Classification of Domestic Cat (Felis catus) Vocalizations by Naive and Experienced Human Listeners

Abstract

To test for possible functional referentiality in a common domestic cat (Felis catus) vocalization, the authors conducted 2 experiments to examine whether human participants could classify meow sounds recorded from 12 different cats in 5 behavioral contexts. In Experiment 1, participants heard singlecalls, whereas in Experiment 2, bouts of calls were presented. In both cases, classification accuracy was significantly above chance, but modestly so. Accuracy for bouts exceeded that for single calls. Overall, participants performed better in classifying individual calls if they had lived with, interacted with, and had a general affinity for cats. These results provide little evidence of referentiality suggesting instead that meows are nonspecific, somewhat negatively toned stimuli that attract attention from humans. With experience, human listeners can become more proficient at inferring positive-affect states from cat meows.

Full text

Classification of Domestic Cat (Felis catus) Vocalizations by Naive
and Experienced Human Listeners
Nicholas Nicastro and Michael J. Owren
Cornell University
To test for possible functional referentiality in a common domestic cat (Felis catus) vocalization, the
authors conducted 2 experiments to examine whether human participants could classify meow sounds
recorded from 12 different cats in 5 behavioral contexts. In Experiment 1, participants heard single calls,
whereas in Experiment 2, bouts of calls were presented. In both cases, classification accuracy was
significantly above chance, but modestly so. Accuracy for bouts exceeded that for single calls. Overall,
participants performed better in classifying individual calls if they had lived with, interacted with, and
had a general affinity for cats. These results provide little evidence of referentiality, suggesting instead
that meows are nonspecific, somewhat negatively toned stimuli that attract attention from humans. With
experience, human listeners can become more proficient at inferring positive-affect states from cat
meows.
Animal communication has been variously described as a pro-
cess of sharing information (Marler, 1984; Smith, 1997), of ma-
nipulation and skeptical reception (Dawkins & Krebs, 1978; Krebs
& Dawkins, 1984), of “assessment/management” (Owings & Mor-
ton, 1998), and of “affect induction” in receivers (Owren & Ren-
dall, 1997), among other approaches. The function of animal
signals has also been diversely characterized as providing emo-
tional readouts, fostering social prestige (Zahavi, 1975), or coding
information in ways similar to human symbolic language (Cheney
& Seyfarth, 1990; Macedonia & Evans, 1993).
Effective information sharing with, manipulation of, or manage-
ment of humans is critically important to animals that depend on
them. Encoding information would seem particularly relevant to
communication in domestic animals. Although a number of wild
species use calls that reportedly feature some degree of semanticity
(for instance, vervet monkeys, Cheney & Seyfarth, 1990; yellow-
bellied marmots, Blumstein & Armitage, 1997; and ravens,
Bugnyar, Kijne, & Kotrschal, 2001), the acquisition of function-
ally referential signals should be especially beneficial for domes-
ticates, insofar as most cats and dogs live in environments where
human referential signaling is ubiquitous.
There is evidence for unsuspected perceptual–cognitive abilities
in domesticated species. Studies of social gaze in dogs have
suggested canine capabilities of “reading” the posture, head posi-
tion, and eyes of humans that compare favorably to those reported
for chimpanzees (Hare & Tomasello, 1999) and even for 3-year-
old children (Soproni, Miklo´si, Topa´l, & Csa´nyi, 2001). Controlled
experiments involving dog–owner play communication have con-
cluded that dogs can competently interpret certain intentional
movements of humans (McKinley & Sambrook, 2000; Rooney,
Bradshaw, & Robinson, 2001). Feddersen-Petersen (2000) re-
cently reported referential-like qualities in canine barks. There is
also an extensive body of results attesting to the referential qual-
ities of food and antipredator calls of domestic chickens (Evans &
Evans, 1999; Gyger, Marler, & Pickert, 1987).
In their review of the problem of meaning in animal signals,
Macedonia and Evans (1993) posited two criteria for functional
referentiality: first, that calls be strongly correlated with a partic-
ular referent and second, that perceiver responses be correlated
with the calls. Here, we describe two controlled experiments
examining referentiality in domestic cat vocal signaling by testing
whether humans can classify instances of a typical domestic cat-
to-human vocalization in a manner consistent with the second of
these criteria.
Moelk (1944) and Bradshaw and Cameron-Beaumont (2000)
divided the vocal repertoire of the domestic cat into three broad
categories: calls “produced with the mouth closed” (e.g., purrs and
trills), “sounds produced while the mouth is held open in one
position” (e.g., spitting or hissing), and calls produced “while the
mouth is open and gradually closed” (Bradshaw & Cameron-
Beaumont, 2000, p. 71). The last category constitutes the primary
subject of this study: the meow or miaow. This call is characterized
by a mean fundamental frequency (F
0
)of⬃400–1200 Hz, a
modulating pitch profile, and a duration ranging from 110 to 3,100
ms in our sample (see Figure 1).
There are several reasons to pay particular attention to the use
and evolution of this call. First, it is the most common cat-to-
human vocalization (Bradshaw & Cameron-Beaumont, 2000); sec-
ond, it is seldom observed in cat-to-cat interactions (Brown, 1993);
and third, although undomesticated felids meow as juveniles, they
rarely produce the call to humans in adulthood (Cameron-
Beaumont, 1997). On the basis of spectrographic analysis, Shipley,
Carterette, and Buchwald (1991) have described the meow as a
periodic sound with F
0
corresponding to the vibration rate of the
Nicholas Nicastro and Michael J. Owren, Department of Psychology,
Cornell University.
Nicholas Nicastro was supported by National Institute of Mental Health
Predoctoral Research Fellowship T32 MN 19389 during preparation of this
study. We thank cat owners Leslie Horowitz, Thomas Volman, Isabel
Tovar, Jill Mateo, Darryl Mayeaux, S. K. List, Karl and Pamela Reichert,
and Robert and Linda Newton. Thanks also to Penny Bernstein for helpful
comments on the manuscript.
Correspondence concerning this article should be addressed to Nicholas
Nicastro, Department of Psychology, Cornell University, 243 Uris Hall,
Ithaca, New York 14853. E-mail: nn12@cornell.edu
Journal of Comparative Psychology Copyright 2003 by the American Psychological Association, Inc.
2003, Vol. 117, No. 1, 44–52 0735-7036/03/$12.00 DOI: 10.1037/0735-7036.117.1.44
44
This
document
is
copyrighted
by
the
American
Psychological
Association
or
one
of
its
allied
publishers.
This
article
is
intended
solely
for
the
personal
use
of
the
individual
user
and
is
not
to
be
disseminated
broadly.