Cat and household demographics
The majority of owners reported fetching behaviour by current cats (n cats = 853), while a substantial minority reported fetching behaviour by previous cats (n cats = 301). There were significantly more male fetching cats (n = 617) than female fetching cats (n = 537), X2 = 5.50 (1, N = 1,154) p = .018. Most cats were younger than 1 year when their owner first noticed the fetching behaviour (n = 701), followed by cats between the ages of 1 and 7 years (n = 415). Fewer cats first displayed their fetching behaviour at 7 years and above (n = 38). The median age of the cat when fetching was first noticed is 7 months (25th percentile = 5 months, 75th percentile = 16 months; range = 0 – 204 months). Most cats fetched 1 – 5 times in the most recent fetching session (n = 635) and had 1 – 10 fetching sessions per month (n = 677). Figure 1 illustrates the dispersion of data for the number of retrievals in the most recent session, number of monthly fetching sessions and the age of the cat when fetching was first noticed. A vast majority of fetching cats (94%) were untrained (n = 1,089) while a minority were intentionally trained (6%, n = 65). Some cats were identified as mixed breed (n = 493), with 160 cats identified as purebred and 501 of unknown breed. A full list of the frequency of all purebreds in our sample can be found in Supplementary Table S5.
Most fetching cats did not share a household with a dog (n = 989). Out of the cats that did share a household with a dog (n = 162), only 106 lived with at least one fetching dog. In contrast, most fetching cats did share a household with at least one other cat (n = 638). However, for the majority of cats who lived with other cats, the other cat(s) did not fetch (n = 479). A summary of owner and household demographics is presented in Table 1 and a full summary of cat and fetching demographics is presented in Table 2.
Favoured objects
The cats’ favourite categorised object to fetch is presented in Fig. 2 and a spreadsheet of all of the objects cats fetched is in Supplementary Information S6. Cat toys were the most frequently cited favourite object type, followed by spherical objects and cosmetics.
Initiation and termination of fetching sessions
There were unequal distributions of who initiates fetching sessions on average (cat, about equal, or owner) (X2 = 124.29 (2, N = 1,154) p < .001). Post-hoc comparisons demonstrated that cats (n = 553) initiate fetching sessions more than owners (n = 249), (X2 = 115.23 (1), p < .001) and more than when initiation was about equal (n = 352), (X2 = 44.64 (1), p < .001). Similarly, owners were significantly more likely to initiate than both the cat and owner equally (X2 = 17.65 (1), p < .001).
There were also significant differences between who ends the fetching sessions on average (X2 = 329.08 (2), p < .001). Cats (n = 675) terminate fetching sessions more than owners (n = 248), (X2 = 197.54 (1), p < .001) and more than when the termination was considered equal (n = 231), (X2 = 217.59 (1), p < .001). There was not a significant difference in frequency between the owner and both the cat and owner equally terminating the session (X2 = 0.60 (1), p = .437). Thus, cats were depicted by their owners as more frequently the agent that controlled the initiations and terminations of fetching sessions. Figure 3 illustrates the relative proportions of who initiates and who ends the fetching sessions on average.
Table 1. Descriptive statistics of household and owner demographics (N = 1,154 cats). Frequencies and percentages are presented along with the mean, SD (in brackets) and ranges of age and duration.
Respondent gender:
|
Mean respondent age in years (SD):
|
Number of people in household:
|
Mean age of household in years (SD):
|
Female
748 (64.8%)
|
34.69 (11.41)
6 – 73
|
1: 220 (19.1%)
2: 565 (49%)
3: 212 (18.4%)
|
34.80 (9.78)
20 – 73
35.59 (10.91)
0 – 72
36.25 (17.95)
0 – 88
|
Male
266 (23.1%)
|
37.22 (11.13)
4 – 64
|
Non-binary
109 (9.4%)
Prefer not to say
31 (2.7%)
|
29.72 (7.52)
7 – 48
27.09 (7.97)
15 – 40
|
4: 100 (9.5%)
|
31.63 (17.69)
0 – 86
|
5: 39 (3.4%)
6: 6 (0.5%)
|
33.55 (20.61)
1 – 87
24.78 (16.12)
2 – 64
|
|
|
7: 2 (0.2%)
|
27 (18.68)
10 – 73
|
Number of dogs in household:
|
Number of fetching dogs in household:
|
Number of cats in household:
|
Number of other fetching cats in household:
|
0: 989 (85.9%)
|
0: 56 (34.6%)
|
1: 451 (41.4%)
|
0: 479 (75.1%)
|
1: 103 (8.9%)
|
1: 81 (50%)
|
2: 383 (35.2%)
|
1: 134 (21%)
|
2: 47 (4.1%)
|
2: 23 (14.2%)
|
3: 156 (14.3%)
|
2: 22 (3.4%)
|
3+: 12 (1%)
|
3: 2 (1.2%)
|
4: 46 (4.2%)
|
3: 3 (0.5%)
|
Would cat have seen the dog fetch prior to fetching themselves?
|
5: 26 (2.4%)
|
|
Yes: 34 (32.1%)
|
6: 12 (1.1%)
|
|
No: 72 (67.9%)
|
7+: 15 (1.4%)
|
|
|
|
|
|
|
Table 2. Descriptive statistics of cat and fetching demographics (N = 1,154 cats). Frequencies and percentages are presented along with the mean, SD (in brackets) and ranges of age and duration.
Cat sex and neuter status:
|
Cat status:
|
Training status:
|
Breed:
|
Female spayed
501 (43.4%)
|
Previous cat
301 (26.1%)
|
Untrained
1,089 (94.4%)
|
Mixed
493 (42.7%)
|
Female intact
36 (3.1%)
|
Current cat
853 (73.9%)
|
Trained
65 (5.6%)
|
Purebred
160 (13.9%)
|
Male neutered
601 (52.1%)
|
|
|
Unknown
501 (43.4%)
|
Male intact
15 (1.3%)
|
|
|
|
Male unknown
1 (0.1%)
|
|
|
|
Cat age in months at time fetching first noticed (SD):
|
Life stage at time fetching first noticed:
|
Duration of time in months the cat had fetched for (SD):
|
15.90 (23.53)
0 - 204
|
Kitten (< 1 year)
701 (60.7%)
|
51.53 (48.72)
0 - 244
|
|
Young adult (< 7 years)
415 (36%)
|
|
|
|
Mature adult (< 10 years)
22 (1.9%)
|
|
|
|
Senior (≥ 10 years)
16 (1.4%)
|
|
|
Count of who initiates the fetching sessions on average:
|
|
Count of who ends the fetching sessions on average:
|
Cat
|
553 (47.9%)
|
|
Cat
|
675 (58.5%)
|
Owner
|
249 (21.6%)
|
|
Owner
|
248 (21.5%)
|
About equal
|
352 (30.5%)
|
|
About equal
|
231 (20%)
|
|
|
|
|
|
|
Number of retrievals in most recent session
There were significant differences between who usually initiates the fetching sessions and the number of retrievals in the most recent fetching session (Kruskal-Wallis X2 = 13.98 (2, N = 1,154), p < .001). There are significantly fewer retrievals if the owner usually initiates the sessions (median = 5, IQR = 4) compared to the cat (median = 5, IQR = 6, Pairwise Wilcoxon p = .009) or if initiation is about equal (median = 6, IQR = 6, Pairwise Wilcoxon p < .001); there is no difference in retrieval number between when the cat usually initiates and when initiation is about equal (Pairwise Wilcoxon p = .995).
There were significant differences between who usually ends the fetching sessions and the number of retrievals in the most recent session (Kruskal-Wallis X2 = 110.55 (2, N = 1,154), p < .001). There are significantly more retrievals when the owner ends the sessions (median = 8, IQR = 5) in comparison to when the cat ends the sessions (median = 5, IQR = 4, Pairwise Wilcoxon p < .001) or when both the cat and owner equally end the fetch sessions (median = 6, IQR = 6, Pairwise Wilcoxon p < .001). There are significantly more retrievals when both the cat and owner equally end the fetch sessions in comparison to when the cat ends the sessions (Pairwise Wilcoxon p < .001).
Number of monthly fetching sessions
There were significant differences between who usually initiates the fetching sessions and the number of monthly fetching sessions (Kruskal-Wallis X2 = 52.95 (2, N = 1,154), p < .001). There are significantly fewer monthly sessions if the owner usually initiates the sessions (median = 4, IQR = 8) compared to the cat (median = 10, IQR = 16, Pairwise Wilcoxon p < .001) or if initiation is about equal (median = 10, IQR = 16, Pairwise Wilcoxon p < .001) but not between the cat and if initiation is about equal (p = .66).
There are significant differences between who usually ends the fetching sessions and the number of monthly sessions (Kruskal-Wallis X2 = 62.97 (2, N = 1,154), p < .001). There are significantly fewer fetching sessions per month if the cat usually ends the sessions (median = 6, IQR = 12.5) in comparison to the owner (median = 10, IQR = 15, Pairwise Wilcoxon p < .001) or if ending is about equal (median = 12, IQR = 20, Pairwise Wilcoxon p < .001). There were no differences in the number of fetching sessions per month between the owner ending the sessions or whether this was about equal (Pairwise Wilcoxon p = 1.00).
Cat sex
There were no significant differences between male and female cats on who usually initiates the fetching sessions (X2 (2) = 1.82, p = .403) or who usually ends the fetching sessions (X2 (2) = 1.97, p = .374). There were no significant differences between male and female cats on the number of retrievals in the most recent fetching session (Mann-Whitney Z = −1.69, p = .091) or the number of fetching sessions per month (Mann-Whitney Z = − .49, p = .622).
Qualitative Analysis
The analysis generated three themes that are illustrated in Fig. 4.
Fetching process. This theme captures some of the behavioural aspects of the fetching procedure and how and when objects would be retrieved to the owner. Respondents often stated they threw an object (intentionally or unintentionally) after which the cat brought it back to them. Other respondents also stated their cat brought the object to them first after which the owner then threw the object. Cats were noted to have incomplete retrievals, for example in dropping the object halfway between where it landed and where it was first thrown, or in gradually dropping the object further and further away from the owner. This theme also captures how fetching developed from playing and how owners reported they unintentionally threw, dropped or moved objects which also led to the occurrence of fetching. Two respondents noted that they unintentionally moved an object and how their cat reacted:
“In opening a newspaper, the rubber band slipped off and flew down the hallway. Waldo chased it down and brought it back, dropping it at my feet. I 'shot' it again, and again he fetched it.”
“When I was making dinner one evening while she was still a kitten, I dropped some macaroni; she scrambled for a piece and brought it back to me. I [praised] her enthusiastically and a night or two later dropped one to see if she'd do the same. It quickly became her favorite game.”
Acquisition process. This theme captures how the owner first noticed their cat beginning to fetch and retrieve items. Some owners stated they knew their cat was able to fetch prior to adoption (for example, if adopted from a family friend or from a shelter that knew about this behaviour). A small proportion of owners intentionally trained their cat to fetch objects using clicker training or reward-based techniques with food or social rewards. Some owners believed their cats could have learnt to fetch from other animals (including siblings, parents, offspring, unrelated conspecifics, and dogs) or that the cat trained them to play fetch. However, many owners stated their cat opportunistically started fetching or retrieving objects to them. One respondent stated that they knew their cat could fetch prior to adoption, a second respondent wrote how they used clicker training and a third respondent explained how their cat trained them:
“Her former owner first noticed it, but told me that Tiny would bring back balls of paper on her own. The first night I brought her home I tried tossing a ball of paper and she brought it right back!”
“She would always chase the baubles but wouldn't bring them back, she would sometimes pick them up in her mouth and take them somewhere better to bat around before she'd meows at you to throw it again so I started to try catching her when she had the bauble in her mouth and calling her to come. If she came with the bauble she got the click and the treat if not she didn't, just got lots of love for coming when called. Treats were not required as motivation for very long once she understood.”
“Dizzy would bring a toy (a mouse, a bundle of feathers, etc.) to me, drop it at my feet, and then attack the toy -- and my ankles. Not being terribly fond of having my ankles bitten ferociously, I would toss the toy a few feet away and encourage Dizzy to play attack it somewhere else. She'd chase it and almost inevitably bring it back, attack again, and so on. Eventually I realized she was chirping and making eye contact as she brought the toy back and (briefly) sitting patiently after she dropped it, and that her attacking my ankles was her way of asking me to get on with throwing the toy. Once I caught on -- that is to say, once she trained me -- we got into a good routine where I'd throw it, she'd chase it and bring it back and then prepare to pounce while I play-feinted and eventually threw it. Rinse, repeat.”
Fetching under specific circumstances. This theme describes how some cats would fetch only under specific circumstances related to the nature of the fetching session. Some cats would fetch an object only if the object was thrown far enough away from the initial thrown location. Some cats were reported to fetch only a few times in one session, whereas others were reported to fetch obsessively for great lengths of time. Some cats were interested in fetching only one specific object, despite, for example, the owner providing an identical toy of a different colour which the cat would not play or fetch with. Some cats would fetch only in specific locations of the house; two common locations were the bedroom (of which cats were sometimes reported to drop the object in the owner’s bed while the owner was asleep) and throwing the object using the stairs for the cat to chase up and down. Some cats were also reported to fetch only with specific people in the household. One respondent reported that their cat will fetch only with a specific object and a second respondent explained that their cat enjoys playing fetch in two locations:
“She's very particular about what she'll fetch, however. She'll smack anything around, but if you want to play fetch it HAS to be this one type of cosmetic swab. (It's like a Q-tip, but much thinner. Used for eyelash extensions. Tulip is interested in regular Q-tips, but she won't fetch them).”
“She only likes playing in one of two places - the best place is in our bedroom, we throw the ball out of the door down the hall and into the bathroom. There’s a step she can leap over which she likes doing - the second place is from the bottom of the stairs going up into the bathroom or vice versa. If we are trying to tire her out, we both play with her throwing it up and down the stairs”