Demonstrating Operant Control of Gazing Behavior in Domestic Dogs (Canis Lupus Familiaris) and Gray Wolves (Canis Lupus ...

MISSING IMAGE

Material Information

Title:
Demonstrating Operant Control of Gazing Behavior in Domestic Dogs (Canis Lupus Familiaris) and Gray Wolves (Canis Lupus Lupus)
Physical Description:
1 online resource (45 p.)
Language:
english
Creator:
Spencer, Jessica Marie
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Psychology
Committee Chair:
Wynne, Clive D
Committee Members:
Kimball, Rebecca T
Dallery, Jesse

Subjects

Subjects / Keywords:
canis -- conditioning -- gaze -- operant -- social
Psychology -- Dissertations, Academic -- UF
Genre:
Psychology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
It has been argued that over domestication, dogs developed a genetic predisposition to gaze at a human's face, forming a complex communication system between dogs and humans that exists independently from individual histories. However, Bentosela et al. (2008) demonstrated that gazing behavior in dogs can be increased or decreased by manipulating reinforcement schedules, presenting evidence that the high occurrence of dogs gazing at their owners may result from operant conditioning. The current experiment demonstrates operant control of gazing using differential reinforcement in dogs and wolves. Gazing at targets was reinforced with food and subsequently extinguished and re-established using a reversal design. Dog subjects were placed in one of three conditions where the target was i) the person delivering reinforcement, ii) a person other than the one providing reinforcement, and iii) an inanimate object. Wolf subjects were placed in one of the latter two conditions. The amount of gazing at each target by dog and wolf subjects increased during acquisition, decreased during extinction, and increased again during re-acquisition. The object condition shows that individuals will gaze at any target when that behavior has been followed by food, decreasing the likelihood that a complex communication system developed between dogs and humans. The demonstrated operant control of the gazing behavior of dogs and wolves suggests that a specially evolved ability is unnecessary to account for dogs' gazing behavior. These experiments encourage the consideration of a more parsimonious explanation for gazing behavior, particularly ontogenetic learning processes.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Jessica Marie Spencer.
Thesis:
Thesis (M.S.)--University of Florida, 2012.
Local:
Adviser: Wynne, Clive D.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-05-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044197:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 DE MONSTRATING OPERANT CONTROL OF GAZING BEHAVIOR IN D OMESTIC DOGS ( CANIS LUPUS FAMILIAR IS ) AND GRAY WOLVES ( CANIS LUPUS LUPUS ) By JESSICA M. SPENCER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Jessica M. Spencer

PAGE 3

3 To all who nurtured my intellectual curiosity, academic interests, and sense of scholarship throughout my lifetime, especially my family

PAGE 4

4 ACKNOWLEDGMENTS I thank the chair and members of my supervisory committee for their mentoring, the staff at University of Doglando and Wolf Park for contributing their time and facilities, undergraduate research assistan ts for their contributions, and the owners of the subjects who volunteered their animals to my study. I thank my family, friends, and fellow lab members for their encouragement and support.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 ABSTRACT ................................ ................................ ................................ ..................... 8 CHAPTER 1 EXPERIMENT 1: OPERANT CONTROL OF GAZING IN DOGS ........................... 10 Intro duction ................................ ................................ ................................ ............. 10 Methods ................................ ................................ ................................ .................. 12 Subjects ................................ ................................ ................................ ............ 12 Procedures ................................ ................................ ................................ ....... 12 Phases ................................ ................................ ................................ ....... 13 Conditions ................................ ................................ ................................ .. 14 Inter Observer Agreement ................................ ................................ ................ 15 Statistical Analysis ................................ ................................ ............................ 15 Control Coding ................................ ................................ ................................ 16 Results ................................ ................................ ................................ .................... 17 Discussion ................................ ................................ ................................ .............. 19 2 EXPERIMENT 2: OPERANT CONTROL OF GAZING IN WOLVES ...................... 30 Methods ................................ ................................ ................................ .................. 30 Subjects ................................ ................................ ................................ ............ 30 Procedures ................................ ................................ ................................ ....... 30 Inter Observer Agreement ................................ ................................ ................ 31 Statistical Analysis ................................ ................................ ............................ 31 Results ................................ ................................ ................................ .................... 31 Discussion ................................ ................................ ................................ .............. 33 General Discussion ................................ ................................ ................................ 34 LIST OF REFERENCES ................................ ................................ ............................... 43 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 45

PAGE 6

6 LIST OF TABLES Table page 1 1 Dog Subject Characteristics ................................ ................................ ............... 24 2 1 Wolf Subject Characteristics ................................ ................................ ............... 39

PAGE 7

7 LIST OF FIGURES Figure page 1 1 Gazing durations for individual dog subjects in Baseline and Test sessions for the Feeding Person, Other Person, and Object conditions. ........................... 26 1 2 Mean number of gazes in each session for dogs in Feeding Person, Other Person, and Object conditions. Error bars indicate 95% Confidence Intervals. .. 27 1 3 Mean durations of gazing during the Extinction phase for dog subjects in the Feeding Person, Other Person, and Object Conditions. Error bars indicate 95% Confidence Inter vals. ................................ ................................ .................. 28 1 4 in Control Coding of Other Pers on condition. ................................ ..................... 29 2 1 Gazing durations for individual wolf subjects in Baseline and Test sessions for Other Person and Obje ct conditions. ................................ ............................. 40 2 2 Mean number of gazes in each session for wolves in Other Person and Object conditions. Error bars indicate 95% Confidence Intervals. ...................... 41 2 3 Mean duration of gazing during the Extinction phase for wolf subjects in the Other Person and Object conditions. Error bars indicate 95% Confidence Intervals. ................................ ................................ ................................ ............. 42

PAGE 8

8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science DEMONSTRATING OPERAN T CONTROL O F GAZING BEHAVIOR IN D OMESTIC DOGS ( CANIS LUPUS FAMILIAR IS ) AND GRAY WOLVES ( CANIS LUPUS LUPUS ) By Jessica M. Spencer M ay 2012 Chair: Clive D.L. Wynne Major : Psychology It has been argued that over domestication, dogs developed a genetic between dogs and humans that exists independently from individual histories However, Bentos ela et al. (2008) demonstrated that gazing behavior in dogs can be increased or decreased by manipulating reinforcement schedules, presenting evidence that the high occurrence of dogs gazing at their owners may result from operant conditioning. The current experiment demonstrates operant control of gazing using differential reinforcement in dogs and wolves. Gazing at targets was reinforced with food and subsequently extinguished and re established using a reversal design. Dog subjects were placed in one of three conditions where the target was i) the person delivering reinforcement, ii) a person other than the one providing reinforcement, and iii) an inanimate object. Wolf subjects were placed in one of the latter two conditions. The amount of gazing at each target by dog and wolf subjects increased during acquisition, decreased during extinction, and increased again during re acquisition. The object condition shows that individuals will gaze at any target when that behavior has been followed by food, decreas ing the likelihood that a complex communication system

PAGE 9

9 developed between dogs and humans. The demonstrated operant control of the gazing behavior of dogs and wolves suggests that a specially evolved ability is unnecessary to ior. These experiments encourage the consideration of a more parsimonious explanation for gazing behavior, particularly ontogenetic learning processes

PAGE 10

10 CHAPTER 1 EXPERIMENT 1: OPERANT CONTROL OF GAZING IN DOGS Introduction The last decade has seen a dram atic rise in scientific interest in dog ( Canis lupus familiaris ) behavior, particularly in their interactions with humans (for a review of recent work see Udell & Wynne, 2008; for an historical review see Feuerbacher & Wynne, ility to follow human pointing gestures in an object choice task has led to several hypotheses regarding the effects of domestication, including arguments for an understanding of human actions independent of exposure to humans (Riedel, Schumann, Kaminski, Call, & Tomasello 2008). Some studies (Hare, Brown, Williamson, & Tomasello 2002; Miklosi et al., 2003) suggested that dogs are more successful than wolves ( Canis lupus lupus ) at following the difficult momentary distal point, in which the distance betwe en the pointing hand and the object pointed to exceeds 50 cm and the cue is removed before the individual makes a choice (though for more recent evidence to the contrary, see Udell, Dorey, & Wynne, 2008). Viranyi et al. (2008) suggested that the success of dogs in following human momentary distal points is due to their tendency to establish and maintain eye contact readily with humans, giving them a better chance to observe the signal before it is removed. Some researchers (e.g., Gasci et al., 2005, Miklos i et al., 2003) have argued further that this difference in gazing patterns between dogs and wolves formed during the phylogenetic process of domestication. They claim that dogs developed a genetic predisposition to eir individual histories, creating a complex

PAGE 11

11 behavior cannot be achieved in wolves even afte r intensive socialization with humans (Miklosi et al., 2003). In contrast, Bentosela, Barrera, Jakovcevic, Elgier, and Mustaca (2008) have demonstrated that gazing behavior in dogs can be increased or decreased by manipulating reinforcement schedules, p resenting empirical evidence that the high occurrence of dogs gazing at their owners may result simply from operant conditioning. During the Acquisition phase, the experimenter delivered a treat to the dog immediately following each gaze toward her face, r esulting in an increase in the duration of gazing over three sessions. In the Extinction/Omission phase, treats were either withheld or resulted in a decrease in gazing gazing behavior changed and the ease with which it could be controlled and manipulated suggest that environmental influences may have more profound effects on this social behavior than previously believed. The present study replicates and extends that of Bentosela et al. (2008) by adding two further conditions. In one, we control for the subject inadvertently gazing at the human as an unintended consequence of orienting toward the source of food. In the seco nd new condition, we remove the human face altogether by making the target of gazes an inanimate object. This condition will test whether individuals will gaze at an arbitrary target to which gazing has previously resulted in reinforcement, thus reducing t he likelihood that gazing has a unique role as part of a complex communication system between dogs and humans. Furthermore, if wolves are able to rapidly learn to gaze at humans, this would suggest that gazing is not a specially evolved ability resulting from

PAGE 12

12 domestication. While these experiments would not completely rule out the possibility of a genetic predisposition, they would advocate that the more parsimonious explanation of ontogenetic learning processes accounting for the emission of gazing behavi or must also be considered. Methods Subjects Twenty four pet domestic dogs (Canis lupus familiaris) living in their present FL. Dogs varied by breed, size, and age, but wer e all at least one year old ( Table 1 1). Only dogs reported to be in good health were used. Procedures Each subject was tested individually and off leash in an indoor room (8 m by 8 m) at the dog daycare center, which allowed some auditory and olfact ory access to other dogs and people outside of the room. Each session began with an approximately five minute habituation phase to allow the dog to investigate the room and become comfortable with the experimenter. Up to ten pieces of high quality dog trea ts were delivered non in the experiment. A dog was considered motivated if it approached and readily consumed the treat from the hand of the experimenter, as well as following the tre at with its eyes as the experimenter picked up each piece from the container. Pieces of Pet Botanics pork or chicken or Pup Peroni beef treats (approximately 0.3 g each) were used as reinforcers depending on any food allergies reported by the owner, but on ly one type of treat was used per dog.

PAGE 13

13 A video camera mounted on a tripod recorded the entire experiment. Experimenter 1 looked at the dog and live coded the onset and offset of its gazes toward the target using J Watcher software on a laptop computer. Exp erimenter 1 also delivered reinforcers when appropriate for all conditions. Phases All subjects, regardless of condition, experienced the same series of phases. This included Baseline, Acquisition, Test, Extinction, and Re Acquisition. The inter session i nterval between every session and phase was between 15 30 seconds. The number and duration of sessions, reinforcement schedules, and interval times were derived from pilot trials. Baseline Baseline involved one two minute session of non contingent reinf the subject. Experimenter 1 delivered one treat on a Variable Time schedule of 15 seconds, meaning that a treat was delivered to the dog on average every 15 seconds. A delay of at least 1.5 seconds was inserted following a gaze at the target before treat delivery to avoid adventitious reinforcement of gazing. Acquisition All subjects received five two minute sessions of Acquisition. Each session consisted of Experimenter 1 del ivering reinforcement on a Fixed Ratio 1 schedule immediately following the onset of a gaze toward the target. This means that a treat was delivered to the dog every time it gazed at the target. Test Test consisted of one two minute session identical to B aseline. This session was included for comparison to Baseline to determine any changes in gazing following Acquisition.

PAGE 14

14 Extinction Gazes toward the target were recorded but no indication was made to the subject, and no reinforcers were delivered. Each sub ject received up to five two minute extinction sessions. However, if the subject did not emit a single gaze toward the target during a session, Extinction was terminated and Re Acquisition began. Re Acquisition Each subject experienced a minimum of two an d maximum of five two minute Re Acquisition sessions. Re Acquisition only continued until it was demonstrated that the response had been re acquired (i.e. response rates were similar to responding before Extinction). The circumstances of Re Acquisition wer e identical to those of Acquisition. Conditions Three different conditions were included, with each dog only experiencing one condition. There were eight dogs in each condition. Feeding Person The first condition involved Experimenter 1 calling the subj name once at the beginning of each session. The face of Experimenter 1 was the target Other Person standing 1 meter away from Experimenter 1. Experimenter 2 ca gazes. Experimenter 2 was instructed to look at the dog but otherwise avoid interacting with it throughout the experiment except for calling its name at the beginning. Object the floor so that it was visible to the dog, but it could not access the toy. The stuffed animal was a tan colored bear approximately 50 cm high and was the target of gazes in

PAGE 15

15 prior to the start of the experiment and was placed behind the object before each of each session, coming from the direction of the object. There was a five second delay between the time the device was placed by the object and the name being called. Inter Observer Agreement The data live coded during the experiments were compared to t hose coded from the video recording by one of two trained independent observers. The coders were instructed to code the onset and offset of gazes at the target, the delivery of reinforcers, and any time the subject moved off screen and returned to view. If the subject moved off screen during a session, that portion was excluded from the calculations. Twenty percent of the total videos were coded for inter observer agreement (IOA). Reliability was calculated by dividing the difference of the two scores by th e smaller value, with a criterion of < .20 or by an absolute measure of no more than one unit difference (number of gazes for Acquisition and Re Acquisition sessions and seconds [s] for Baseline, Test, and Extinction sessions). Satisfying either criterion was considered a hit. The IOA percentages were calculated by dividing the number of hits by the total number of sessions coded by the independent observer for each condition. The reliability was 95.7% for Feeding Person, 84.0% for Other Person, and 81.8% f or Object. Statistical Analysis The frequency of gazing during each two minute session was analyzed. Additionally, duration of gazing was analyzed for Baseline, Test, and Extinction sessions because frequency alone does not account for the variability in length of individual gazes. For example, a single gaze could last a fraction of a second or 30 seconds. Duration was not analyzed for Acquisition or Re Acquisition sessions because

PAGE 16

16 the immediate presentation of reinforcement interrupted and truncated each gaze, artificially limiting the duration. We tested whether reinforcement increased the duration of gazing and whether the effect was the same across conditions by conducting a two way ANOVA with within subjects factor Session (Baseline vs. Test) and bet ween subjects factor Condition (Feeding person; Other person; Object). Furthermore, we ran a two way ANOVA to assess changes in gaze frequency across the acquisition and extinction sessions with respect to the Condition (Feeding Person; Other Person; Objec t). Then we conducted a two way ANOVA with repeated measures to assess if the gaze durations in Extinction decreased. We also assessed re acquisition of the gazing response following extinction by comparing the frequency of gazing between Acquisition 1 and Re acquisition 1 with respect to the Condition (Feeding person; Other person; Object). post hoc test. The degrees of freedom for within subjects comparisons were adjusted for dev iance from sphericity (Greenhouse Geisser) to decrease Type 1 errors. An alpha level of .05 was adopted throughout statistical analyses. Control Coding In order to account for the possibility that the presentation of food occasioned gazing in general, ga were coded from video. Similar trends (i.e. an increase in gazing during acquisition and a decrease in gazing during extinction) between the Other Person condition and the Control Coding would suggest that the mere presentation of food was responsible for changes in gazing rather than the schedule of reinforcement for gazing at a particular target. If these trends do not occur, this would strengthen the evidence for the argument

PAGE 17

17 that the schedul e of reinforcement is the controlling factor. Because the camera position was not intended to capture gazing toward Experimenter 1 in this condition, only video from five subjects could be coded in this way. Inter observer agreement was also obtained for 2 0% of these videos. Results Figure 1 1 shows the durations of gazing at the target for individual dogs across Baseline and Test sessions for Feeding Person, Other Person, and Object conditions. Overall, the dogs in all three groups increased the duration of gazing from Baseline to Test. Also, dogs in the Feeding Person condition appeared to gaze at their respective target longer than dogs in the Object condition.The two way ANOVA of gaze durations from baseline to the test sessions confirmed a significant main effect of Session, F(1, 21) = 52.33, p < .001, and a main effect of Condition, F(2, 21) = 13.86, p < .001. However, there was no interaction between the factors, F(2, 21) = 3.27, p > .05. Post hoc tests revealed that the only significant differences i n Conditions lay between In all three conditions, the subjects increased the number of gazes across the five Acquisition sessions (Figure 1 2). A two way ANOVA confirmed the effects of Session, F(2.13, 44.82) = 20.22, p < .001 with Greenhouse Geisser adjustment, and Condition, F(2, 21) = 19.80, p < .001. However, there was no interaction between the factors, F(4.27, 44.82) = 1.68, p > .05 with Greenhouse Geisser adjustment. A test for an increasi ng linear trend across Sessions in the Acquisition phase revealed significant effects for Feeding Person, t(7) = 2.79, p < .05, Other Person, t(7) = 4.49, p < .01, and Object, t(7) = 2.52, p < .05.

PAGE 18

18 Duration of gazing decreased during the Extinction phase for subjects in all three conditions, particularly in Feeding Person and Other Person (Figure 1 3). A two way ANOVA with within subjects factor Session (Extinction 1 5) and between subjects factor Condition (Feeding Person; Other Person; Object) was condu cted. Zeroes were used for sessions that were not conducted because the subject met the criterion of no gazes during a full session (two subjects, two sessions each). There were significant main effects of Session, F(2.85, 59.82) = 6.64, p = .001 with Gree nhouse Geisser adjustment, and Condition, F(2, 21) = 12.43, p < .001. However, there was no interaction between the factors, F(5.7, 59.82) = 1.31, p > .05 with Greenhouse Geisser adjustment. A test for a decreasing linear trend across Sessions in the Extin ction phase revealed significant effects for Feeding Person t(7) = 2.78, p < .05, and Other Person, t(7) = 2.53, p < .05. However, there was no significant linear trend for Object, t(7) = 1.64, p > .05. Overall, gazing in all conditions was re acquir ed following the Extinction phase as demonstrated by the increase in frequency of gazing from Acquisition 1 to Re acquisition 1 (Figure 1 2). The two way ANOVA with within subjects factor Session (Acquisition 1 vs. Re acquisition 1) and between subjects fa ctor Condition (Feeding person; Other person; Object) confirmed significant main effects of Session, F(1, 21) = 28.18, p < .001, and Condition, F(2, 21) = 9.66, p = .001. There was no interaction between the factors, F(2, 21) = 1.51, p > .05. Control Codin g Results Figure 1 4 displays the duration of gazing toward Experimenter 1during the Other Person condition. A visual analysis reveals that there was not an increase in gazing during Acquisition, rather, there was a decrease. If the

PAGE 19

19 presence of food occas ioned gazing in general by causing an increase in all activity, then high rates of gazing toward Experimenter 1 would have maintained. Additionally, there was an increase of gazing activity during Extinction, when no food was presented. This further illust rates that the occurrence of gazing was not merely controlled by the presentation of food. Discussion This experiment was designed to demonstrate that manipulating reinforcement could impact the gazing behavior of dogs. As hypothesized, the results show t hat the duration of gazing toward targets increased for subjects in all three conditions from Baseline to the Test session as a consequence of the intervening Acquisition phase. The rate of gazing also increased during Acquisition. When reinforcement was r emoved in the Extinction phase, the duration of gazing decreased. Finally, the reintroduction of the reinforcement schedule in the Re acquisition phase resulted in another increase in the rate of gazing. Differences in gaze durations between the groups hi ghlight some important points. The duration of gazing for subjects in the Feeding Person group was higher than for the Object group in both Baseline and Test sessions. However, because this could be attributed to the dogs simply orienting toward the source of the reinforcers in the Feeding Person condition a comparison of Object and Other Person groups may be more illuminating. At the group level, there was no significant difference between the two conditions, but visual inspection highlights that although durations were generally similar in Baseline, half of the Other Person subjects gazed longer in Test than any of the subjects in the Object group.

PAGE 20

20 The potential difference in gazing durations in Test between Other Person and Object groups may have several possible causes. As argued in Miklosi et al. (2003), there could be an evolved cognitive mechanism in dogs selected for during domestication that increases the likelihood of gazing interactions with humans. Although this is possible, a more parsimonious e xplanation lies in the conditioning experiences of each individual subject throughout its lifetime. Because only ten minutes of explicit reinforcement for gazing can significantly alter the rate and duration of gazing at either a person or an object, it is possible that extensive informal conditioning experience on a daily basis before the experiment could have affected the rate of gazing at a particular target. Another alternative explanation for these results is that they could be an artifact of the expe rimental setting. For example, the object was kept inside a crate with the dog on the outside, restricting tactile access and constructing a partial visible barrier to the object. This was done to ensure that the dog could not remove the object and damage it. Other research (Udell, Dorey, & Wynne, 2008) has shown that dog responsiveness to a human cue when viewed through a fence is greatly reduced. Meanwhile, the human in the Other Person condition had no such barrier and could be clearly seen and touched, which may have led to higher rates of gazing. Furthermore, the inanimate object was completely motionless, but the human in the Other Person condition did exhibit some subtle movements, such as breathing and yawning, for example, despite efforts to remain attention and increased the chances of subjects contacting the contingency of reinforcement.

PAGE 21

21 The sample of subjects used in this experiment consisted of pet dogs obtained from a dog daycare facility in the Orlando, Florida area. The generality of these results to other populations may be limited. Past research (Udell, Dorey, & Wynne, 2010) has demonstrated that behavior of dogs living in homes as pets compared to those in shelt ers can differ significantly during their interactions with humans. In a comparison of the gazing behavior of shelter dogs and pet dogs, Barrera, Mustaca, and Bentosela (2011) observed shorter durations in Extinction for shelter dogs, suggesting that the d ifferent living conditions and learning experiences through everyday interactions can contribute to differences in social behavior. Future research could investigate the factors affecting this behavior in animals from other populations. Another limitation of this study is that breeds were not systematically analyzed for differences due to a limited number of subjects per breed type. Jakovcevic, Elgier, Mustaca, & Bentosela (2010) compared Retrievers, Shepherds, and Poodles using the Bentosela et al. paradi gm and discovered no differences in acquisition but a slower rate of extinction in Retrievers. They also found a higher duration of gazing in Retrievers during a baseline assessment with no prior gaze training. The major mechanism behind this breed differe nce has not yet been identified (i.e. genetic or environmental), so future research could try to tease apart this difference and include the new conditions of the present study. Furthermore, this particular daycare center provides its dogs with intra speci fic socialization and obedience training onsite, which may have also impacted the behavior of the dogs in this setting. For example, if the subjects previously received obedience training using positive reinforcement methods in the testing room, the settin g of the

PAGE 22

22 room itself could serve as a discriminative stimulus for attending to the human as a source of treats. This may have inflated baseline gazing durations. Likewise, if any subjects experienced aversive training techniques in the testing room, the ro om itself may illicit avoidance responses, decreasing baseline rates of gazing. Future research should investigate the rates of reinforcement in interactions between dogs and humans in natural settings. If it could be shown that owners are more likely to d eliver reinforcement in the form of treats or attention when their dog is looking at them than when it is not, more evidence for the argument of operant control would be provided. In addition, a functional analysis, a procedure used by applied behavior ana lysts to determine which stimulus in the environment is maintaining a particular behavior in humans (e.g. Iwata, Dorsey, Slifer, Bauman, & Richman, 1982) or non human individuals (e.g. Dorey, Rosales Ruiz, Smith, & Lovelace, 2009), could be conducted. For example, potential conditions to be arranged could include tangible reinforcers (e.g. treats), attention (e.g. verbal or tactile), play, escape from a demand, and an ignore condition. Future experiments could also study dogs from infancy, manipulating rat es of reinforcement for gazing at various targets. This would test the hypothesis that reinforcement histories before the experiment are contributing to the patterns we obtained in our data. If dogs raised without receiving reinforcement for interacting wi th humans were shown to have reduced gazing toward humans, there would be further evidence that the individual conditioning experiences are the most important factors. In the current study, the experimenters looked at the subject during the experiment, wh ich may serve as a discriminative stimulus or predictor of reinforcement.

PAGE 23

23 On the other hand, prolonged gaze from the experimenter may be perceived as a threat and elicit avoidance of eye contact or escape behaviors in the subjects, as had been found in hum ans (Greenbaum & Rosenfeld, 1978, cited in Kleinke, 1986). Future up to one where the experimenter does not look at the dog and avoids eye contact.

PAGE 24

24 Table 1 1. Dog Subject Characteristics Name Breed Age Sex Condition Simone Australian Shepherd mix 2 yrs F Feeding Person T Bear Laborador Retriever 3 yrs M Feeding Person Dixon Laborador Retriever 2 yrs M Feeding Person Bella Border Collie 2 yrs F Feeding Person Heidi Laborador Retr iever mix 3 yrs F Feeding Person Datz German Shepherd 4 yrs M Feeding Person Court Weimara ner 3 yrs M Feeding Person Linus Australian Shepherd 5 yrs M Feeding Person Stash Australian Shepherd 2 yrs M Other Person Maya Boxer 2 yrs F O ther Person Sparkle Australian Shepherd 4 yrs F Other Person Scamper Australian Shepherd 4 yrs M Other Person Laila B Boxer 2 yrs F Other Person Hudson Laborador Retriever 2 yrs M Other Person Scout German S.H. Pointer 7 yrs M Other Person Sam Laborador Retriever mix 3 yrs M Other Person Laila C Boxer 4 yrs F Object Sage Hound mix 5 yrs F Object Rain ey G erman Shepherd mix 2 yrs F Object Bo Laborador Retriever 5 yrs M Object Belle Laborador Retriever 3 yrs F Object Sam C German Shepherd 3 yrs F Object

PAGE 25

25 Table 1 1. Continued Name Breed Age Sex Condition Jag American Bulldog mix 2 yrs M Object Kepa Pitbull Terrier mix 4 yrs M Object

PAGE 26

26 Figure 1 1. Gazing durations for individual dog s ubject s in Baseline and Test sessions for the Feeding Person, Other Person, and Object conditions.

PAGE 27

27 Figure 1 2. Mean number of gazes in each session for dogs in Feeding Person, Other Person, and Object conditions. Error bars indicate 95% Confidence In tervals.

PAGE 28

28 Figure 1 3. Mean durations of gazing during the Extinction phase for dog subjects in the Feeding Person, Other Person, and Object Conditions. Error bars indicate 95% Confidence Intervals.

PAGE 29

2 9 Figure 1 ing toward Experimenter 1 in each session in Control Coding of Other Person condition.

PAGE 30

30 CHAPTER 2 EXPERIMENT 2: OPERANT CONTROL OF GAZING IN WOLVES Methods Subjects Eight captive wolves ( C l lupus ) were studied at Wolf Park, Battle Ground, IN. They range d in age from 11 months to 12 years at the time of data collection (Table 2), and were hand reared from 10 14 days of age and extensively socialized by experienced human caretakers using the methods described in Klinghammer & Goodman (1987). All subjects h ad varying amounts of experience with unrelated behavioral tests. Procedures Only Other Person and Object conditions were tested with wolves due to the orienting toward the source of reinforcement. Also, seven of the eight wolf subjects had condition a year prio r to data collection. The procedures were as identical as possible to those used with the dogs with exceptions due to the practicalities of the facilities and necessary safety precautions. Tests were conducted outdoors in physical isolation from other wol ves but with visual, auditory, and olfactory access to the other wolves and people at the park. For five resident enclosure, which consisted of a long, narrow space measuring 3 m wide and 10 m long. For the other three subjects, the experiment took place in their individual

PAGE 31

31 resident enclosures measuring approximately 15 m by 15 m where they were housed alone. The individuals acting as Experimenter 1 and 2 were familiar to the w olves and stood inside the testing enclosure against a fence, while the first author remained outside the enclosure and live coded responses. Two video cameras mounted on tripods outside the enclosure were used for recording the experiment. Pieces of summe r sausage (about 0.5 g each) were used as the reinforcers, and they were away from Experimenter 1 but inside the testi ng enclosure in case of an emergency situation and did not interact with the subject during the experiment. The stuffed animal approximately 15 cm from the fence. Inter Observe r Agreement As in the previous experiment, 20% of the videos were coded by an independent observer to calculate inter observer agreement. The reliability was 100% for Other Person and 87.5% for Object. Statistical Analysis We performed the same tests that were used to analyze the dog data from Experiment 1. Results Figure 2 1 shows the gazing durations for individual wolf subjects in Baseline and Test sessions for Other Person and Object conditions. The gazing durations increased from Baseline to Test for both groups, and the data for three of the four wolves in Object condition appear similar to the data for the wolves in Other Person. A two way ANOVA

PAGE 32

32 with repeated measures on factor Session (Baseline vs. Test) and between subjects factor Condition (Other person vs. Object) confirmed a significant main effect of session, F(1, 6) = 40.70, p < .001. There was no significant effect of condition, F(1, 6) = 4.03, p > .05, nor any interaction between the factors, F(1, 6) = 0.81, p > .05. The frequency of gazing increased during the Acquisition phase for subjects in both conditions (Figure 2 2). A two way ANOVA with within subjects factor Session (Acquisition 1 5) and between subjects factor Condition (Other Person vs. Object) confirmed significant main effects of Session, F(3.28, 19.67) = 8.35, p = .001 with Greenhouse Geisser adjustment, and Condition, F(1, 6) = 12.51, p < .05. However, there was no interaction between the factors, F(3.28, 19.67) = 0.54, p > .05 with Greenhouse Geisser adjustment. A test for lin ear trend across Sessions in the Acquisition phase revealed significant effects for Other Person, t(3) = 3.56, p < .05, and Object, t(3) = 3.70, p < .05. The duration of gazing decreased during the Extinction phase for subjects in the Object condition. T he subjects in Other Person decreased their duration of gazing across the first four sessions but showed a small increase in Extinction 5, although the mean of the final session remained below that of the first session of Extinction (Figure 2 3). A two way ANOVA with within subjects factor Session (Extinction 1 5) and between subjects factor Condition (Other Person vs. Object) was conducted to assess changes in gazing duration during the Extinction phase. There was a significant main effect of Session, F(2. 13, 12.77) = 15.47, p < .001 with Greenhouse Geisser adjustment. However, there was no effect of Condition, F(1, 6) = 2.66, p > .05. There was an interaction between the factors, F(2.13, 12.77) = 4.15, p < .05 with Greenhouse

PAGE 33

33 Geisser adjustment. A test for trend across Sessions in the Extinction phase revealed significant effects for linear trend in Object, t(3) = 4.31, p < .05, and a significant cubic trend for Other Person, t(3) = 6.72, p < .01. Following Extinction, gazing was re acquired during the Re acquisition phase for subjects in both conditions (Figure 2 2). A two way ANOVA with within subjects factor Session (Acquisition 1 vs. Re acquisition 1) and between subjects factor Condition (Other person vs. Object) confirmed significant main effects of Session, F(1, 6) = 10.64, p < .05, and Condition, F(1, 6) = 53.59, p < .001. There was no interaction between the factors, F(1, 6) = 5.43, p > .05. Discussion Similar to the results in Experiment 1, the gazing durations and frequencies for the wolf subjec ts increased and decreased in the appropriate phases as predicted. Acquisition and Re acquisition showed an effect of condition, with the frequency of gazing in Object condition remaining higher than that of Other Person condition. The individual data are shown for the Baseline Test comparison to demonstrate that the durations of gazing were similar among the subjects in both conditions except for one individual wolf in Object condition, who gazed considerably longer than the other subjects in either condit ion. The apparently higher rate of gazing for the Object group is clearly caused by the elevated level of one subject, highlighting the problem with analyzing group averages using small numbers of subjects. In spite of this, there was consistency in the d ata in terms of the trend followed by all subjects across sessions, suggesting tight experimental control. Furthermore, because each individual subject served as its own control in a reversal design (i.e. Acquisition, Extinction, Re Acquisition), the data could

PAGE 34

34 be visually analyzed for individual subjects in addition to statistical analysis by group. Regardless, the generality of these results can only be established via replication with more subjects. Another limitation of this study is that some of the c hanges to the experimental set up due to the nature of working with wolves versus dogs prevent a fair direct comparison between the two canid types. For example, the wolves may have had more environmental distractions during the experiment than the dogs be cause they were tested outdoors, with birds flying overhead, other wolves in sight, extra humans present, and occasional bouts of howling by all animals in the park. This being said, both canid types were tested in settings to which they were more accustom ed, so testing the dogs outdoors in an attempt to equalize the distraction factor would have put them at an unfair advantage since they typically spend more time indoors. General Discussion Every interaction an individual has with its environment has the p otential to reinforce or punish some behavior (Skinner, 1953). If a dog looks at a human and subsequently receives reinforcement in the form of a treat, a scratch behind the ears, or verbal attention (which previously could have been paired with a primary reinforcer such as treats), then the dog will be more likely to look at the human again in the future. Because pet dogs experience extensive interactions with humans starting at birth, they are likely to have a long history of reinforcement for paying atte ntion to humans throughout their lives. At the same time, dogs could have a weaker history of reinforcement for looking at an inanimate object if reinforcement seldom follows this behavior. Even though the wolves used in this study are socialized to humans their

PAGE 35

35 individual experiences undoubtedly differ from those of pet dogs, resulting in potentially different gazing patterns. In 2005, Gasci et al. conducted a test similar to the current experiment on wolf and dog pups at five and nine weeks of age, whe re a plate of meat was placed in sight but out of reach, and gazes to the experimenter were reinforced with food. During the four minute session, the dog pups increased their rate of gazing, but the wolf pups did not. This stands in stark contrast to the r esults of the current study, which demonstrated clearly that wolves can learn to gaze at humans and other targets. This difference may be because adult wolves were studied rather than pups. Another possibility is that the wolf pups may have required more t ime and chances to contact the contingency of reinforcement than the dog pups, which would have resulted in acquisition of gazing but at a slower rate. This seems unlikely to be the only factor because the dog and wolf subjects in the current study show si milar rates of acquisition for the first two session (i.e. duration of training session in Gasci et al. study). Using a paradigm designed to compare the gazing behavior of dogs and wolves, Miklosi et al. (2003) presented subjects with physical tasks that a t first they could solve, but subsequently were made unsolvable. The behavior of interest was turning and looking back at the owner, which was interpreted as soliciting help. They found that dogs had a shorter latency and a higher duration of looking at th e owner compared to wolves, which led the authors to conclude that there is a genetic difference in dogs and interpretation is that the effects of environmental influences precedi ng the experiment were not considered. If the dogs in the study had extensive experience with their

PAGE 36

36 owners solving difficult tasks for them, such as opening food containers, while the wolves had more experience destroying containers and obtaining food on t heir own, then these results would not be surprising. The dogs would have learned that looking to the owner results in reinforcement, while the wolves learned that persistence pays off. Furthermore, Marshall Pescini, Passalacqua, Barnard, Valsecchi, and P rato Previde and rescue, and no training and found that those trained in agility, which relies heavily on watching human cues, gazed at the human longer in the unsolvab le task. Viranyi et al. (2008) measured latency to establish eye contact in dog and wolf subjects being tested on an object choice task. When tested at four months old, dogs had a shorter latency to gaze than wolves of the same age, even though the authors report that the dog subjects were reared under similar conditions of socialization. behavior; rather, it is also the specific experience each individual encounters. For example, caretakers dropping food or toys or playing fetch with dogs but not with wolves received formal training on following point types and were tested again at 11 months contact were the same as for the dogs, suggesting that experience does matter. The current study is not suggesting that dogs are explicitly trained to gaze at humans. Instead, gazing may arise if owners (consciously or not) are more likely to provide reinforcers to the dog following gazes. Recent evidence has demonstrated that a dogs gaze releases oxytocin in the owner (Nagasawa, Kikusui, Onaka, & Ohta, 2009 ),

PAGE 37

37 and therefore could be reinforcing to an owner. Another possibility is that gazing could be adventitiously reinforced, merely coinciding temporally with the delivery of food or attention. Dogs may simply learn that paying attention to humans often resul ts in something rewarding for them. The fact that the dogs in the Feeding Person group exhibited a higher baseline of gazing than those in the Other Person group shows that the presence of food will necessarily result in more gazing at the nearest person, supporting the idea that reinforcement is an important factor. are sufficient to account for other types of social behaviors in dogs, such as point following (Elgier, Jakovcevi c, Barrera, Mustaca, & Bentosela, 2009; Udell, et al., 2010) and preferential begging from humans of varied attentional states (Udell, Dorey, & Wynne, 2011). Because gazing has been clearly shown to be an operant behavior and easily manipulable in a short period of time, the possibility that it is also shaped during with humans on a daily basis are most likely sufficient to account for gazing behavior, as they have with th e other social behaviors mentioned above, though future research should investigate this directly. The results of this study demonstrate that gazing responds to operant manipulations, suggesting that claims about the innate tendencies of dogs to gaze at hu mans may be premature. The fact that we can control gazing not only toward humans but also inanimate objects supports the idea that gazing is just as subject to contingencies of reinforcement as any other behavior, and dogs do not require a specially evolv ed connection with humans to exhibit high rates of gazing at them.

PAGE 38

38 Furthermore, the finding that wolves similarly show operant control of gazing casts doubt upon the assumption that gaze differences between dogs and wolves result solely from domestication Attempts to explain behavior in any organism must consider individual histories as well as genetic influences if we are to gain a better understanding of the effects of domestication in general and the dynamic interactions between humans and dogs.

PAGE 39

39 Table 2 1 Wolf Subject Characteristics Name Age Sex Condition Wolfgang 6 yrs M Other Person Erin 1 3 yrs F Other Person Dharma 1 yrs F Other Person Ruedi 7 yrs M Other Person Renki 7 yrs M Object Kailani 7 yrs F Object Ayla 7 yrs F Object Wotan 6 yrs M Object

PAGE 40

40 Figure 2 1. Gazing durations for individual wolf subjects in Baseline and Test sessions for Other Person and Object conditions.

PAGE 41

41 Figure 2 2. Mean number of gazes in each session for wolves in Other Person and Object conditions. Error bars indicate 95% Confidence Intervals.

PAGE 42

42 Figure 2 3. Mean duration of gazing during the Extinction phase for wolf subjects in the Other Person and Object conditions. Error bars indicate 95% Confiden ce Intervals.

PAGE 43

43 LIST OF REFERENCES Barrera, G., Mustaca, A., & Bentosela, M. (2011) Communication between domestic dogs and humans: Effects of shelter housing upon the gaze to the human. Animal Cognition 14 727 734. Bentosela, M., Barrera, G., Jako vcevic, A., Elgier, A., & Mustaca, A. (2008) Effect of reinforcement, reinforcer omission and extinction on a communicative response in domestic dogs (Canis familiaris). Behavioural Processes 78 464 469. Bentosela, M., Jakovcevic, A., Elgier, A, Papin i, M., & Mustaca, A. (2009) Incentive contrast in domestic dogs ( Canis familiaris). Journal of Comparative Psychology 123 2, 125 130. Dorey, N. R., Rosales Ruiz, J., Smith, R., & Lovelace, B. (2009) Functional analysis and treatment of self injury i n a captive olive baboon. Journal of Applied Behavior Analysis 42 785 794. Dorey, N., Udell, M. A. R., & Wynne, C. D. L. (2010) When do domestic dogs, Canis familiaris start to understand human pointing? The role of ontogeny in the development of inte rspecies communication. Animal Behaviour 79 37 41. Elgier, A. M., Jakovcevic, A., Barrera, G., Mustaca, A. E., & Bentosela, M., (2009) Communication between domestic dogs ( Canis familiaris ) and humans: Dogs are good learners. Behavioural Processes 81 402 408. Feuerbacher, E., & Wynne, C. D. L. (2011) A History of Dogs as Subjects in North American Experimental Psychological Research. Comparative Cognition & Behavior Reviews, 6 46 71. Gacsi, M., Gyori, B., Miklosi, A., Viranyi, Z., Kubinyi, E., Topal, J., & Csanyi, V. (2005) Species specific differences and similarities in the behavior of hand raised dog and wolf pups in social situations with humans. Developmental Psychobiology 47 111 122. Hare, B., Brown, M., Williamson, C., & Tomasello, M (2002) The Domestication of Social Cognition in Dogs. Science 298 5598, 1634 1636. Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1982) Toward a functional analysis of self injury. Analysis and Intervention in Develop mental Disabilities 2 3 20. Jakovcevic, A, Elgier, A, Mustaca, A, & Bentosela, M. (2010) ( Canis familiaris ) gaze to the human face. Behavioural Processes 84 602 607. Kleinke, C. L. (1986) Gaze and eye contact: A researc h review. Psychological Bulletin 100 1, 78 100.

PAGE 44

44 Klinghammer, E., & Goodmann, P. A. ( 1987) Socialization and management of wolves in captivity In: Frank, H. (ed.) Man and Wolf: Advances, Issues, and Problems in Captive Wolf Research W. Junk Dordrech t, pp. 31 61 Marshall Pescini, S., Passalacqua, C., Barnard, S., Valsecchi, P., & Prato Previde, E. (2009) behaviour in socio cognitive tasks. Behavioural Processes 81 416 422. Mik losi, A., Kubinyi, E., Topal, J., Gacsi, M., Viranyi, Z., & Csanyi, V. (2003) A simple reason for a big difference: Wolves do not look back at humans, but dogs do. Current Biology 13 763 766. Nagasawa, M., Kikusui, T., Onaka, T., & Ohta. (2009) Dog' s gaze at its owner increases owner's urinary oxytocin during social interaction. Hormones and Behavior 55 434 441. Riedel, J., Schumann, K., Kaminski, J., Call, J., & Tomasello, M. (2008) The early ontogeny of human dog communication. Animal Behavio ur 75 1003 1014. Skinner, B. F. (1953) Science and Human Behavior. MacMillan, New York, pp. 90 100. Udell, M. A. R., Dorey, N. R., & Wynne, C. D. L. (2008) Wolves outperform dogs in following human social cues. Animal Behaviour 76 1767 1773. Udell M. A. R., Dorey, N. R., & Wynne, C. D. L. (2010) The performance of stray dogs ( Canis familiaris ) living in a shelter on human guided object choice tasks. Animal Behaviour 79 717 725. Udell, M. A. R., Dorey, N. R., & Wynne, C. D. L. (2011) Can your dog read your mind? Understanding the causes of canine perspective taking. Learning & Behavior 39 4, 289 302. Udell, M. A. R., & Wynne, C. D. L. (2008) Canis familiaris ) human like behaviors: Or why behavior analysts shou ld stop worrying and love their dogs. Journal of the Experimental Analysis of Behavior 89 247 261. Viranyi, Z., Gacsi, M., Kubinyi, E., Topal, J., Belenyi, B., Ujfalussy, D., & Miklosi, A. (2008) Comprehension of human pointing gestures in young human reared wolves ( Canis lupus ) and dogs ( Canis familiaris ). Animal Cognition 11 373 387. Wynne, C. D. L., Udell, M. A. R., & Lord, K. (2008) dog communication. Animal Behaviour 76 e1 e4.

PAGE 45

45 BIOGRAPHICAL SKETCH Jessica Marie Spencer was born in Granbury, Texas, as the youngest of six children. She grew up mostly in Pennsylvania, but graduated from Watkins Mill High School in Gaithersburg, Maryland, in 2004. She received her Bachelor of Science degree in p sychology at Stetson University of Deland, Florida, in 2008, graduating Magna c um Laude During her time at Stetson, she also spent a semester at the University of Edinburgh in the study abroad program. She received her Master of Science degree from the Department of P sychology in the behavior analysis area at t he University of Florida in 2012