Vocabulary Learning in a Yorkshire Terrier: Slow Mapping of Spoken Words

Please note: The following is a summary of a research article, referenced at the end of this page, by Ulrike Griebel and D. Kimbrough Oller.


Some animals are known to learn languages. Rico, a border collie, was observed to recognize over 200 words (Griebel, Oller, 2012). Amazingly, Chaser, also a border collie, learned more than 1000 names (Griebel, et al., 2012). In 2004, researchers interpreted Rico’s abilities as an example of fast mapping (Griebel, et al., 2012). In Griebel and Oller’s study, a Yorkshire terrier named Bailey was given similar tests to Rico’s. Griebel and Oller tried to identify if Yorkshire terriers, which were not bred for work and command obedience, could have similar results as Rico’s. Furthermore, the authors questioned if Rico’s tests could prove his ability to fast map (Griebel, et al., 2012).

Bailey is a 12-year-old lap dog who learned the names of more than 120 toys while growing up (Griebel, et al., 2012). The tests were set up so that Bailey’s toys were in a pile located in front of the fireplace, connected to (but visually separated from) the person giving commands in a hallway (Griebel, et al., 2012). First, Bailey accurately fetched toys by name when given a command from her owner (Griebel, et al., 2012). Bailey also performed well on the same test but with unfamiliar voices--“one male, one female, with different accents” (Griebel, et al., 2012). In the Initial Exclusion test, Bailey was further asked to fetch a new toy from a pile of known ones (Griebel, et al., 2012). The novel items were requested along with known items in a random order; the known items served as controls (Griebel, et al., 2012). During this period of testing, Bailey resisted the commands and barked when asked to retrieve a novel item (Griebel, et al., 2012). Although her results were not above chance, she showed signs of “discriminating the new items from the known ones” (Griebel, et al., 2012).

Afterwards, Bailey received various training sessions (during which she learned two novel toys and their names), followed by retention tests (Griebel, et al., 2012). She successfully retrieved the correct novel toys in “exclusion and retention tests,” but failed in the “two-choice task” (Griebel, et al., 2012). In the two-choice task, Bailey failed to accurately associate the novel name with the novel toy when presented with only two novel items (Griebel, et al., 2012). Notably, Bailey showed special attachment to the novel item Triceratops (Griebel, et al., 2012).

Lastly, the owner kept training Bailey until she retrieved the two novel items with enough accuracy to suggest that she knew their names (Griebel, et al., 2012). The results ultimately showe that she did not fast map, and thus calls into question the deduction that Rico could fast map (Griebel, et al., 2012). Instead, he may have had success in certain trials “by a process of extended exclusion” (Griebel, et al., 2012). By extension, this study suggests that many animals may “have evolved [the] cognitive structures required for language such as symbol and category formation” (Griebel, et al., 2012).


Reviewed by Sofia Hsu; 11/26/18

Article Source: Vocabulary Learning in a Yorkshire Terrier: Slow Mapping of Spoken Words


Griebel U, Oller DK (2012) Vocabulary Learning in a Yorkshire Terrier: Slow Mapping of Spoken Words. PLOS ONE 7(2): e30182.https://doi.org/10.1371/journal.pone.0030182