Poster #114 - Knowledge of Spoken Words by Domestic Dogs: A New Instrument for Use With Dog Owners

• In Event: 2-055 - Poster Session 06
  In Poster Session: PS 06 Section - Language, Communication

Integrative Statement

Dogs are sensitive to human social cues (Miklosi & Topal, 2013) and use such cues to gain information about new situations (Merola et al., 2012) or successfully complete tasks (Kaminski & Nitzschner, 2013). As early as 1928, scientists have sought to assess empirically dogs’ ability to comprehend what we say (Kaminski et al., 2004; Pilley & Reid, 2011; Warden & Warner, 1928). Although these studies investigated exceptional dogs, they demonstrate dogs’ potential to respond consistently and selectively to spoken words. However, no reliable tool exists for determining which and how many words dogs know.
The purpose of the current study was to develop a reliable and valid measure of the number of words to which dogs respond differentially and systematically. Our tool was modeled after the MacArthur-Bates CDI (Fenson et al., 1994), a parent-report instrument that assesses children’s understanding and production of early language. Before children speak, they demonstrate clear receptive language abilities (e.g., directing gaze towards appropriate objects upon hearing select words; Fernald et al., 1998; performing gestures like waving “bye-bye” that correspond with a spoken word; Fenson et al., 1994). Although some may argue that parents are biased, untrained observers, ample research using the CDI with young children suggests otherwise. In developing the CDI, researchers found that accurate and reliable representations of children’s language abilities were obtained when parents used vocabulary checklists rather than generating words. As such, we designed our instrument as a checklist of words separated into categories for use by dog owners.
Our first goal was to assess the content validity and usability of the online checklist with dog owners, while another was to identify predictors of dogs’ word understanding. Participants included 168 owners of dogs of various ages and breeds. Owners answered questions about themselves and their dog, and identified and rated the level at which their dog responded to each of 276 words subdivided into 1 of 8 categories of words. Dog owners reported that their dogs responded selectively to M=89.3 words (SD=35.7, Figure 1) and variables such as owners’ experience with dog training, how quickly dogs reportedly learned new tricks, and breed type showed predictive utility. Principal component analyses revealed that words to which dogs respond encompassed three distinct, yet related components: general words, verbs, and nouns (Table 1).
In sum, we developed a detailed inventory that comprehensively captures dogs’ receptive language abilities, and in doing so, we identified important predictors of word knowledge. The importance of dogs to humans is illustrated by the many roles they serve in humans’ lives, whether as valued family members or as specially trained working dogs. Their ability to interpret human social and verbal cues at a level unmatched by other species (Miklosi & Soproni, 2006) is central to these roles. Our tool not only allows for a standardized measurement of dogs’ receptive abilities, but also may become an important tool for predicting the potential of working dogs and most germane to developmental psychologists, for use in comparative studies of language development with humans and dogs.

Authors

• Catherine Reeve, Dalhousie University
  Presenting Author

• Sophie Jacques, Dalhousie University
  Non-Presenting Author