust, 5 = complete trust). The resulting scores conrmed that
the panel of observers generally rated familiar pairs diferently from unfamiliar
pairs; specically, familiar pairs were scored as being consistently more trust-
ing, more cooperative, more attentive, and more relaxed than unfamiliar pairs.
Narrative description and interviews: identifying themes. Although at rst
sight, most of the lmed sequences appeared to show quiet, well-behaved
horses plodding alongside humans, detailed observation of the lms at slow
speed suggested several subtle diferences. Some of these were also identied
by handlers in interviews, thus providing cross-validation. The sports analy-
sis software enabled close scrutiny of relative positions and actions of both
the horse and human, enabling a detailed narrative. Several themes emerged
in parallel from researchers’ notes and from interviews, some of which were
similar to observations made by the student panel. The most signicant
themes were:
Paying attention: eld notes and interviews broadly supported the nd-
ings from the observer panel, namely that some pairings appeared to be more
attentive or tuned in to each other, others less so. More precisely, both horses
and handlers may sometimes look away from where they were going, attend-
ing to something outside the arena. Thus, “horse looking around” appeared
consistently in eld notes, but occurred more often with unfamiliar pairings.
This diference was also noted in interviews: one rider (a young man in his
20s) said of his horse that, “she was looking around a lot more” while work-
ing with the unfamiliar person (human participant L1). Another participant
(a woman in her 20s) observed that her horse with the other handler was
SOAN 23.1_81-100_Birke.indd 89 12/15/2014 5:55:49 PM
90
() -
“looking around a little bit more, and his ears were pricked forward” (L3). So,
looking around more, being less attentive to the task at hand, was more often
associated with unfamiliar humans and horses.
Moving together, cooperating: some horse-human pairs moved in apparent
harmony and uidity, as the observer panel noted, and our eld notes corrobo-
rated that. In other cases, movements seemed to be obviously initiated by the
human.
Diferences in how horses moved with the handler were also commented
upon by interviewees. Thus, one caregiver, who ran a livery yard taking care of
a number of other horses, pointed out that her horse was diferent with the
unfamiliar handler: with her, he was
...kind of to the side of me, his head was very low, he was investigat-
ing the ground and the obstacles as we went past them, whereas when
I was observing him following D around, he was, he was very close to
her...almost burying his head in her back really, and just saying, I’ll
follow you and not look at what’s going on and I know we’ll get there in
the end. (C1)
The horse was older and very familiar with being handled by diferent people,
and both handlers were experienced. Yet there were clear diferences; here, the
caregiver was interpreting her horse’s behavior, noting his diferent position-
ing, and apparent ease alongside her by contrast to his less condent behavior
with the other handler. Other participants noted that some pairings appeared
much more coordinated, much more together, and relaxed.
Following quietly versus resistance. These horses were all used to human
handling, and it may be expected that they would usually follow quietly. Some
indeed did so, but others displayed occasional bouts of resistance (for instance,
pulling back, stopping; also see Chamove et al., 2002). This might have been
a result of miscommunication, or perhaps, as some interviewees suggested,
these horses are “natural leaders” rather than “followers.”
Some interviewees made explicit reference to lack of cooperation. For
example, one handler, a college lecturer (W3), noted that her Exmoor pony
“did do a lot more stopping with the unfamiliar handler, a lot more standing
around....She did seem a lot less forward going with the unfamiliar handler
than with me.”
Livery yards are a common form of boarding other people’s horses in the United Kingdom.
They are very diverse in size and type, with some being dedicated stable yards while others
use conversions of old farm buildings.
SOAN 23.1_81-100_Birke.indd 90 12/15/2014 5:55:49 PM
91 :
() -
Some resistances were not necessarily seen by horse caregivers, however.
Indeed they recognized the individuality and agency of their horses and some-
times portrayed them as having fun, “messing around” whenever (say) they
stopped and pulled back: thus, one young woman noted her horse’s tendency
simply to stop and not go forward, commenting that “it made me laugh when
he kept stopping, he didn’t want to walk forward and that’s just because there’s
people around....He wanted to stop and talk to the people” (C3).
Tension versus relaxation. In some cases, either or both of the partners
appeared tense in the way they carried their bodies (eld notes and observer
panel). Apart from overt resistance, this seemed to be accompanied by (for the
human) holding the hand high on the rope, with the lead rope more taut, and
for the horse, moving the ears and raising the head more often.
Handlers also noted the diference: one handler, a teacher in her 50s, noticed
variation in rope tension, observing that her horse was “perhaps a bit more
willing to come with [her] [than with the unfamiliar handler],” continuing,
“...um, I had a oat in the rope, I had a bit of loose in my rope, whereas C had
quite a tight rope on him...whether that’s because I’m just used to how fast
he walks...” (W6).
In addition to the themes noted by external observers, we also identied
two other features in eld notes. First, there were diferences in leading styles.
For some handlers, the usual style was to have the horse walking alongside,
with the neck approximately parallel to the human’s shoulder (generally con-
sidered to be the “correct” way to lead a horse). However, several handlers
tended to walk out in front (especially if the horse was unfamiliar), expecting
the horse to follow behind. Leading from the front meant that the human had
to look back more often to see where the horse was. Second, some handlers
seemed to try to block the forward movement of the horses if they wanted to
change directions. This involved the whole body moving in front of the horse
(especially if leading from the front), or it simply involved moving the hand
(and lead rope) across the horse’s front.
These qualitative assessments, then, indicate a general consensus between
research eld notes, interviews, and external observers. There were consistent
themes of coordination, condence, and cooperation, as well as trust and
appearing relaxed, which were—unsurprisingly—more often associated with
familiar pairs (see Figures 1 and 2 for illustration).
Adding Quantitative Measures
Although there was broad consensus among observers, we subsequently added
quantitative measurements of specic components as a form of validation.
These were based on themes emerging from the qualitative part of the study.
SOAN 23.1_81-100_Birke.indd 91 12/15/2014 5:55:49 PM
92
() -
Example of a familiar pair showing mutual coordination and
attention. With this pair, there were no overt signals to indicate a
change of direction; rather, movements were luid and together.
Example of unfamiliar pair, not working together. Here, the horse resists the
handler’s attempts to ask him to move forward.
SOAN 23.1_81-100_Birke.indd 92 12/15/2014 5:55:49 PM
93 :
() -
We focused on three components from the qualitative analysis: attention, ten-
sion, and coordination. For each, we took specic measurements in order to
see how robust the qualitative observations were.
They were as follows:
1. Attention: obvious shifts of attention were indicated by head movements
(horse or human looking at something outside the arena, for example).
We recorded frequencies of such attention shifts, away from the direction
of movement, for both participants. We also recorded how often the
human looked directly at the horse—“checking in.”
2. Tension: we assessed tension for humans through rating lead rein tension
on a scale of 1 to 3 (relaxed to tense) as the pair moved through specic
parts of the course (chicane, slalom, and subsequent gates). Rope tension
might have been high during one of these, but not in the others, so
we used the mean score over all three components. For the horse, we
assessed tension through the frequency of ear movements in specic
parts of the course (chicane, halt, and nal gates) and also assessed resis-
tances over the whole course, that is, when the horse pulled back or
raised his or her head. These measures are similar to those used by
Chamove et al. (2002), who argued that ear movements increased when
horses were tense.
3. Greater coordination: we sought an assessment of the apparent greater
coordination among familiar pairs. Analysis of video records at a very
slow speed suggested that coordination may be associated with less vari-
ation in interindividual distances. Familiar pairs may move together, irre-
spective of how far apart the pair were routinely. Thus, a pair could
maintain a relatively constant distance even if the human routinely
walked in front of the horse.
We took measurements of interindividual distances on 7 occasions while
the pair moved in parallel to the long side of the arena (thus perpendicu-
lar to the camera) in the rst circuit of the arena. From the frame-by-
frame analysis, we measured the distance between the center of the
surcingle worn by each horse and the back of the human (usually shoul-
der, depending on relative heights of each).
The surcingle carried a heart rate monitor to provide continuous assessment of changes
in heart rate. These data will be presented separately.
SOAN 23.1_81-100_Birke.indd 93 12/15/2014 5:55:49 PM
94
() -
The rst point was a calibration, to allow for variation in distances of
diferent horses from the camera and for diferent leading styles; this was
set at 50 cm for all horse-human pairs. The remaining six measurements
were then used to calculate mean deviation from the calibration distance
as horse and human moved along the side of the arena.
Quantitative Results
Results from these comparisons between familiar and unfamiliar were ana-
lyzed using paired t-tests (data were normally distributed), and broadly con-
rmed the diferences identied in qualitative analyses. In general, unfamiliar
pairings were less coordinated, less attentive, and exhibited more tension.
Specically, looking away (attention shift) occurred approximately twice as
often in unfamiliar pairs (mean frequency human looking away: familiar = 2.7;
unfamiliar = 4.6; p < 0.005; for the horse: familiar = 2.0; unfamiliar = 4.1;
p = 0.006). Human looking back was more frequent in unfamiliar pairs (mean
= 11.5) than familiar ones (mean = 5.7; p = 0.005).
Unfamiliar pairs were also more likely to show higher lead-rein tension
(p < 0.05), and these horses had higher frequencies of ear movements
(familiar = 8.3; unfamiliar = 13.0; p = 0.001) and exhibited resistance more often
(mean number of resistances in familiar pairs = 1. 9; unfamiliar = 3.6; p = 0.02).
Some of these resistances seemed to follow immediately after the handlers
“blocked” the horses’ forward movements with their hands or body—75% of
resistances for unfamiliar pairs were preceded by blocking, compared to 25%
for familiar pairs.
Our nal measurement was used to provide a measure of the observed coor-
dination: this showed that greater apparent coordination among familiar pairs
was associated with a smaller mean variation in distance between horse and
handler. The mean deviation from initial calibration for familiar pairs was 2.4,
while the deviation for unfamiliar pairs was 4 (p = 0.004).
Taken together, these measurements support the primary qualitative assess-
ments. Interactions between familiar horse-human pairs are more likely to
be coordinated, less tense, more attentive, and less likely to show resistance.
These qualities of experienced partnerships in turn become evident in spe-
cic details that can sometimes be quantied: less distractible, less tense, more
moving together. It is these qualities that can at times give the overall impres-
sion to observers of greater harmony—even in the apparently straightforward
task of human leading horse.
SOAN 23.1_81-100_Birke.indd 94 12/15/2014 5:55:49 PM
95 :
() -
Discussion
Some of the horse-human interactions we observed showed few diferences;
both horse and person seemed much the same whether familiar or unfamiliar.
They were, on the whole, older horses who were quite used to walking around
with humans. But for most, there was a discernible—though often subtle—
diference. To external observers, as well as to the researchers, established
partnerships were indeed more coordinated, giving an impression of moving
together, of harmony and predictability.
Our study has some limitations, which should be addressed here. One
important limitation of our approach was that it was restricted to a snapshot in
time: we focused only on how the two partners interacted at a specic moment,
doing a specic task, and in a specic place. We know little about the previous
history of that horse with that person, or, indeed, with people in general. As
Noske (2003) noted, looking at one moment in an ongoing relationship is to
take it out of context, away from the ongoing and contextualized engagement
of one with the other (also see Segerdahl’s comments on working with great
apes: Segerdahl, 2012). On the contrary, relationships are processes, they pro-
duce biographies: they both happen in context and create context themselves
(Franklin et al., 2007). “Snapshots” would thus yield somewhat limited infor-
mation about how those horses and humans behaved in that time and place.
Constraints also arise because of the nature of the task: the horse and person
must move within areas prescribed by our layout and by the arena, while they
are observed from a distance. But more important perhaps, the horses them-
selves are physically constrained. These are horses who are very accustomed to
being handled and moved around; they cannot move with complete freedom,
but they are required to do what humans ask, at the end of a lead rope. In
this sense, what we observe is a horse whose expressiveness is constrained and
whose agency is thereby limited. Moreover, judgments about what constitutes
a “good relationship” are inevitably human ones. While we can make eforts to
understand potential negative impacts on the horses (making assessments of
their stress for instance), we can never know for sure how they would evaluate
a particular relationship, or, indeed person.
On the other hand, making comparisons at a specic moment can tell us
something about how horses and people interact now, the current quality of
their engagement. What this study emphasizes is that there are indeed quali-
ties about established relationships on which observers can generally agree.
When these working relationships function well, both partners are atten-
tive to each other and to the task in hand, less ready to be distracted by out-
side inuences. There is mutual trust and cooperation, giving an impression
SOAN 23.1_81-100_Birke.indd 95 12/15/2014 5:55:49 PM
96
() -
of harmony. In that sense, the horse has some agency, and both horse and per-
son work together, even within the obvious physical constraints.
Clearly, experiencing horses in general helps people engage with them (and
experiencing humans helps horses). But there is also a specic trust, born out of
learning to judge the other’s actions and moods within ongoing relationships;
over time, the participants learn to read each other, creating expectancies—
whether that relationship is within a species (e.g., see Kochanska, Aksan,
Prisco, & Adams, 2008, for studies of how human mother-infant pairs develop
mutuality) or it is interspecies (e.g., Irvine, 2004). That ability to predict and
read the other’s body language is an important factor in producing the cohe-
sion evident here in the movements of familiar partners.
Unfamiliar partners, by contrast, are less trusting, less relaxed, and less
coordinated. Human handlers need to look often at the horse, to ensure she
or he is where she or he is expected to be; the horses, in turn, sometimes react
to unfamiliar people with overt resistance, or with neck or ear tension (see
Chamove et al., 2002). The net efect is generally one of less harmony. Both
know what is expected in general, but they do not know how to read that
specic other. Thus, a simple hand movement, intended to guide, may not be
immediately understood by the horse, who may react by resisting.
We chose a qualitative approach here because we did not want to frame
our horse-human interactions within assumptions of linearity, which would
not seem to do justice to the ne-tuning that we see when we observe good
relationships: when we see these, we see partners moving together. In stud-
ies of human behavior, synchrony in movement encourages observers to judge
the pair as a unit, as having rapport and showing mutual attention (Lakens &
Stell, 2011). While this judgment applies to a third party, Lakens and Stell also
pose the question: would “spontaneous synchronization with an interaction
partner create an even stronger bond?” (p. 11).
Humans can synchronize identical movements, such as walking in stride
or waving, while humans and nonhuman animals cannot easily do so.
Nonetheless, what these studies of human interactions underline is that coor-
dinating behavior with another entails an emotional component, which both
reects and produces the relationship, and so does interspecies coordination,
even if their behavior is not identical (see Maurstad, Davis, & Cowles, 2013,
with regard to human-horse relationships and emotional connection). This
cooperation, and associated afect, is the “being one with the horse” to which
horse people usually aspire.
While this study does not go far enough in meeting the sentience/agency
of the animal other, it does indicate that qualitative approaches to understand-
ing the minutiae of human-animal encounters might be fruitful. Indeed, this
study could be the basis for further, longitudinal, investigations of developing
SOAN 23.1_81-100_Birke.indd 96 12/15/2014 5:55:49 PM
97 :
() -
horse-human bonds, and it may also be used in studies of other species who
live alongside humans. It is such longitudinal and multidisciplinary projects
that Franklin et al. (2007) emphasized with regard to studying human-dog
encounters: relationships, they remind us, build up over time as each party
learns to engage with the other.
More detailed studies, with various species, are crucial. While we may never
know for sure how animals experience their interactions with us, a good rela-
tionship, with close bonding and predictability during interactions, is surely
preferable to a bad one, and we need to know more about how relationships
with other animals work (or not) and how they develop (or not). Too many
companion animals are treated badly (or unpredictably) by humans, usu-
ally with dire consequences for them; on these grounds, we would argue that
knowing more about what constitutes close bonds, and how they are built, has
some potential to improve animal lives.
Physical spaces also frame social relationships in ways that merit further
research attention. Writing about animal cultures and semiotics, Lestel (2002)
emphasized the “joint occupation of spaces” by humans and many animals,
within which human-animal communication is taking place. “Humans,” he
suggests, “are those living beings which invent new and original means of
access to other living beings using non-living mediators” (p. 58). In our study,
the mutuality we observe is framed within particular physical structures
(such as arenas, stables, and the layout of the obstacles), and we cannot tell
from these observations how those particular structures shape the interspe-
cies engagements. We need, suggests Lestel, to explore further the very subtle
sharing of space of humans and nonhumans, to consider an “architecture of
human/animal relations” (p. 58). In the case of horses and people observed
here, we may ask: how are these apparently coordinated behavior sequences
produced by the spaces in which human and equine lives are shared?
However much spaces and emotions have contributed to the relationship,
what we see is a mutual negotiation of space between horse and person. The
horse is an architect of the ne movements just as the person is. This is a kind
of performativity—an example of the choreography, or dynamic dance, to
which King (2004) referred. In this choreography, no one is overtly thinking,
“if I move this way, then the horse/human will do this”; on the contrary, there
is uidity, a moving-together that cannot be easily reduced. Just as children are
caught up in the primacy of movement and experience the world kinetically
(Sheets-Johnstone, 2009), so did the horses and people engaging in our study.
For discussion of the idea of performativity in relation to thinking about animals, see
Birke, Bryld, and Lykke (2004).
SOAN 23.1_81-100_Birke.indd 97 12/15/2014 5:55:49 PM
98
() -
These are not animals simply plodding around at the behest of a human, but
they are mindful of how to read the human from moment to moment—mindful
in moving and being moved. Argent (2012) notes how horses, social animals
that they are, are able to coordinate their movements, to become entrained
with another, even a human. “Moving together,” she suggests, “is the way horses
belong, and horses who have been well and gently socialized to participate in
joint action with a rider seem to want to, and try very hard to, get the synchro-
nous movements ‘right’” (Argent, 2012, 123, emphasis in original).
At best, their combined actions are movements of shared uidity and
connection. Riders have sometimes paid tribute to the wonderful relation-
ships they have with their horses during riding—a oneness evoking the mythi-
cal centaur (Game, 2000; Thompson, 2011). But togetherness is also evident
on the ground; indeed, it is part and parcel of how horses and humans can—
sometimes—be with each other in all kinds of situations. Sadly, we humans
all too often misread what animals are trying to tell us, often with dire con-
sequences for the animals. We may never know for sure how they themselves
experience the relationships, but knowing how they behave within it does tell
us something. A better understanding of how togetherness and partnership
are built, by humans and by companion animals, would surely benet us all.
Acknowledgements
We want to thank the University of Lincoln livery yard and the caregivers of the
three smaller yards for help and participation in the research, as well as Tamsin
Young, Sarah Redgate, and Kristen Armstrong Oma for help in carrying it out.
We are also grateful to Françoise Wemelsfelder for discussions of whether rela-
tionship “quality” may be assessed. But above all, we want to thank the various
equine and human participants in this research, and especially one horse, Kir,
who epitomized the contrast between familiar and unfamiliar handlers and
whose reactions reminded us how important it is to think about how animals
perceive what we do.
References
Argent, G. (2012). Toward a privileging of the nonverbal: Communication, corporal syn-
chrony and transcendence in horses and humans. In J.A. Smith & R.A. Mitchell
(Eds.), Experiencing animal minds: An anthology of animal-human encounters
(pp. 111-128). New York, : Columbia University Press.
SOAN 23.1_81-100_Birke.indd 98 12/15/2014 5:55:49 PM
99 :
() -
Birke, L. (2007). “Learning to speak horse”: The culture of “natural horsemanship”.
Society & Animals, 15, 217-240.
———. (2008). Talking about horses: Control and freedom in the world of “natural
horsemanship”. Society & Animals, 16, 107-126.
Birke, L., & Brandt, K. (2009). Mutual corporeality: Gender and human/horse relation-
ships. Women’s Studies International Forum, 32, 189-197.
Birke, L., & Hockenhull, J. (2012). On investigating human—animal bonds: Realities,
relatings, research. In L. Birke & J. Hockenhull (Eds.), Crossing boundaries: Inves-
tigating human-animal relationships (pp. 15-36). Leiden, Netherlands: Brill.
Birke, L., Bryld, M., & Lykke, N. (2004). Animal performances: An exploration of inter-
sections between feminist science studies and studies of human/animal relation-
ships. Feminist Theory, 5, 167-183.
Brandt, K. (2004). A language of their own: An interactionist approach to human-horse
communication. Society & Animals, 4, 299-316.
Bryman, A. (2004). Social research methods. Oxford, United Kingdom: Oxford Univer-
sity Press.
Chamove, A.S., Crawley-Hartrick, O.J.E., & Staford, K.J. (2002). Horse reactions to
human attitudes and behavior. Anthrozoos, 15, 323-331.
Cudworth, E. (2010). Social lives with other animals: Tales of sex, death and love.
Basingstoke, United Kingdom/New York, : Palgrave.
Franklin, A., Emmison, M., Haraway, D., & Travers, M. (2007). Investigating the thera-
peutic benets of companion animals: Problems and challenges. Qualitative
Sociology Review, 3, 42-58.
Game, A. (2001). Riding: Embodying the centaur. Body and Society, 7(4), 1-12.
Hockenhull, J., & Creighton, E. (2013). The use of equipment and training practices and
the prevalence of owner-reported ridden behaviour problems in leisure horses.
Equine Veterinary Journal, 45, 15-19.
Irvine, L. (2004). If you tame me: Understanding our connection with animals.
Philadelphia, : Temple University Press.
———. (2007). The question of animal selves: Implications for sociological knowledge
and practice. Qualitative Sociology Review, 3, 5-22.
Keaveney, S.M. (2008). Equines and their human companions. Journal of Business
Research, 61, 444-454.
King, B.J. (2004). The dynamic dance: Non-vocal communication in African great apes.
Cambridge, : Harvard University Press.
Kirksey, S.E., & Helmreich, S. (2010). The emergence of multispecies ethnography.
Cultural Anthropology, 25, 545-576.
Kochanska, G., Aksan, N., Prisco, T.R., & Adams, E.E. (2008). Mother-child and father-
child mutually responsive orientation in the rst 2 years and children’s outcomes at
preschool age: Mechanisms of inuence. Child Development, 79(1), 30-44.
SOAN 23.1_81-100_Birke.indd 99 12/15/2014 5:55:50 PM
100
() -
Lakens, D., & Stell, M. (2011). If they move in sync, they must feel in sync: Movement
synchrony leads to attributions of rapport and entitativity. Social Cognition, 29, 1-14.
Lestel, D. (2002). The biosemiotics and phylogenesis of culture. Social Science
Information, 41, 35-68.
Maurstad, A., Davis, D., & Cowles, S. (2013). Co-being and intra-action in horse-human
relationships: a multi-species ethnography of be(com)ing human and be(com)ing
horse. Social Anthropology, 21(3), 322-335.
Noske, B. (2003). Horse images and the human self-image in equine research. In F. de
Jonge & R. van den Bos (Eds.), The human-animal relationship: Forever and a day.
Assen, Netherlands: Royal van Gorcum.
Segerdahl, P. (2012). Human-enculturated apes: Towards a new synthesis of philosophy
and comparative psychology. In L. Birke & J. Hockenhull (Eds.), Crossing boundar-
ies: Investigating human-animal relationships (pp. 139-160). Leiden, Netherlands:
Brill.
Sheets-Johnstone, M. (2009). The corporeal turn. Exeter, United Kingdom: Imprint-
Academic.
Taylor, N. (2012). Animals, mess, method: Post-humanism, sociology and animal stud-
ies. In L. Birke & J. Hockenhull (Eds.), Crossing boundaries: Investigating human-
animal relationships (pp. 37-50). Leiden, Netherlands: Brill.
Thompson, K. (2011). Theorising rider-horse relations: An ethnographic illustration of
the centaur metaphor in the Spanish bullght. In N. Taylor & T. Signal (Eds.),
Theorising animals (pp. 221-253). Leiden, Netherlands: Brill.
Wemelsfelder, F. (2007). How animals communicate quality of life: The qualitative
assessment of animal behaviour. Animal Welfare, 16(S), 25-31.
———. (2012). A science of friendly pigs...Carving out a conceptual space for address-
ing animals as sentient beings. In L. Birke & J. Hockenhull (Eds.), Crossing boundar-
ies: Investigating human-animal relationships (pp. 223-250). Leiden, Nether lands:
Brill.
Wemelsfelder, F., Hunter, E.A., Mendl, M.T., & Lawrence, A.B. (2001). Assessing the
“whole animal”: A free-choice proling approach. Animal Behaviour, 62, 209-220.
SOAN 23.1_81-100_Birke.indd 100 12/15/2014 5:55:50 PM
