Dogs do look at images: eye tracking in canine cognition ... - Springer

Anim Cogn (2012) 15:163–174DOI 10.1007/s10071-011-0442-1ORIGINAL PAPERDogs do look at images: eye tracking in canine cognition researchSanni Somppi • Heini Törnqvist • Laura Hänninen •Christina Krause • Outi VainioReceived: 11 February 2011 / Revised: 6 July 2011 / Accepted: 1 August 2011 / Published online: 23 August 2011Ó Springer-Verlag 2011Abstract Despite intense research on the visual communicationof domestic dogs, their cognitive capacitieshave not yet been explored by eye tracking. The aim of thecurrent study was to expand knowledge on the visualcognition of dogs using contact-free eye movement trackingunder conditions where social cueing and associativelearning were ruled out. We examined whether dogsspontaneously look at actual objects within pictures andcan differentiate between pictures according to their noveltyor categorical information content. Eye movements ofsix domestic dogs were tracked during presentation ofdigital color images of human faces, dog faces, toys, andalphabetic characters. We found that dogs focused theirattention on the informative regions of the images withoutany task-specific pre-training and their gazing behaviordepended on the image category. Dogs preferred the facialimages of conspecifics over other categories and fixated ona familiar image longer than on novel stimuli regardless ofthe category. Dogs’ attraction to conspecifics over humanfaces and inanimate objects might reflect their naturalinterest, but further studies are needed to establish whetherdogs possess picture object recognition. Contact-free eyemovement tracking is a promising method for the broaderexploration of processes underlying special socio-cognitiveskills in dogs previously found in behavioral studies.S. Somppi (&) L. Hänninen O. VainioFaculty of Veterinary Medicine, University of Helsinki,P. O. Box 57, 00014 Helsinki, Finlande-mail: sanni.somppi@helsinki.fiH. Törnqvist C. KrauseFaculty of Behavioral Sciences, University of Helsinki,P. O. Box 9, 00014 Helsinki, FinlandKeywords Canine Domestic dogs Eye movementtracking Visual cognitionIntroductionClose co-evolution, sharing the same living habitat, andsimilarities in canine and human socio-cognitive skillsmake the dog a unique model for comparative cognitionstudies (Hare and Tomasello 2005; Miklósi et al. 2007;Topál et al. 2009). Domestication has equipped dogs withsophisticated sensitivity to respond to human visual communicativecues, including gaze direction and pointinggestures (Hare and Tomasello 2005; Miklósi et al. 2003;Soproni et al. 2002; Virányi et al. 2004). Despite theintense research around the visual communication ofdomestic dogs, their cognitive capacities have not yet beenexplored by eye tracking.Animals with developed visual systems control theirgaze direction using eye movements (Land 1999). Eyemovements are linked to attention (Buswell 1935); thevisual-cognitive system directs a gaze toward importantand informative objects, and vice versa, gaze directionaffects several cognitive processes (Henderson 2003).Based on behavioral observations, dogs can perceive twodimensionalvisual information and can be trained to performvisual tasks, such as classifying photographs of naturalstimuli by means of a perceptual response rule (Rangeet al. 2008). They can correlate the photographs and voiceof their owner (Adachi et al. 2007) and are able to distinguishthe facial images of two individual humans or dogs(Racca et al. 2010).However, in previous experiments, dogs have beeneither manually restrained (Adachi et al. 2007; Guo et al.2009; Racca et al. 2010) or their choices were reinforced123

164 Anim Cogn (2012) 15:163–174(Range et al. 2008), which may have affected the outcome.Dogs are highly sensitive and respond to humangestures and attentional states (reviewed in Miklósi et al.2007; Topáletal.2009), so the owner or experimentermay even subconsciously influence the dog’s expectationsand behavior; the phenomenon called Clever Hanseffect (Pfungst 1907). In the previous studies, this effecthas been minimized by avoiding the human interferencewith the dogs or leaving the restrainer ignorant of thepurpose of the experiment (Adachi et al. 2007; Guo et al.2009; Racca et al. 2010). The Clever Hans effect mightbe particularly prevalent in dogs (Lit et al. 2011); thus,tasks that are performed without human presence couldbe useful.In traditional visual discrimination tests the lookingbehavior is measured manually from the orientationbehavior (Adachi et al. 2007; Guo et al. 2009; Racca et al.2010; Faragó et al. 2010), which may include both bouts ofactive information processing and blank stares (Aslin2007). When only behavioral measures are used, it isimpossible to define which features the dogs’ attention isdrawn to. Is the dog actually looking at the picture? Eyemovement tracking, a technique for directly assessinggazing behavior, enables measurement of the sub-elementsof the looking behavior, which likely indicate the underlyingmechanisms of the performance better than traditionalparadigms (Quinn et al. 2009).Eye tracking provides an objective method for understandingthe ongoing cognitive and emotional processes,such as visual-spatial attention, semantic information processing,and motivational states of animals (Henderson2003; Kano and Tomonaga 2009). Eye movement trackingin dogs has been used to diagnose ocular motor abnormalitiessuch as nystagmus (Jacobs et al. 2009; Dell’Ossoet al. 1998), and a head-mounted eye-tracking camera hasbeen recently tested on a single dog (Williams et al. 2011),but to date, the method has not been utilized in caninecognitive studies. Although dogs’ visual acuity is lowerthan that of humans and they have comparatively reducedcolor perception (dichromatic color vision), the visualsystem of the dog does not impose crucial limitations onthe use of the eye movement tracking method. The canineoptic chiasm has a crossover of about 75%, which allowsfor good binocular vision (Miller and Murphy 1995). Dogscan detect details, even in very small photographs presentedat a short distance (Range et al. 2008).The spontaneous visual discrimination abilities of humaninfants and non-human primates have traditionally beentested with novelty preference paradigms (Fantz 1964;Joseph et al. 2006; Dahl et al. 2007; Colombo and Micthell2009). In a commonly used protocol, a single stimulus isfirst presented repeatedly (familiarization) and then changedto a novel stimulus. If the subject discriminates thesimilarity of the repeating presentation, there is a decline inlooking, and a rebound in looking occurs when the stimulusis shifted if the subject discriminates a novel stimulus froma familiar one (Houston-Price and Nakai 2004). Recently, ithas been demonstrated that dogs exhibit an orientingbehavior toward novel pictures, but their preferences aredependent on image category (Racca et al. 2010). Thisfinding indicates that dogs may have the ability to categorizevisual objects. Forming categories by sorting objectsinto associative classes is one of the fundamental cognitiveabilities of humans, but little is known about spontaneouscategorization in animals (Murai et al. 2005).The purpose of our study was to expand knowledge onthe cognitive abilities of dogs using a new method, contactfreeeye movement tracking. Our aim was to assess howdogs view two-dimensional photographs under free viewingconditions where social cueing is ruled out. To studythis, we tested whether dogs differentiate between picturesaccording to their novel or categorical information contentwithout any task-specific pre-training.Materials and methodsAll experimental procedures were approved by the EthicalCommittee for the use of animals in experiments at theUniversity of Helsinki.Animals and experimental setupSix privately owned dogs, aged 1–5 years (Table 1), participatedin the study. All experiments were conducted atthe Veterinary Faculty of the University of Helsinki.Prior to conducting the experiments (1–2 months), thedogs were trained by their owners as instructed by the firstauthor. Dogs were trained using an operant-positive conditioningmethod (clicker) to lie on a 10-cm-thick Styrofoammattress and lean their jaw on a purpose-designedu-shaped chin rest for up to 60 s (see Fig. 1 for experimentaldetails). During the last 2–3 weeks of the trainingperiod, the dogs and their owners visited the experimentalTable 1 Breed, age, and gender of the dogs participating in theexperimentID Breed Age (years) GenderD1 Beauce shepherd 1 FemaleD2 Beauce shepherd 2 FemaleD3 Beauce shepherd 2.5 FemaleD4 Rough collie 1 FemaleD5 Hovawart 4 CastratedD6 Great pyreness 5 Female123

170 Anim Cogn (2012) 15:163–174Fig. 4 An example of eye gaze patterns during presentation of theDOG image (a) and the BLANK screen (b). The focus map representsaveraged fixations of the right eye of five dogs and three repetitionspresented consecutively. The color coding represents the average offixation durations; minimum 5 ms indicated by light blue and themaximum of 100 ms or over by bright red. The area corresponding tothe image size and placement is overlaid on the BLANK screen as adashed lineFig. 5 The effect of novelty status of the image on the number offixations per frame (mean ± SEM) and the total duration of fixationsper frame (mean in ms ± SEM) in six dogs. In the novelty paradigm,the same stimulus image repeats 3–5 times (last frame is consideredas FAMILIAR) and then changes (SHIFT) to another stimulus imagefrom the same stimulus category (NOVEL). Statistically significantdifferences between the 1st, FAMILIAR, and NOVEL frames arepresented as different letters (MIXED, P \ 0.05)LETTER n = 113) were found for the number of fixations,the total duration of fixations, and the relative fixationduration (Table 3). The dogs fixated on DOG images moreoften (2.0 ± 0.3) than on HUMAN (1.6 ± 0.3, P = 0.014),ITEM (1.2 ± 0.3, P = 0.006), or LETTER images(0.5 ± 0.5, P = 0.002). LETTER images gathered fewerfixations than other categories (LETTER vs. DOG P =0.000; LETTER vs. HUMAN P = 0.025; LETTER vs.ITEM P = 0.043; Fig. 7).The mean duration of a single fixation was 214 ms (95%CI 154–289 ms) on average and did not differ between thecategories. However, the total fixation duration among thefour image categories differed statistically significant(P = 0.000; Table 3). Dogs fixated longest on the imagesFig. 6 The effect of the repeated image presentation on a the numberof fixations (mean ± SEM) and the duration of a single fixation(mean in ms ± SEM) per frame b the total duration of fixations perframe (mean in ms ± SEM) and the relative fixation duration of thefixations targeted to the object (% ± SEM) in six dogs during anexperimental trial where a total of six frames of two different imagesfrom the same stimulus category are shown as a novelty paradigm.Statistically significant differences between the frames are presentedas different letters (MIXED, P \ 0.05)of dogs (534 ± 80 ms) and shortest (94 ± 120 ms) on theimages of alphabetic characters (DOG vs. HUMANP = 0.024; DOG vs. ITEM P = 0.001; DOG vs. LETTER123

Anim Cogn (2012) 15:163–174 171Fig. 7 The effect of the image category (DOG, HUMAN, ITEM, andLETTER) on the number of fixations (mean ± SEM) and the totalduration of fixations (mean in ms ± SEM) per frame in six dogs.Statistically significant differences between the image categories arepresented as different letters (MIXED, P \ 0.05)P = 0.000; LETTER vs. HUMAN P = 0.005; LETTERvs. ITEM P = 0.029; Fig. 7). The main effect of the imagecategory on the relative fixation duration of the object wasstatistically significant (P = 0.042), but the pairwisecomparisons did not specify which categories differed fromeach other (DOG 65.4 ± 6.4%; HUMAN 56.2 ± 6.7%;ITEM 60.4 ± 8.4%; and LETTER 39.8 ± 13.3%).DiscussionThe current study produced evidence on the canine visualcognition by a new method: a contact-free eye movementtracking. Dogs focused their attention on the informativeregions of the images without any task-specific pre-training,and their gazing behavior depended on the imagecategory.Dogs focused most of their fixation duration on theactual image compared with the surrounding monitor andon the actual object compared with the background image,as it has previously been reported for humans, chimpanzees,and monkeys (Yarbus 1967; Nahm et al. 1997; Kanoand Tomonaga 2009). Dogs spontaneously prefer imagesof conspecifics over human faces and inanimate objects,suggesting they might be able to discriminate images ofdifferent categories.Various animal species can form visual categories whentrained using a match-to-sample procedure (Bovet andVauclair 2000). However, animals quickly learn to repeatthe reinforced behaviors (Skinner 1938), even uncharacteristicones, and hence, the natural looking behavior couldremain hidden (Dahl et al. 2007). Explicit rewarding forcertain pre-defined criterion could lead to atypical responsestrategies and limit the comparability between studies donewith naïve human subjects (Murai et al. 2005; Dahl et al.2009). In the current study, the differences between stimuluscategories arise from unprompted attention measureddirectly as fixational eye movements.The categories gathered different numbers of fixations,while the average duration of a single fixation was the samein all categories. The role of the fixation is to keep the gazestable enough for stimulus encoding in the photoreceptors(Land 1999; Yarbus 1967); thus, a sufficiently long durationfor single fixation is needed to identify the object. Inhumans, targets interpreted as informative or interestingattract more re-fixations (Buswell 1935; Henderson andHollingworth 1999). In the current study, the images ofletters received the least number of fixations and thus theshortest overall fixation duration. The finding correspondswith behavioral observations in a recent study of Faragóet al. (2010) who suggested that dogs might consider naturalobjects more interesting than abstract ones. On theother hand, these complex pictures might contain moreinformation to process. The lowest level for the pictureperception can occur on the basis of physical features (i.e.,color, intensity, contrasts, or edge orientation), which doesnot require the understanding the representational contentof the image (Fagot et al. 1999, Bovet and Vauclair 2000).In eye movements, the targets of the fixations could bedriven not only by higher level conceptual information ofthe objects but also by low-level visual saliency (Henderson2003; Henderson and Hollingworth 1999). In fact, themore complex the image, the more it contains details thatmight attract fixations and hence increase the time it takesto view and process the image. Therefore, the categorydependentgazing behavior could also be the result of differencesin physical complexity. In humans, the high- andlow-level mechanisms of the guidance of eye movementsare dependent on the task and could alternate by turns(Einhäuser et al. 2008a, b). Thus, based only on eyemovement data, we cannot yet draw any conclusions as towhether the attention of dogs was mainly directed bystimulus features or semantic information, or both.The letter images were much simpler in their featuresthan were the other categories, but the images of dog faceswere not apparently more complex from their physicalproperties than presented human faces. However, dogsfixated for more times and longer total duration on caninefaces than on human faces. A preference for conspesificfaces has been suggested to indicate expertise in the perceptionof the faces of own species (Dahl et al. 2009;Hattori et al. 2010). Recent behavioral studies in dogs alsofound species-dependent looking behavior when viewinghuman and canine faces (Guo et al. 2009; Racca et al.2010). Face detection plays an important role in non-verbalcommunication in primates and probably also in other123

172 Anim Cogn (2012) 15:163–174social mammals (Leopold and Rhodes 2010). The dogmight perceive conspecific faces to be more interesting orinformative than human faces due to different social relevanciesand communication strategies between dogs anddogs, and dogs and humans. One factor that might haveaffected category preferences is a variation within a category;the dog faces represent many breeds, while humanfaces were all Caucasian.As has been found in human infants (Fantz 1964;reviewed in Colombo and Micthell 2009) and monkeys(Joseph et al. 2006), the first frame attracted the highesttotal looking time, which decreased when the image wasrepeated, probably indicating habituation to the stimulus.However, we could not detect expected rebound in lookingwhen the novel picture was presented. Instead, the totalduration of fixations decreased after the stimulus changed.It is likely that dogs did not notice the change or they mayhave generalized familiar pictures to novel ones. In generalization,two stimuli, even if readily distinguishable, aresimilar enough with respect to their physical properties toevoke the same response (Shepard 1994, Chirlanda andEnqvist 2003; generalization of natural 2D pictures hasbeen demonstrated in dogs by Range et al. 2008). Most ofthe familiar and the familiar–novel pairs were rather similar;for example, humans were paired according to gender,age, hair color, and expression. When the stimuli arecomplex, dogs’ performance could be limited as comparedwith other species with better visual acuity and color perception(Range et al. 2008). It is possible that the dogs donot perceive the minor details of the picture and thuscannot discriminate the images within the same category.An interaction between category and novelty status wasnot established, which is partly contradictory to the recentstudy of Racca et al. (2010) in which dogs also directedshorter looking times to novel canine faces, but longerdurations at novel human faces and objects. Categorydependentnovelty responses have also been reported forhuman observers, who prefer familiarity in human facesbut novelty in natural scenes, and had no clear preferencefor geometric figures (Park et al. 2010). The differentresults may also be due to many methodological reasons;we examined fixational eye movements, not overall orientingtime as Racca et al. (2010). The overall orientingtime includes also the inactive viewing (‘‘blank stares’’;Aslin 2007). Moreover, the presentation setup differed. Inthe current study, the design was a modified version of thatused for monkeys (Joseph et al. 2006). The stimulus wasrepeated several times, and dogs were rewarded after thesixth frame. All dogs were highly motivated, but thedecrease in the number and total duration of fixationsduring the last frames of the trial suggest that they mighthave got tired of the monotony of the task. It is also likelythat they anticipated the reward and therefore were notfocusing their attention at the monitor at the end of the trial.Varying the length of trials could have represented a betterdesign for canine research. Dogs might detect the differencebetween familiar and novel images better if theimages are presented side by side as in visual pairedcomparison study of Racca et al. (2010). Different methodologieshave led to contradicting novelty responses alsoin infant studies (Houston-Price and Nakai 2004).The absence of preferential looking does not necessarilymean the absence of discrimination (Aslin 2007). Noveltypreferences can vary individually; the subjects might haveperceived certain categories more attractive or might haveused individual strategies for detecting a new stimulus, assuggested previously for monkeys and humans (Josephet al. 2006). Preference may even focus on certain details ofthe object, for example, the head area in whole-body picturesQuinn et al. (2009). Also the physical similarity anddifferences in attractiveness between consecutive imagescould affect novelty responses. A rebound in looking islikely to occur when consecutive images differ more fromeach other and when the novel image is more interesting incontent than the previous one (Dahl et al. 2007).The dogs in the current study performed the tasksindependently while the owners and experimentersremained hidden behind an opaque barrier. The dogs weretrained neither to fixate on the monitor nor to discriminateimages. However, all measured eye movement variablesindicated that the dogs were more interested in looking atthe monitor when images were displayed than when thescreen was blank. This finding confirms that the dogs hadnot learned to fix their gaze on the monitor in anticipationof a reward or a response to social cueing. To the authors’knowledge, blank screen viewing has not been previouslymeasured in animal studies. Dogs targeted some fixationstoward the blank screen, which is typical also in humanlooking behavior because the re-activation of memoryrepresentations drives the eyes to previously viewed locations(Ferreira et al. 2008).Clearly, we cannot yet say unequivocally what the dogssee in the pictures. In humans, a picture is something inwhich objects can be recognized, even though the objectsthemselves are not actually present. It is under debatewhether animals recognize pictures as representations ofreal-world objects (Bovet and Vauclair 2000; Jitsumori2010). Dogs can associate visual image with acousticalinformation, suggesting they are capable of forming mentalrepresentations through pictures (Adachi et al. 2007, Faragóet al. 2010). It has also been demonstrated through afetching task that at least some dogs are able to matchphotographs of items to actual objects (Kaminski et al.2009). Nevertheless, eye movement tracking is a promisingmethod for comparing visual perception strategies andabilities between humans and dogs.123

Anim Cogn (2012) 15:163–174 173In conclusion, contact-free eye movement tracking canbe used to assess canine visual cognition. This promisingmethod represents a tool for the broader exploration ofprocesses underlying special socio-cognitive skills in dogspreviously established through behavioral studies. Dogs’attraction to conspecific and human faces over inanimateobjects might reflect the natural interests of dogs, but furtherstudies are needed to establish if dogs possess pictureobject recognition.Acknowledgments This work was financially supported by theAcademy of Finland and University of Helsinki. The authors aregrateful to Antti Flyckt, Matti Pastell, Aleksander Alafuzoff, TeemuPeltonen, Jaana Simola, Timo Murtonen, and Kristian Törnqvist fortheir support in conducting the experiment. Authors also thank IKEAgroup for the permission to use the photos of children’s toys.Conflict of interestof interest.ReferencesThe authors declare that they have no conflictAdachi I, Kuwahata H, Fujita K (2007) Dogs recall their owner’s faceupon hearing the owner’s voice. Anim Cogn 10:17–21Aslin RN (2007) What’s in a look? Dev Sci 10:48–53Bovet D, Vauclair J (2000) Picture recognition in animals andhumans. Behav Brain Res 109:143–165Buswell GT (1935) How people look at pictures; a study of thepsychology of perception in art. University of Chicago Press,ChicagoChirlanda S, Enqvist M (2003) A century of generalization. AnimBehav 66:15–36Colombo J, Micthell DW (2009) Infant visual habituation. NeurobiolLearn Mem 92:225–234Dahl CD, Logothetis NK, Hoffman KL (2007) Individuation andholistic processing of faces in rhesus monkeys. Proc Rev Soc BBiol Sci 274:2069–2076Dahl CD, Wallraven C, Bülthoff HH, Logothetis NK (2009) Humansand macaques employ similar face-processing strategies. CurrBiol 19:509–513Dell’Osso LF, Williams RW, Jacobs JB, Erchul DM (1998) Thecongenital and see-saw nystagmus in the prototypical achiasmaof canines: comparison to the human achiasmatic prototype. VisRes 38:1629–1641Einhäuser W, Rutishauser U, Koch C (2008a) Task-demands canimmediately reverse the effects of sensory-driven saliency incomplex visual stimuli. J Vis 8:1–19Einhäuser W, Spain M, Perona P (2008b) Objects predict fixationsbetter than early saliency. J Vis 8:1–26Fagot J, Martin-Malivel J, Dépy D (1999) What is the evidence for anequivalence between objects and pictures in birds and nonhumanprimates? Curr Psychol Cogn 18:923–949Fantz RL (1964) Visual experience in infants: decreased attention tofamiliar patterns relative to novel ones. Sci 146:668–670Faragó T, Pongràcz P, Miklósi Á, Huber L, Virányi Z, Range F(2010) Dogs’ expectation about signalers’ body size virtue oftheir growls. PLoS One 12:1–8Ferreira F, Apel J, Henderson JM (2008) Taking a new look atlooking at nothing. Trends Cogn Sci 12:405–410Guo K, Meints K, Hall C, Hall S, Mills D (2009) Left gaze bias in humans,rhesus monkeys and domestic dogs. Anim Cogn 12:409–418Hare B, Tomasello M (2005) Human-like social skills in dogs?Trends Cogn Sci 9:439–444Hattori Y, Kano F, Tomonaga M (2010) Differential sensitivity toconspecific and allospecific cues in chimpanzees and humans: acomparative eye-tracking study. Biol Lett 6:610–613Henderson JM (2003) Human gaze control during real-world sceneperception. Trends Cogn Sci 7:498–504Henderson JM, Hollingworth A (1999) High-level scene perception.Annu Rev Psychol 50:243–271Houston-Price C, Nakai S (2004) Distinguishing novelty andfamiliarity effects in infant preference procedures. Infant ChildDev 13:341–348Jacobs JB, Dell’Osso LF, Wang ZI, Acland GM, Bennett J (2009)Using the NAFX to measure the effectiveness over time ofgene therapy in canine LCA. Invest Ophthalmol Vis Sci50:4685–4692Jitsumori M (2010) Do animals recognize pictures as representationsof 3D objects? Comp Cogn Behav Rev 5:136–138Joseph JE, Powell DK, Andersen AH, Bhatt RS, Dunlap MK, FoldesST, Forman E, Hardy PA, Steinmetz NA, Zhang Z (2006) fMRIin alert, behaving monkeys: an adaptation of the human infantfamiliarization novelty preference procedure. J Neurosci Methods157:10–24Kaminski J, Tempelmann S, Call J, Tomasello M (2009) Domesticdogs comprehend human communication with iconic signs. DevSci 12:831–837Kano F, Tomonaga M (2009) How chimpanzees look at pictures: acomparative eye-tracking study. Proc Biol Sci 276:1949–1955Land MF (1999) Motion and vision: why animals move their eyes.J Comp Physiol A Neuroethol Sens Neural Behav Physiol185:341–352Leopold DA, Rhodes G (2010) A comparative view of faceperception. J Comp Psychol 124:233–251Lit L, Schweitzer JB, Oberbauer AM (2011) Handler beliefs affectscent detection dog outcomes. Anim Cogn 14:387–394Miklósi Á, Kubinyi E, Topál J, Gacsi M, Viranyi Z, Csanyi V (2003)A simple reason for a big difference: wolves do not look back athumans, but dogs do. Curr Biol 13:763–766Miklósi Á, Topál J, Csányi V (2007) Big thoughts in small brains?dogs as a model for understanding human social cognition.Neuroreport 18:467–471Miller PE, Murphy CJ (1995) Vision in dogs. J Am Vet Med Assoc207:1623–1634Murai C, Kosugi D, Tomonaga M, Tanaka M, Matsuzawa T, Itakura S(2005) Can chimpanzee infants (Pan troglodytes) form categoricalrepresentations in the same manner as human infants (Homosapiens)? Dev Sci 8:240–254Nahm FKD, Perret A, Amaral DG, Albright TD (1997) How domonkeys look at faces? J Cogn Neurosci 9:611–623Park J, Shimojo E, Shimojo S (2010) Roles of familiarity and noveltyin visual preference judgments are segregated across objectcategories. Proc Natl Acad Sci USA 107:14552–14555Pfungst O (1907) Das Pferd des Herrn von Osten (der Kluge Hans):Ein Beitrag zur experimentellen Tier-und Menchenpsychologie.Johann Ambrosius Barth, LeipzigQuinn PC, Doran MM, Reiss JE, Hoffman JE (2009) Time course ofvisual attention in infant categorization of cats versus dogs:evidence for a head bias as revealed through eye tracking. ChildDev 80:151–161Racca A, Amadei E, Ligout S, Guo K, Meints K, Mills D (2010)Discrimination of human and dog faces and inversion responsesin domestic dogs (Canis familiaris). Anim Cogn 13:525–533Range F, Aust U, Steurer M, Huber L (2008) Visual categorization ofnatural stimuli by domestic dogs. Anim Cogn 11:339–347Shepard R (1994) Perceptual-cognitive universals as reflections of theworld. Psychon Bull Rev 1:2–28123

174 Anim Cogn (2012) 15:163–174Skinner (1938) The behavior of organisms: an experimental analysis.D. Appleton-century company. New York, p 457Soproni K, Miklósi Á, Topál J, Csányi V (2002) Dogs’ (Canisfamiliaris) responsiveness to human pointing gestures. J CompPsychol 116:27–34Topál J, Miklósi Á, Gácsi M, Dóka A, Pongrácz P, Kubinyi E,Virányi Z, Csányi V (2009) The dog as a model forunderstanding human social behavior. In: Brockmann HJ, RoperTJ, Naguib M, Wynne-Edwards KE, Mitani JC, Simmons LW(eds) Advances in the study of behavior, vol 39. Academic Press,Burlington, pp 71–116Virányi Z, Topál J,Gácsi M, Miklósi Á, Csányi V (2004) Dogsrespond appropriately to cues of humans’ attentional focus.Behav Process 66:161–172Williams FJ, Mills DS, Guo K (2011) Development of a headmounted,eye-tracking system for dogs. J Neurosci Methods94:259–265Yarbus AL (1967) Eye movements and vision. Plenum Press, New York123