The String-Pulling Paradigm in Comparative Psychology

Ivo F. Jacobs and Mathias Osvath Lund University

String pulling is one of the most widely used paradigms in comparative psychology. First documented 2 millennia ago, it has been a well-established scientific paradigm for a century. More than 160 bird and mammal species have been tested in over 200 studies with countless methodological variations. The paradigm can be used to address a wide variety of issues on animal cognition; for example, what animals understand about contact and connection as well as whether they rely on perceptual feedback, grasp the functionality of strings, generalize across conditions, apply their knowledge flexibly, and possess insight. Mammals are typically tested on a horizontal configuration, birds on a vertical one, making the studies difficult to compare; in particular, pulling a string vertically requires better coordination and attention. A species’ performance on the paradigm is often influenced by its ecology, especially concerning whether limbs are used for foraging. Many other factors can be of importance and should be considered. The string-pulling paradigm is easy to administer, vary, and apply to investigate a wide array of cognitive abilities. Although it can be and has been used to compare species, divergent methods and unclear reporting have limited its comparative utility. With increasing research standards, the paradigm is expected to become an even more fundamental tool in comparative psychology.

Keywords: string pulling, means-end understanding, comparative psychology, animal cognition, insight

One of the most widely used and well-known experimental gnon; still-life pictures of fruit with goldfinches pulling water paradigms in comparative psychology is string-pulling. The basic buckets (see Figure 1). Overall, the practice seems to have had a task—pulling in an out-of-reach reward attached to a string—is wider cultural and historical impact than any other tests of animal simple but can be varied in a vast number of ways to address an intelligence. array of different psychological questions. Perhaps people found it appealing to watch birds pull strings The history of using this practice with animals is far older than because it appears unusually clever. That said, although previously comparative psychology itself. The first documented reference is regarded as an interesting feat (Ray, 1678; Zorn, 1743), in the 19th from the Roman naturalist Pliny the Elder (23–79 AD), who century making captive birds work for their food and water was describes goldfinches pulling up small buckets of water (Bierens heavily criticized as unnatural and cruel and, therefore, not suitable de Haan, 1933; Rackham, 1947). A source of entertainment, the for studies by naturalists (Bierens de Haan, 1933). practice became so common that, since the end of the Middle Given the practice’s long history and widespread use, it is not Ages, the goldfinch has been called putter in Dutch; meaning one surprising that comparative psychologists early on became in- who draws water from a well. Similar names were present in terested in using it as a research paradigm. Since the first German, English, and French in the 19th century (Audubon, 1831; studies a century ago (Hobhouse, 1915; Kinnaman, 1902; Bierens de Haan, 1933; Brückner, 1933). It spread to America Köhler, 1917/1927; Shepherd, 1910, 1915), the paradigm has (Audubon, 1831), and may have originated independently in Ja- been used to test 163 mammal and bird species in 208 studies, pan, using varied tits (Thorpe, 1959). The popularity of the prac- involving at least 50 variations on so-called string patterns (see tice is reflected in two 17th century paintings by Abraham Mi- Figure 2 for an historical overview). The creativity exhibited in varying the routine appears endless, from using virtual strings

This document is copyrighted by the American Psychological Association or one of its allied publishers. (Wasserman, Nagasaka, Castro, & Brzykcy, 2013)toattaching This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. aliverattooneendofthestring,soamonkeycouldthrowit Ivo F. Jacobs and Mathias Osvath, Department of Cognitive Science, Lund University. to fetch his out-of-reach reward (Klüver, 1937). The cognitive We thank The Crafoord Foundation for funding this work. We thank mechanisms investigated range from insight and means-end Helena Osvath for doing the illustrations. We are grateful to Josep Call, understanding to instinct and associative learning. Arno Cimmadon, Erin Colbert-White, Annie Ellison, Nathan Emery, Jolle In the first half of the 20th century, researchers mostly used Jolles, Anastasia Krashenninikova, Héctor Manrique, Felipe Medina, string-pulling as a test for learning speed and for interspecies Tanya Obozova, Stefanie Riemer, and Christian Schloegl for sharing their comparisons, as dictated by ecology. This was followed by a unpublished work. The unpublished research on rooks was done by IFJ at period of increasing interest in brain function, performing le- DEPE, CNRS, Strasbourg under the supervision of Valérie Dufour. We sions, and related interventions on nonhuman animals and in- also thank Liesbeth Sterck and four anonymous reviewers for comments on earlier versions of this work. vestigating the effects on string-pulling. The 1970s saw a shift Correspondence concerning this article should be addressed to Ivo F. in focus toward developmental and sensorimotor aspects of Jacobs, Department of Cognitive Science, Lund University, Helgonavägen cognition under the influence of Piaget. Nowadays, the string- 3, 221 00, Lund, Sweden. E-mail: [email protected] pulling paradigm is mostly used for making phylogenetic com-

89 90 JACOBS AND OSVATH

Figure 1. Details of two paintings by Abraham Mignon (1640–1679). Left: “Still life with fruit and a goldfinch.” Right: “Fruit still-life with squirrel and goldfinch.” Adapted from http://www.the-athenaeum.org. See the online article for the color version of this figure.

parisons and studying the cognitive abilities of nonhuman an- Horizontal and Vertical Orientation of imals in more detail. String-Pulling Setups The aim of this article is to review comparative psychology’s use of the string-pulling paradigm. The problem space is defined As the string-pulling paradigm involves different presentations by an out-of-reach reward or reward container—attached to a of the strings it is beneficial to first consider some of the common various patterns and orientations. Strings are usually oriented string or similar—that the subject can reel in. String-pulling is not either in a horizontal or a vertical fashion. Typically, a horizontal typically considered as tool use because the manipulator is not string can be reeled in with a single pull, whereas a vertical string orienting the “tool,” and the reward is already part of the tool requires better coordination and multiple-step motor planning; (Piaget, 1952; Shumaker, Walkup, & Beck, 2011; St Amant & reach down, grasp and pull, create a loop, stand on it, and repeat Horton, 2008; Van Lawick-Goodall, 1970). Basic string-pulling several times—depending on the length of the string. Some spe- indeed appears to be less taxing than tool use; great apes perform cies might also use other techniques, ranging from turning the better on tasks where strings are used instead of rakes (Herrmann, body while holding the string to consuming the string and then Wobber, & Call, 2008), whereas human children functionally releasing it (Ellison, Watson, & Demers, 2015; Werdenich & discriminate strings earlier than stick tools (Brown, 1990). Huber, 2006). All though require more complex coordination than This document is copyrighted by the American Psychological Association or one of its allied publishers. First, we describe some of the most widely used orientations and a horizontal string, in no small part because the reward will fall if This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. patterns of strings. We then discuss some of the most common the string is released. Furthermore, the subject has a better view of areas of cognition investigated using the string-pulling paradigm. things in the horizontal configuration and can more easily afford We also consider various factors that might influence performance staying focused on the reward rather than attending to details like in such experiments, and we speculate on the ecological, evolu- motor coordination, which might reduce motivation. Finally, it is tionary, and anatomic questions raised. We close with recommen- easier to arrange different horizontal experimental setups. Almost dations for future research. invariably, the easiest arrangement has the reward balancing on a A major part of this review is an extensive catalogue of string- ledge and the attached string hanging down. The subject can obtain pulling studies, including which species have been tested on what the reward with a single tug, requiring only a notion of object variation of the paradigm. In the main text, we focus on studies permanence to represent the hidden reward; but this method has that report their methods and results clearly and that use sound not been used recently. The behaviors required to lift a bucket of statistics based on multiple individuals. Such studies are almost water in the classic goldfinch cage are very similar to reeling in a exclusively recent. For sake of completeness, we still reach studies “wagon” loaded with food, where the wagon will roll back down that do not reach these criteria, where relevant. the inclination unless the string is stepped on (see Figure 3). STRING-PULLING PARADIGM 91

speculate why one orientation seems more difficult for one species than another. Comparing mammals and birds on these tasks is problematic; mammals are typically tested on horizontal problems, birds on vertical ones. In principle, vertical problems should be easier for the mammals that are usually tested than birds; mammals can rest on their hind legs and pull the string “hand over hand,” simply a repetition of reach down and pull, whereas birds usually need to anchor the string with at least one foot. Unless otherwise mentioned, these differences are implied throughout this review.

Common Cognitive Investigations With Different String Patterns As we discuss later in this article (see Ecology, Evolution, and Autonomy), string pulling may be ecologically relevant for some species. Nonetheless, purely innate responses appear insufficient for solving string-pulling tasks—whereas behavioral innovations and forms of learning coupled with physical cognition do appear to be important. Much has been written about the cognitive skills required to pull a string; for example, associative learning, trial and error, causal cognition, means-end understanding, imagination, and insight (Wasserman et al., 2013). We outline some of the most Figure 2. The number of string-pulling studies published per decade. The common areas that the string-pulling paradigm addresses (summa- studies that include multiple species are also accounted for in the graphs of rized in Table 1) and briefly discuss factors that might influence or the separate taxa (e.g., a study with multiple bird species is represented in confound performance (see Table 2). The most common tests use the birds and multiple species lines). No string-pulling studies exists that a single string (see Table 3), but variations can easily be created test both mammals and birds. The number of studies before 1909 is (see Table 4). Indeed, to reveal what strategies animals use in underrepresented; most of those, however, describe observations or anec- solving string-pulling problems, it is important that they be tested dotes on goldfinches or other songbirds in the string-pulling cages (see on multiple setups. Bierens de Haan, 1933). The graph reveals that mammals have been more tested than birds, and that there is a trend toward more studies on multiple bird species. In the 1930s many string-pulling studies tested multiple Means-End Understanding species, which contrast with the general picture of comparative psychology In presenting his theory of infant sensorimotor development, at the time (Shettleworth, 2009b). Piaget (1952) argues that string pulling tests means-end understanding. It is the typical test for Stage 4 (8–12 months), where the The differences between horizontal and vertical configurations are empirically well established. The ability to solve a vertical string-pulling problem comes about 2 months later in both human and gorilla infants (Redshaw, 1978); and a few days later in yellow-crowned parakeets (Funk, 2002). Many animals require more time to solve vertical problems, if they can solve them at all (Adams, 1929; Bagotskaya, Smirnova, & Zorina, 2012; Heinrich, 1995; Hobhouse, 1915; Jolly, 1964a, 1964b; McDougall & Mc- This document is copyrighted by the American Psychological Association or one of its allied publishers. Dougall, 1927; Obozova & Zorina, 2013). This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. However, there are some exceptions. For example, although crossbills were apt vertical string pullers they ignored horizontal strings (Obozova, Bagotskaya, Smirnova, & Zorina, 2014). Wild chipmunks, unable to solve horizontally posed problems, manage to solve vertical ones—perhaps because they occasionally must pull shoots that are suspending food in the air but are unaccus- tomed to food being out of reach horizontally (Gordon, 1938). Heubel (1940) reports a palm civet who performs better at a vertically oriented task, probably for the same reason. Surpris- ingly, experience with horizontal string-pulling did not improve Figure 3. Example of a typical cage, used for centuries, where songbirds vertical string-pulling in orange-winged amazons (Krashenin- were required to pull up a water bucket to drink, and a wagon with food to nikova & Schneider, 2014). Few studies test the same animals on eat. For old video footage, see Britsh Pathé, 1932. (Reprinted from Bierens both horizontally and vertically oriented problems, leaving one to de Haan, 1933.) 92 JACOBS AND OSVATH

Table 1 Summary of How Different Cognitive Skills Could Be Tested in the String-Pulling Paradigm

Cognitive skill Shown by (solving)

Means-end understanding Goal-directedness (solving perpendicular condition, Figure 4) No proximity error (solving slanted, crossed, or contact conditions, Fig. 4) Flexible solutions (see Functional Generalization and Fixedness) No dependence on immediate feedback (see Perceptual Feedback) Understanding contact Contact/no-contact condition Understanding connection Differentiating types of contact (e.g., that a connection is formed through support or physical continuity of material) Not moving away with the reward while the string is still attached Independence of visual feedback No visual feedback (restricted visual access, or solving perpendicular coiled condition, Figure 5) Negative visual feedback (conditions where the reward moves away or sideways before coming closer) Perceiving visual continuity No dependence on visually different strings Functional generalization Generalizing to novel string types Using different techniques No functional fixedness (not pulling a reward that is too heavy; not pulling when the food can be obtained from elsewhere) Insight Crossed condition (Figure 4) with no previous experience after an initial impasse; not a result of chance, trial-and-error, visual feedback or innate processes Note. String-pulling might not be particularly suitable for investigating insight (see Insight).

first signs of “truly intelligent” behavior are manifested through 1987; Parker & McKinney, 1999; Pepperberg, 2002; Vauclair, intentional coordination of two independent schemata; final (end) 2012), one reason why string-pulling tests are so common in and transitional (means). Piaget’s framework has often been ap- animal cognition (Chevalier-Skolnikoff, 1982, 1983; Dumas & plied to nonhuman animals (e.g., Antinucci, 1989; Doré & Dumas, Doré, 1991; Frank & Frank, 1985; Hallock & Worobey, 1984;

Table 2 Several Other Factors That May Affect String-Pulling Performance

Factor Explanation and examples

Age Older animals often perform better because they are more cognitively developed and less playful (Davis, Lovelace, & McKenna, 1964; Mason & Harlow, 1961), although juveniles might be more successful because they can be more persistent (Vince, 1958, 1961). Attention Animals with poor attention are more likely to fail (Warden, Koch, & Fjeld, 1940) or pull unbaited strings without monitoring feedback (Beck, 1967). Divided attention could explain the difficulty of more complex conditions (Heinrich & Bugnyar, 2005; Nissani, 2004). Captivity and rearing Although little studied, the effects of rearing and testing environment seem to be minimal, and can usually be explained by the influence of associated other factors (Funk, 2002; Heinrich, 1995; Huber & Gajdon, 2006; Singh, 1966). Inhibition Animals might pull strings at random if they have poor inhibition, which is especially a problem when tested on patterned problems after the single string condition (Seibt & Wickler, 2006). Limb use and laterality Animals often fail if they lack the dexterity to grasp a string or cannot step on it for anchoring (Beck, 1967; Newton, 1967). Limb lateralization seems to be beneficial (Magat & Brown, 2009). Motivation A lack of motivation might result in animals pulling strings randomly, so testing if they prefer to obtain easier reward can be helpful (Pfuhl, 2012). Using preferred rewards or moderate food deprivation can increase performance (Birch, 1945; Crutchfield, 1939).

This document is copyrighted by the American Psychological Association or one of its alliedNeophobia publishers. Neophobic animals can take a long time to even approach a string, so providing them with loose strings to

This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. interact with before testing can be beneficial (Heinrich, 2000). Neophilia and play Neophilic and playful animals often pull a string for its self-rewarding value (Mason, Harlow, & Reuping, 1959; Schuck-Paim, Borsari, & Ottoni, 2009). Giving them extended experience with strings prior to testing can help. Object permanence Hiding the reward in a container is problematic for animals without object permanence (Heubel, 1939). Even if they have it they likely perform better if they see the reward directly (Fischel, 1936). Personality Personality traits such as boldness, curiosity, dominance, and sociality can influence string-pulling performance (Jolles, Ostojic, & Clayton, 2013; Pfuhl, Gattemayr, & Bugnyar, 2014). Side biases Because a typical patterned string problem is composed of only two strings, a side bias is a successful strategy half of the time. Adding strings or only baiting the nonpreferred side are common solutions (Gagne, Levesque, Nutile, & Locurto, 2012; Hobhouse, 1915). String type and length Functional generalization tests have shown that the type of string used is often not very important, as long as it is visible and can easily be grasped. For better comparability between species, relative string length should be the same, such as twice the body length for birds (Krasheninnikova, 2013). Visual acuity Better string-pulling performance might be partially attributable to better visual acuity (Harris & Meyer, 1971b). STRING-PULLING PARADIGM 93

Table 3 Overview of Species, and Their performance, on the Single String Condition

First Number success in Species N Orientation of trials Success trial Reference

Apes (Hominoidae) Black crested gibbon (Hylobates concolor) 2c H 1 Yes 1 Guillaume & Meyerson, 1931 Bonobo (Pan paniscus) 5c H 12 Yes Buttelmann, Carpenter, Call, & Tomasello, 2008 Bornean orangutan (Pongo pygmaeus) 7c H 12 Yes Buttelmann et al., 2008 3c H Yes Chevalier-Skolnikoff, 1982 4c H Yes Chevalier-Skolnikoff, 1983 3c H 1 Yes 1 Guillaume & Meyerson, 1931 Chimpanzee (Pan troglodytes) 6c H 30 Yes Birch, 1945 16c H 12 Yes Buttelmann et al., 2008 1c H Yes Chevalier-Skolnikoff, 1982 2c D 1 Yes 1 Drescher & Trendelenburg, 1927 30c H Ն2,000 Yes Finch, 1941a 15c H 1 Yes 1 Guillaume & Meyerson, 1931 2c H Yes Hallock & Worobey, 1984 1c H Yes Hayes, 1951 1c V Yes Hayes, 1951 1c H Yes Hobhouse, 1915 4c H 1 Yes 1 Köhler, 1927 7c V 1 Yes 1 Köhler, 1927 2c H 4–9 Yes Յ3 Mathieu et al., 1980 2c H Ն2 1S 1U Poti & Spinozzi, 1994 7c H Ն8 Yes 1 (6) Povinelli, 2012 Gorilla (Gorilla sp.) 1c D 7 Yes 1 Yerkes, 1927 Orangutan (Pongo sp.) 1c H 32 Yes 11–20 Laidler, 1978 1c H Yes Miles, 1990 1c V Yes Miles, 1990 Western gorilla (Gorilla gorilla) 5c H 12 Yes Buttelmann et al., 2008 1c H Yes Chevalier-Skolnikoff, 1982 1c H 1 Yes 1 Guillaume & Meyerson, 1931 4c H Yes Redshaw, 1978 4c V Yes Redshaw, 1978 3c H Yes 1 (1) Riesen, Greenberg, Granston, & Fantz, 1953 1c H Yes Spinozzi & Natale, 1989 White-handed gibbon (Hylobates lar) 4c H 2–6 Yes 1 Beck, 1967 Old world monkeys (Cercopithecoidae) Barbary macaque (Macaca sylvanus) 1c H 1 Yes 1 Guillaume & Meyerson, 1931 Black crested mangabey (Lophocebus aterrimus) 2c H 1 Yes 1 Guillaume & Meyerson, 1931 Campbell’s monkey (Cercopithecus campbelli) 1c H 1 Yes 1 Guillaume & Meyerson, 1931 De Brazza’s monkey (Cercopithecus neglectus) 1c H 1 Yes 1 Guillaume & Meyerson, 1931 Golden-bellied mangabey (Cercocebus chrysogaster) 1c H 1 Yes 1 Guillaume & Meyerson, 1931 This document is copyrighted by the American Psychological Association or one of its allied publishers. Green monkey (Chlorocebus This article is intended solely for the personal use of the individual user and is notsabaeus to be disseminated broadly. ) 1c H 1 Yes 1 Guillaume & Meyerson, 1931 Grivet (Chlorocebus aethiops) 1c H 1 Yes 1 Guillaume & Meyerson, 1931 Guinea baboon (Papio papio) 1c H 1 Yes 1 Guillaume & Meyerson, 1931 Japanese macaque (Macaca fuscata) 1c H Yes Antinucci, Spinozzi, Visalberghi, & Volterra, 1982 1c H 1 Yes 1 Guillaume & Meyerson, 1931 1c H Yes Potí, 1989 Long-tailed macaque (Macaca fascicularis) 1c D 2 Yes 1 Drescher & Trendelenburg, 1927 1c H 1 Yes 1 Drescher & Trendelenburg, 1927 1c H Yes Fischel, 1930 2c H 2 Yes Klüver, 1961 1c H Yes Klüver & Bucy, 1939 1c H Yes Potí, 1989 (table continues) 94 JACOBS AND OSVATH

Table 3 (continued)

First Number success in Species N Orientation of trials Success trial Reference

Mandrill (Mandrillus sphinx) 1c H 1 Yes Guillaume & Meyerson, 1931 Mona monkey (Cercopithecus mona) 1c D 3 Yes 1 Bierens de Haan, 1930 1c H 1 Yes 1 Bierens de Haan, 1930 Northern plains gray langur (Semnopithecus entellus) 2c H Yes Chevalier-Skolnikoff, 1982 Olive baboon (Papio anubis) 62w V 3S 10F 1 (1) Laidre, 2008 Patas monkey (Erythrocebus patas) 6c H 16 Yes 1 Hall & Mayer, 1966 Pig-tailed macaque (Macaca 2c D 2 Yes 1 Bierens de Haan, 1930 nemestrina) 2c H 1 Yes 1 Bierens de Haan, 1930 Putty-nosed monkey (Cercopithecus nictitans) 1c H 1 Yes 1 Guillaume & Meyerson, 1931 Red-tailed monkey (Cercopithecus ascanius) 1c H 1 Yes 1 Guillaume & Meyerson, 1931 Rhesus macaque (Macaca mulatta) 15c H Yes Christensen & Pribram, 1977 16c H Yes Davis & McDowell, 1953 15c H Yes Davis et al., 1964 1ca D 2 Yes 1 Drescher & Trendelenburg, 1927 1ca H 1 Yes 1 Drescher & Trendelenburg, 1927 1c H 1 Yes 1 Hobhouse, 1915 1c H 30 Yes 1 Kinnaman, 1902 1c H Yes Klüver & Bucy, 1938 14c H Yes Klüver & Bucy, 1939 81c H Ն300 Yes Mason & Harlow, 1961 1c D 2 Yes 1 Nellmann & Trendelenburg, 1926 1c H Ն2 Yes 1 Nellmann & Trendelenburg, 1926 9c H 10 Yes 1 (9) Shepherd, 1910 12c H Yes Singh, 1966 17c H Yes Ungerleider & Brody, 1977 12c H Yes Ungerleider & Pribram, 1977 5c H Յ20 Yes Warden et al., 1940 2c H Yes Warden, Barrera, & Walt, 1942 Yellow baboon (Papio cynocephalus) 1c D 2 Yes 1 Drescher & Trendelenburg, 1927 1c H 1 Yes 1 Drescher & Trendelenburg, 1927 1c H Ն2 Yes Ն2 Nellmann & Trendelenburg, 1926 White-eyelid mangabey (Cercocebus 2c D Ն2 Yes 1 Bierens de Haan, 1930 sp.) 1c H Ն2 Yes Ն2 Bierens de Haan, 1930 New world monkeys (Platyrrhini) Capuchin monkey (Cebus sp.) 1c H Yes Klüver, 1937 Common marmoset (Callithrix jacchus) 2c H 3–20 Yes 1 (1) Gagne et al., 2012 Common squirrel monkey (Saimiri 10c H 20 Yes Cha & King, 1969 sciureus) 3c H Ն5 Yes Fife & Kamback, 1970 5c H Ն5 Yes Harris & Meyer, 1971a 6c H Ն24 Yes Su, 1982 Cotton-top tamarin (Saguinus 8c H 12 Yes Chapman & Weiss, 2013 oedipus) 14c V 6S 8F 2 (1) Dillis, Humle, & Snowdon, 2010 19c V 8S 11F Humle & Snowdon, 2008 Geoffroy’s spider monkey (Ateles geoffroyi) 2c H Յ20 Yes Warden et al., 1940

This document is copyrighted by the American Psychological Association or one of its allied publishers. Red-bellied tamarin (Saguinus

This article is intended solely for the personal use of the individual user and is notlabiatus to be disseminated broadly. ) 18c V Ն2 Yes Prescott & Buchanan-Smith, 1999 Red-faced spider monkey (Ateles paniscus) 2c H 1 Yes 1 Guillaume & Meyerson, 1931 Saddle-backed tamarin (Saguinus fuscicollis) 18c V Ն2 Yes Prescott & Buchanan-Smith, 1999 Tufted capuchin monkey (Cebus 3c H Yes Spinozzi, 1989 apella) 1c H Յ20 Yes Warden et al., 1940 White-headed capuchin (Cebus 1c D Ն2 Yes 1 Bierens de Haan, 1930 capucinus) 1c H 1 Yes 1 Bierens de Haan, 1930 1c H Yes Klüver & Bucy, 1939 2c H Յ20 Yes Warden et al., 1940 1c H Yes Warden et al., 1942 Prosimians (Prosimii) Black lemur (Eulemur macaco) 1c H Yes Fischel, 1930 Brown lemur (Eulemur fulvus) 4c H Ն2 2S 2U 1 (1) Jolly 1964a, 1964b STRING-PULLING PARADIGM 95

Table 3 (continued)

First Number success in Species N Orientation of trials Success trial Reference

4c V Ն2 3S 1U Jolly 1964a, 1964b Greater galago (Otolemur 1c H Ն2 Jolly 1964a, 1964b crassicaudatus) 1c V Ն2 Yes Jolly 1964a, 1964b Mongoose lemur (Eulemur mongoz) 2c D Ն2 Yes 1–3 Bierens de Haan, 1930 2c H 1S 1F 1 (1) Bierens de Haan, 1930 Potto (Perodicticus potto) 2c H Ն2 1S 1U Jolly 1964a, 1964b 2c V Ն2 1S 1U Jolly 1964a, 1964b Red-fronted lemur (Eulemur rufifrons) 1c H Yes Fischel, 1930 Red slender loris (Loris tardigradus) 1c H Ն2 Jolly 1964a, 1964b 1c V Ն2 Jolly 1964a, 1964b Ring-tailed lemur (Lemur catta) 1c D Ն2 Yes 1 Bierens de Haan, 1930 1c H Ն6 Yes 3 Bierens de Haan, 1930 1c H Ն2 Yes Jolly 1964a, 1964b 1c V Ն2 Yes Jolly 1964a, 1964b 1c H 25 Yes 1 Klüver, 1961 Senegal bushbaby (Galago senegalensis) 4c H Ն2 Jolly 1964a, 1964b 4c V Ն2 Yes Jolly 1964a, 1964b Other mammals Asian elephant (Elephas maximus) 2c H Ն100 Yes 1 Nissani, 2004 12c H Յ8 Yes Plotnik, Lair, Suphachoksahakun, & de Waal, 2011 Black rat (Rattus rattus) Ն18w V Yes Ewer, 1971 Brown rat (Rattus norvegicus) 243c H Ն50 Yes Crutchfield, 1939 11c V 240 Yes Յ10 Hamilton & Ellis, 1933 7c V Ն20 Yes Kolb, Cioe, & Comeau, 2008 82c H Ն10 Yes 1 McCulloch, 1934 120c H Ն10 Yes McCulloch & Pratt, 1934 3c H Ն2 Yes 1 McDougall & McDougall, 1927 5c V Ն6 Yes 1 (2) McDougall & McDougall, 1927 135c H Ն4 Yes Pratt, 1938 Ն1c H Yes Tolman, 1937 20c V Ն84 Yes Tomie & Whishaw, 1990 33c V Ն14 Yes Whishaw, Tomie, & Kolb, 1992 53c V Ն14 Yes Whishaw & Tomie, 1991 10c V Ն14 Yes Whishaw & Tomie, 1995 Chipmunk (Tamias sp.) 2w D Yes Gordon, 1938 2w H No Gordon, 1938 2w V Yes Gordon, 1938 Common treeshrew (Tupaia glis) 3c H Ն2 2S 1U Jolly 1964a, 1964b 3c V Ն2 Jolly 1964a, 1964b Dingo (Canis lupus dingo) 13c H Յ6 12S 1F 1 (10) Smith & Litchfield, 2013 Domestic cat (Felis silvestris catus) 10c H 5S 5F Adams, 1929 4c V 1S 3F Adams, 1929 5c H Yes Dumas & Doré, 1991 Ն2c D No Drescher & Trendelenburg, 1927 Ն2c H No Drescher & Trendelenburg, 1927 9c H Յ30 Yes Herbert & Harsh, 1944

This document is copyrighted by the American Psychological Association or one of its allied publishers. 3c D Յ34 Yes 1 (1) Hobhouse, 1915

This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 14c H Ն2 6S 6F 1 (2) Trueblood & Smith, 1934 15c H Յ20 Yes Whitt, Douglas, Osthaus, & Hocking, 2009 Domestic dog (Canis lupus 1c D 1 Yes 1 Bierens de Haan, 1932 familiaris) 1c H 1 Yes 1 Bierens de Haan, 1932 4c D Yes de Jong, 1919 3c D No Drescher & Trendelenburg, 1927 3c H No Drescher & Trendelenburg, 1927 3c D 1 No Fischel, 1933 2c H Ն2 Yes 1 Fischel, 1933 4c D Ն2 Yes Hobhouse, 1915 1c V 11 Yes Hobhouse, 1915 1c V No Köhler, 1927 9c H 6 Yes Miklosi et al., 2003 1c D Ն2 No Nellmann & Trendelenburg, 1926 (table continues) 96 JACOBS AND OSVATH

Table 3 (continued)

First Number success in Species N Orientation of trials Success trial Reference

1c H Ն2 Yes 1 Nellmann & Trendelenburg, 1926 16c H 20 Yes Osthaus, Lea, & Slater, 2005 7c H 1 2S 5F 1 Sarris, 1937 3c D 3 2S 1F 1 Sarris, 1937 3c H No Shepherd, 1915 Elephant (Elephantidae sp.) 1c V Ն15 Yes 1 Hobhouse, 1915 Fruit bat (Pteropus sp.) Ն2c V Yes Atkinson, 1994 Grey wolf (Canis lupus lupus) 1c H 9 Yes 1 Grzimek, 1942 9c H 6 Yes Miklosi et al., 2003 Kangaroo (Macropus sp.) 1c D Yes Heubel, 1939 1c H No Heubel, 1939 Palm civet (Paradoxurinae sp.) 1c D 2 Yes 1 Heubel, 1940 1c H Ն3 No Heubel, 1940 1c V 2 Yes 1 Heubel, 1940 Quokka (Setonix brachyurus) 2c H Yes Bonney & Wynne, 2002 Raccoon (Procyon lotor) 2c D Ն5 Yes 1–5 Bierens de Haan, 1932 2c H Yes 1 Bierens de Haan, 1932 1c H Ն5 Yes 1 McDougall & McDougall, 1931 1c V Ն3 Yes 1 McDougall & McDougall, 1931 Red fox (Vulpes vulpes) Ն4c H Ն15 Yes 1 (Ն1) Schmid, 1936 Red squirrel (Sciurus vulgaris) 1c D Ն3 No Bierens de Haan, 1932 1c H Ն8 Yes 6 Bierens de Haan, 1932 Rock squirrel (Spermophilus variegatus) 4c H 3–4 Yes King & Witt, 1966 Tayra (Eira barbara) 1c H Yes Schmid, 1936 White-nosed coati (Nasua narica) 2c D Ն5 Yes 3 Bierens de Haan, 1932 2c H Yes 1 Bierens de Haan, 1932 Corvids (Corvidae) American crow (Corvus Ն1w V Yes Burns, 1895 brachyrhynchos) 2c V No Heinrich, 1995 Carrion crow (Corvus corone 1c V Ն9 Yes 2 Hertz, 1926 corone) 5w V No Schiestl, 2013 Carrion/hooded crow hybrid (Corvus corone corone x cornix) 68w V No Schiestl, 2013 Common raven (Corvus corax) 5c V 4 3S 2F 2–3 Bagotskaya, Smirnova, & Zorina, 2010 36c V Ն2 17S 19F Heinrich, 1995 Ն50w V No Heinrich, 1995 6c V 1 5S 1F 1 Heinrich, 2000 6c V Ն5 5S 1F Heinrich & Bugnyar, 2005 Ն1w V Yes Holmberg, 1957 Ն2w V Yes Larsson, 1958 9c V Yes Pfuhl et al., 2014 Eurasian jay (Garrulus glandarius) 1c V Ն10 No Fischel, 1936 1c V 5 Yes Thorpe, 1963 Hooded crow (Corvus corone cornix) 8c V 4 4S 4F 1 Bagotskaya et al., 2010 Ն7w V Yes Holmberg, 1957 Ն2w V Yes Larsson, 1958 42w V No Schiestl, 2013

This document is copyrighted by the American Psychological Association or one of its allied publishers. Jackdaw (Corvus monedula) 14c V 8 5S 9F 1 (1) Cimadom, 2013

This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 1c H Yes 1 Dücker & Rensch, 1977 1c V Ն8 Yes 2 Hertz, 1926 1c V Yes Thorpe, 1963 Ն1c V Yes Zorn, 1743 New Caledonian crow (Corvus Auersperg, von Bayern, Gajdon, moneduloides) 5c H Ն8 Yes 1 Huber, & Kacelnik, 2011 5c V 8 3S 2F 1 (1) Cimadom, 2013 12c V 10 10S 2F 1 (7) Medina, 2012 7c V Ն10 Yes Taylor, Elliffe, Hunt, & Gray, 2010 4c V 10 Yes 1 (3) Taylor, Medina, et al. 2010 Rook (Corvus frugilegus) 1w V Yes Cross, 1947 1w V Yes Inglet, 1967 19c V Ն35 Yes Jolles et al., 2013 1w V Yes Richards, 1973 Ն2w V Yes Thorpe, 1944 STRING-PULLING PARADIGM 97

Table 3 (continued)

First Number success in Species N Orientation of trials Success trial Reference

Ն1w V Yes Washington, 1974 Parrots (Psittaciformes) African grey parrot (Psittacus 1c V Yes Krasheninnikova, 2014 erithacus) 1c V Yes Krasheninnikova, Bovet, Busse, & Péron, 2012 4c V 2S 2F 1 (2) Pepperberg, 2004 3c V Yes C. Schloegl, personal communication Australian king parrot (Alisterus scapularis) 5c V 10 Yes 1 (1) Magat & Brown, 2009 Blue-and-gold macaw (Ara ararauna) 2c V Yes 1 Fischel, 1936 Blue-fronted amazon (Amazona aestiva) 2c V 7 Yes 1 (1) Schuck-Paim et al., 2009 Blue-headed macaw (Ara couloni) 7c V No Krasheninnikova et al., 2012 Blue-throated macaw (Ara glaucogularis) 2c V Yes Krasheninnikova et al., 2012 Budgerigar (Melopsittacus undulatus) 3c H Yes 5 (1) Dücker & Rensch, 1977 2c V No Fischel, 1936 5c V Yes Krasheninnikova, 2014 13c V 5S 8F Krasheninnikova et al., 2012 5c V 10 No Magat & Brown, 2009 Cockatiel (Nymphicus hollandicus) 10c V 10 Yes 1 (9) Krasheninnikova, 2013 5c V 10 No Magat & Brown, 2009 Cuban amazon (Amazona leucocephala) 1c V Yes C. Schloegl, personal communication Eclectus parrot (Eclectus roratus) 10c V Yes Krasheninnikova, 2014 1c V Yes C. Schloegl, personal communication Galah (Eulophus roseicapilla) 16c V 10 Yes 1 (5) Krasheninnikova, 2013 5c V 10 Yes Magat & Brown, 2009 Gang-gang cockatoo (Callocephalon fimbriatum) 5c V 10 Yes 1 (2) Magat & Brown, 2009 Golden conure (Guaruba guarouba) 8c V No Krasheninnikova et al., 2012 2c V 7 No Schuck-Paim et al., 2009 Greater Patagonian conure 3c V Yes Krasheninnikova, 2014 (Cyanoliseus patagonus) 7c V 3S 4F Krasheninnikova et al., 2012 Greater Vasa parrot (Coracopsis vasa) 10c V Yes Krasheninnikova, 2014 Green-winged macaw (Ara 4c V 25 2S 2F Krasheninnikova, Bräger, & Wanker, chloroptera) 2013 2c V Yes C. Schloegl, personal communication Hyacinth macaw (Anodorhynchus hyacinthius) 4c V 7 Yes 1 (2) Schuck-Paim et al., 2009 Illiger’s macaw (Primolius maracana) 2c V Yes Krasheninnikova, 2014 2c V Yes Krasheninnikova et al., 2012 Kea (Nestor notabilis) 6c H Ն8 Yes 1 Auersperg et al., 2011 7c V 10 6S 1F 1 (6) Werdenich & Huber, 2006 Lear’s macaw (Anodorhynchus leari) 4c V 7 2S 2F 1 (1) Schuck-Paim et al., 2009 Monk parakeet (Myiopsitta

This document is copyrighted by the American Psychological Association or one of its allied publishers. monachus) 7c V Yes Krasheninnikova, 2014

This article is intended solely for the personal use of the individual userOrange-winged and is not to be disseminated broadly. Amazon (Amazona 1c V Yes 1 Fischel, 1936 amazonica) 45c V 10 22S 23F Krasheninnikova & Schneider, 2014 Peach-faced lovebird (Agapornis roseicollis) 15c V Yes Krasheninnikova, 2014 12c V 6S 6F Krasheninnikova et al., 2012 Rainbow lorikeet (Trichoglossus haematodus) 10c V 25 8S 2F Krasheninnikova et al., 2013 Red-lored Amazon (Amazona autumnalis) 1c V Yes C. Schloegl, personal communication Red-spectacled Amazon (Amazona pretrei) 2c V Yes Krasheninnikova, 2014 7c V 2S 5F Krasheninnikova et al., 2012 Red-tailed black cockatoo (Calyptorhynchus banksii) 5c V 10 Yes 1 (1) Magat & Brown, 2009 (table continues) 98 JACOBS AND OSVATH

Table 3 (continued)

First Number success in Species N Orientation of trials Success trial Reference

Senegal parrot (Poicephalus senegalus) 12c V Yes Krasheninnikova, 2014 1c V Yes C. Schloegl, personal communication Scarlet macaw (Ara macao) 1c V Yes 1 Capener, 2011 Slender-billed cockatoo (Cacatua tenuirostris) 6c V Yes Krasheninnikova, 2014 6c V Yes Krasheninnikova et al., 2012 Spectacled parrotlet (Forpus conspicillatus) 13c V 9S 4F 1 (8) Krasheninnikova & Wanker, 2010 8c V 25 6S 2F Krasheninnikova et al., 2013 Sulphur-crested cockatoo (Cacatua galerita) 3c V 25 Yes Krasheninnikova et al., 2013 5c V 10 Yes 1 (1) Magat & Brown, 2009 Sun parakeet (Aratinga solstitialis) 2c V Yes Krasheninnikova et al., 2012 Superb parrot (Polytelis swainsonii) 5c V 10 Yes Magat & Brown, 2009 Thick-billed parrot (Rhynchopsitta 3c V Yes Krasheninnikova, 2014 pachyrhyncha) 7c V 1S 6F Krasheninnikova et al., 2012 Yellow-crowned parakeet (Cyanoramphus auriceps) 11c V Yes 1 (Ն5) Funk, 2002 11c H Yes 1 (Ն4) Funk, 2002 Non-passerine non-psittacine birds Glaucous-winged gulls (Larus Ն2c H No T.A. Obozova, personal glaucescens) communication Great grey owl (Strix nebulosa) 12c H 6S 6F Obozova & Zorina, 2013 Rock pigeon (Columba livia) Ն1c V No Obozova & Zorina, 2013 4c H Yes Schmidt & Cook, 2006 Turkey vulture (Cathartes aura) 6c V 15 3S 3F 1 (2) Ellison et al., 2015 Other passerine birds American goldfinch (Carduelis tristis) 1c V Yes Audubon, 1831 Blackbird (Turdus merula) 40c VU Yes Sasvari, 1985 Blue tit (Parus caeruleus) 9c V Yes Altevogt, 1953 2w V Yes 1 (1) Brooks-King, 1941 Ն2w V Yes Brooks-King & Hurrell, 1958 2c V 1S 1F 1 (1) Herter, 1940 7c V 4 5S 2F 1 (5) Obozova et al., 2014 Ն1w V Yes Thorpe, 1959 Ն1w V Yes Thorpe, 1963 Brewer’s blackbird (Euphagus cyanocephalus) Ն1c V No Millikan & Bowman, 1967 Bullfinch (Pyrrhula pyrrhula) Ն1c V Yes Murray, 1882 Ն1c V No Thorpe, 1944 Chickadee (Poecile sp.) Ն1w V Yes Heinrich, 1999 Cocos finch (Pinaroloxias inornata) Ն1c V No Millikan & Bowman, 1967 Common crossbill (Loxia Ն1c V Yes Newton, 1967 curvirostra) 12c H No T.A. Obozova, personal communication 12c V 4 4S 8F 1 (2) Obozova et al., 2014 Common myna (Acridotheres tristis) 1c H Ն6 Yes 6 Dücker & Rensch, 1977

This document is copyrighted by the American Psychological Association or one of its allied publishers. Common wren (Troglodytes

This article is intended solely for the personal use of the individual user and is nottroglodytes to be disseminated broadly. ) 2c V No Thorpe, 1963 Chaffinch (Fringilla coelebs) 2c V No Bierens de Haan, 1933 Ն1w V No Thorpe, 1944 Ն1c V Yes Thorpe, 1963 16c V Ͻ12 No Vince, 1958 Cuban grassquit (Tiaris canorus) Ն1c V No Millikan & Bowman, 1967 Domestic canary (Serinus canaria) 1c V Yes Bierens de Haan, 1933 1c V No Murray, 1882 11c V Յ12 5S 6F Vince, 1958 Eurasian nuthatch (Sitta europaea) 1c V Ն3 No Herter, 1940 Ն1 V Yes Thorpe, 1963 European robin (Erithacus rubecula) Ն2w V No Thorpe, 1944 3c V No Thorpe, 1963 Garden warbler (Sylvia borin) 1c D 6 No Teyrovský, 1930 1c H Yes Teyrovský, 1930 STRING-PULLING PARADIGM 99

Table 3 (continued)

First Number success in Species N Orientation of trials Success trial Reference

Goldfinch (Carduelis carduelis) 3c V 2S 1F Bierens de Haan, 1933 Ն3c V 1SՆ1F Brückner, 1933 1c V Yes “Hector the ‘mule’,” 1950 Ն1c V Yes Hertz, 1937 1c V Yes Murray, 1882 Ն1c V Yes Pliny the Elder, 79 AD (Rackham, 1947) Ն1c V Yes Ray, 1678 52c V 10 12S 40F 1 (3) Seibt & Wickler, 2006 Ն1c V Yes Zorn, 1743 Goldfinch/canary hybrid (Carduelis x 1c V Yes “Hector the ‘mule’,” 1950 Serinus) 1c V Yes Murray, 1882 Great tit (Parus major) Ն2w V Yes Bierens de Haan, 1933 2w V Yes Brooks-King, 1941 1c V Yes Clementius, 1933 365c V 1 91S 274F 1 Cole et al., 2011 Ն1w V Yes Erhardt, 1933 1c V Ն3 No Herter, 1940 Ն1w V Yes Hertz, 1926 1w V Yes Rijk, 1934 1w V Yes Thorpe, 1943 Ն1w V Yes Thorpe, 1959 28c V Յ5 4S 24F Thorpe, 1963 Ն1w V Yes Thorpe, 1963 12c V Յ4 1S 11F 1 (1) Vince, 1956 6c V Yes Vince, 1964 Greenfinch (Carduelis chloris) 2c V No Bierens de Haan, 1933 1c V No Bierens de Haan, 1933 1w V Yes Thorpe, 1944 Ն1c V Yes Thorpe, 1963 18c V Յ12 3S 15F Vince, 1958 House sparrow (Passer domesticus) Ն1w V No Hertz, 1937 1w V Yes Thorpe, 1944 Large cactus finch (Geospiza conirostris) Ն1c V Yes Millikan & Bowman, 1967 Large tree finch (Camarhynchus psittacula) Ն1c V Yes Millikan & Bowman, 1967 Linnet (Carduelis cannabina) 2c V No Bierens de Haan, 1933 1c V Yes Bierens de Haan, 1933 4c V 2S 2F Newton, 1967 Ն1c V Yes Ray, 1678 Loggerhead shrike (Lanius ludovicianus) Ն1c V Yes Millikan & Bowman, 1967 Medium tree finch (Camarhynchus pauper) Ն1c V Yes Millikan & Bowman, 1967 Northern mockingbird (Mimus polyglottos) Ն1c V No Millikan & Bowman, 1967 Oak titmouse (Baeolophus inornatus) Ն1c V Yes Millikan & Bowman, 1967 Redpoll (Carduelis flammea) Ն1c V Yes Crozier, 1910

This document is copyrighted by the American Psychological Association or one of its allied publishers. Ն1c V No Thorpe, 1944

This article is intended solely for the personal use of the individual userRed-winged and is not to be disseminated broadly. blackbird (Agelaius phoeniceus) Ն1c V No Millikan & Bowman, 1967 Siskin (Carduelis spinus) 2c V No Bierens de Haan, 1933 1c V Yes Bierens de Haan, 1933 1c V No Fischel, 1936 Ն1c V Yes Hertz, 1937 Ն1c V Yes Murray, 1882 29c V 10 18S 11F 1 (10) Seibt & Wickler, 2006 Ն2c V Ն1S Ն1F Thorpe, 1944 Small tree finch (Camarhynchus parvulus) Ն1c V Yes Millikan & Bowman, 1967 Song thrush (Turdus philomelos) 40c VU Yes Sasvari, 1985 Tufted titmouse (Baeolophus bicolor) 1w V Yes Dickinson, 1969 Varied tit (Parus varius) Ն1c V Yes Thorpe, 1959 (table continues) 100 JACOBS AND OSVATH

Table 3 (continued)

First Number success in Species N Orientation of trials Success trial Reference

Vegeterian finch (Platyspiza crassirostris) Ն1c V No Millikan & Bowman, 1967 White-crowned sparrow (Zonotrichia leucophrys) Ն1c V No Millikan & Bowman, 1967 Woodpecker finch (Camarhynchus pallidus) Ն1c V Yes Millikan & Bowman, 1967 Note. This condition does not include non-straight strings that test for perceptual feedback. A total of 68 mammal and 85 bird species were tested in 163 studies on the single string condition. N ϭ maximum number of individuals tested (c: captive, w: wild). Italics indicate studies where it is unclear whether this species was tested. Orientation ϭ down (D), horizontal (H), vertical (V), vertical unattached (VU) (the string is not attached and placed in a glass cylinder to be pulled out, which is, therefore, more similar to H). Success ϭ whether the subjects were successful in the number of trials displayed in the previous column (Yes ϭ all subjects were successful; No ϭ no subjects were successful; S ϭ number of subjects that were successful; F ϭ number of subjects that failed; U ϭ number of subjects for which the results are unknown). First success in trial ϭ the first trial in which at least one subject was successful, with the number in parentheses showing the number of subjects (all succeeded if no number in parentheses is shown). Missing information is left blank. a Although this species was not tested in this reference, they were previously tested by the same authors and extensively reported here, which allows for inclusion in the table.

Mathieu, Daudelin, Dagenais, & Décarie, 1980; Miles, 1990; Potì Laidler, 1978; Mason & Harlow, 1961; Mason et al., 1959; Obo- & Spinozzi, 1994; Redshaw, 1978; Spinozzi & Natale, 1989). zova et al., 2014; Riesen et al., 1953; Schuck-Paim et al., 2009; Means-end behavior is said to involve the deliberate, planned Settlage, 1939; Taylor, Medina, et al., 2010; Warden et al., 1940). execution of a sequence of steps to achieve a goal; it is revealed in Reliance on proximity is ineffective in another situation; two situations where an obstacle must be removed to reach the goal rewards are present, but only one can be obtained. Under the (Huber & Gajdon, 2006; Piaget, 1952). In the case of string- so-called contact/no contact conditions (see Figure 4), the reward pulling, the obstacle is the distance to an out-of-reach reward. is placed close to—but not touching—the “incorrect” string and However, pulling a string does not always require means-end attached to a broken string. The subject needs to choose the piece understanding, for a number of reasons. An animal may reach for of string that contacts the reward (see Understanding Contact, the reward directly and accidently touch the string, causing the below). The two final common patterns—the double-crossed and reward to move via the string and thereby impresses an association pseudocrossed conditions (see Figure 4)—might seem more com- between the two (Thorpe, 1963). Each time the string is pulled, the plex because they are nonlinear (Birch, 1945; Harlow & Settlage, reward moves closer, an action that might be repeated purely from 1934), but they can still be solved using a proximity bias. A that association (Dumas & Doré, 1991; Taylor, Knaebe, & Gray, potential problem with all patterned string problems, but especially 2012). This is one reason why it is informative to test animals on the more complex ones, is that subjects could develop a side bias conditions where perceptual feedback is limited (see Perceptual (see Table 2). Feedback, below) or where the reward is too heavy to pull (see An animal can reasonably be said to have means-end under- Functional Generalization and Fixedness, below). standing when the string pulling is goal-directed (solving the A sometimes overlooked problem is that the pulling action or perpendicular setup), not dependent on proximity (solving the the string itself may be rewarding. This appears true for several slanted, crossed, or contact/no contact conditions); used flexibly species (e.g., Altevogt, 1954; Beck, 1967; Riesen et al., 1953; (see Functional Generalization and Fixedness, below), and not Schuck-Paim et al., 2009; Whitt et al., 2009). Sometimes an animal dependent on perceptual feedback (see Perceptual Feedback, be- pulls an unbaited string at an equal rate to a baited one, implying low). that the string pulling is rewarding in itself. In this light, the perpendicular configuration in Figure 4 tests goal-directedness, This document is copyrighted by the American Psychological Association or one of its allied publishers. revealed when animals repeatedly choose the baited string even

This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Learning and Experience though scrambling toward the reward can still result in accidental success (Mason & Harlow, 1961). Contrary to claims that the behaviors necessary for string- Presented with strings in different patterns, dogs almost invari- pulling are innate and not cognitively taxing (see Ecology, Evo- ably paw at the string closest to the food (Osthaus et al., 2005; lution, and Anatomy, below), learning and experience make many Range, Möslinger, & Virányi, 2012; Riemer, Müller, Range, & important contributions; for example, goldfinches’ and siskins’ Huber, 2014). Slanting of the correct string toward the inside, so experiences handling branches facilitate their string-pulling per- that its proximal end is farther from the reward than the incorrect formance (Seibt & Wickler, 2006) whereas, by contrast, the most string (slanted condition), or positioning it in the crossed condition successful juvenile greenfinches and canaries were reared without (see Figure 4), always leads to failure. This so-called proximity access to grasses or other string-like materials (Thorpe, 1963; error is very widespread. Pulling the string closest to the reward Vince, 1958). Meanwhile, in experiments testing for understanding might be the most frequent strategy animals use when faced with of contact, chimpanzees and orangutans performed equally well patterned string-pulling problems (Bagotskaya et al., 2012; Cha & regardless of whether or not they had experience with such exper- King, 1969; Gagne et al., 2012; Klüver, 1961; Köhler, 1927; imental materials as ropes and cloths (Herrmann et al., 2008). STRING-PULLING PARADIGM 101

Table 4 Overview of the Species Tested on Various Patterned String Problems (See Figure 4)

Species N Or Pe Co Sl Ps Cr Do Ot Reference

Apes (Hominoidae) Black crested gibbon (Hylobates concolor) 2c HxGuillaume & Meyerson, 1931 Bonobo (Pan paniscus) 5c HxAmici, Barney, Johnson, Call, & Aureli, 2012 4c HxHerrmann et al., 2008 34c H x Herrmann, Hare, Call, & Tomasello, 2010 6c Hx xMayer et al., 2014 5c Vx Schrauf & Call, 2011 12c H x Wobber & Hare, 2011 Bornean orangutan (Pongo pygmaeus) 5c HxAmici et al., 2012 3c HxGuillaume & Meyerson, 1931 32c H x Herrmann, Call, Hernández-Lloreda, Hare, Tomasello, 2007 37c HxHerrmann et al., 2008 6c Vx Schrauf & Call, 2011 Chimpanzee (Pan troglodytes) 18c HxAmici et al., 2012 6c HxxxxBirch, 1945 8c Hxx xxxxFinch, 1941b 4c H x Gonzalez, Gentry, Bitterman, 1954 15c H x Guillaume & Meyerson, 1931 1c H x x Hayes, 1951 106c H x Herrmann et al., 2007 124c HxHerrmann et al., 2008 7c Hx xxKöhler, 1927 12c Hx xMayer et al., 2014 7c HxPovinelli, 2000 7c Hx Povinelli, 2012 9c Vx Schrauf & Call, 2011 12c H x Wobber & Hare, 2011 Gorilla (Gorilla sp.) 1c D,H x Yerkes, 1927 Orangutan (Pongo sp.) 4c Hxx xxxxFischer & Kitchener, 1965 1c Hx x Laidler, 1978 Western gorilla (Gorilla gorilla) 5c HxAmici et al., 2012 3c Hxx xxxxFischer & Kitchener, 1965 1c H x x Guillaume & Meyerson, 1931 5c HxHerrmann et al., 2008 3c Hxx xxxxRiesen et al., 1953 1c Vx Schrauf & Call, 2011 White-handed gibbon (Hylobates lar) 4c HxBeck, 1967 Old world monkeys (Cercopithecoidae) Baboon (Papio sp.) 1c Hxx xxxHarlow & Settlage, 1948 Barbary macaque (Macaca sylvanus) 1c HxGuillaume & Meyerson, 1931 Black crested mangabey (Lophocebus aterrimus) 2c HxGuillaume & Meyerson, 1931 Campbell’s monkey (Cercopithecus campbelli) 1c HxGuillaume & Meyerson, 1931 De Brazza’s monkey (Cercopithecus neglectus) 1c HxGuillaume & Meyerson, 1931 Drill (Mandrillus leucophaeus) 1c H x x x x x x Balasch et al., 1974 Formosan rock macaque (Macaca cyclopsis) 3ca Hxx xxx Yagi, 1964

This document is copyrighted by the American Psychological Association or one of its allied publishers. Golden-bellied mangabey (Cercocebus chrysogaster) 1c HxGuillaume & Meyerson, 1931 This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Green monkey (Chlorocebus sabaeus) 1c HxGuillaume & Meyerson, 1931 1c Hxx xxxHarlow & Settlage, 1948 Grivet (Chlorocebus aethiops) 1c HxGuillaume & Meyerson, 1931 Guinea baboon (Papio papio) 1c HxGuillaume & Meyerson, 1931 Japanese macaque (Macaca fuscata) 6c Hxx xxx Yagi, 1964 1c HxGuillaume & Meyerson, 1931 Long-tailed macaque (Macaca fascicularis) 12c HxAmici et al., 2012 6c Hxx xxxHarlow & Settlage, 1934 7c Hxx xx xKlüver, 1961 1c Hx xKlüver & Bucy, 1939 13c H x Schmitt, Pankau, & Fischer, 2012 Mandrill (Mandrillus sphinx) 3c Hxx xxxxBalasch et al., 1974 1c HxGuillaume & Meyerson, 1931 3c Hxx xxxHarlow & Settlage, 1934 (table continues) 102 JACOBS AND OSVATH

Table 4 (continued)

Species N Or Pe Co Sl Ps Cr Do Ot Reference

Mona monkey (Cercopithecus mona) 1c Hxx xxxHarlow & Settlage, 1934 Olive baboon (Papio anubis) 1c Hx x xBolwig, 1962 5c H x Schmitt et al., 2012 Pig-tailed macaque (Macaca nemestrina) 6c Hxx xxxHarlow & Settlage, 1934 1c Hxx xxxSettlage, 1939 Putty-nosed monkey (Cercopithecus nictitans) 1c Hxx xxxxBalasch et al., 1974 1c HxGuillaume & Meyerson, 1931 Red-tailed monkey (Cercopithecus ascanius) 1c HxGuillaume & Meyerson, 1931 Rhesus macaque (Macaca mulatta) 17c H x x Akert, Orth, Harlow, & Schiltz, 1960 10c H x Callaghan, McQueen, Scott, & Bigelow, 1954 15c Hxx xxx Christensen & Pribram, 1977 16c H x x x x Davis, McDowell, Deter, & Steele, 1956 15c H x x x x Davis, McDowell, Grodsky, & Steele, 1958 16c HxxxxDavis & McDowell, 1953 16c HxxxxxDavis & Steele, 1963 15c Hx Davis et al., 1964 8c Hxx xxxHarlow & Settlage, 1934 31c H x x x x x Harlow & Settlage, 1948 4c Hxx xxxHarlow, 1939 2c HxHarlow et al., 1955 8c H x x x Klüver, 1937 1c H x Klüver & Bucy, 1938 14c H x Klüver & Bucy, 1939 19c H x x Kruper, Patton, & Koskoff, 1971 6c Hxx xx Mason et al., 1956 81c Hx xx xMason & Harlow, 1961 11c Hx xx Raisler & Harlow, 1965 7c Hxx xxxRiopelle, Alper, Strong, & Ades, 1953 9c Hxx xxxSettlage, 1939 12c Hxx xxx Singh, 1966 17c Hxx xxx Ungerleider & Brody, 1977 12c Hxx xxx Ungerleider & Pribram, 1977 5c Hxxx Warden et al., 1940 10c Hxx xxx Warren, Leary, Harlow, & French, 1957 4c Hxx x xWeiner & Harlow, 1952 9c Hxx xxxxWilson & Mishkin, 1959 Sooty mangabey (Cercocebus atys) 3c Hxx xxxHarlow & Settlage, 1934 1c Hxx xxxHarlow & Settlage, 1948 Yellow baboon (Papio cynocephalus) 1c Hxx xxxHarlow & Settlage, 1934 New world monkeys (Platyrrhini) Capuchin monkey (Cebus sp.) 1c D, H x Klüver, 1937 2c Hx Klüver, 1961 Common marmoset (Callithrix jacchus) 2c Hxxx x xGagne et al., 2012 13w Hx Halsey, Bezerra, & Souto, 2006 Common squirrel monkey (Saimiri sciureus) 10c HxxxxxxxCha & King, 1969 3c HxxxxFife & Kamback, 1970 This document is copyrighted by the American Psychological Association or one of its allied publishers. 5c Hxx xxxHarris & Meyer, 1971a This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 2c Hx Klüver, 1961 6c Hx x Su, 1982 Geoffroy’s spider monkey (Ateles geoffroyi) 12c HxAmici et al., 2012 1c Hxx xxxHarlow & Settlage, 1934 2c Hx Klüver, 1961 5c Hxx xxx Lashley, 1948 2c Hxxx Warden et al., 1940 Red-faced spider monkey (Ateles paniscus) 2c HxGuillaume & Meyerson, 1931 Tufted capuchin monkey (Cebus apella) 12c HxAmici et al., 2012 9c Hx xMayer et al., 2014 6c HxSu, 1982 1c Hxxx Warden et al., 1940 White-headed capuchin (Cebus capucinus) 2c Hxx xxxHarlow & Settlage, 1934 2c Hx Klüver, 1961 1c Hx xKlüver & Bucy, 1939 STRING-PULLING PARADIGM 103

Table 4 (continued)

Species N Or Pe Co Sl Ps Cr Do Ot Reference

2c Hxxx Warden et al., 1940 Prosimians (Prosimii) Ring-tailed lemur (Lemur catta) 1c Hx Klüver, 1961 Other mammals Asian elephant (Elephas maximus) 2c Hx xNissani, 2004 Brown rat (Rattus norvegicus)Hx Adams, 1933 82c Hx McCulloch, 1934 120c Hx McCulloch & Pratt, 1934 135c Hx Pratt, 1938 20c Vx Tomie & Whishaw, 1990 33c Vx Whishaw et al., 1992 53c Vx Whishaw & Tomie, 1991 10c Vx Whishaw & Tomie, 1995 Domestic cat (Felis silvestris catus) 12c Hxx xAdams, 1929 14c Hxx Trueblood & Smith, 1934 15c Hx x Whitt et al., 2009 Domestic dog (Canis lupus familiaris) 3c Hx Fischel, 1933 4c HxFrank & Frank, 1985 4c Dx Hobhouse, 1915 64c Hx x x Osthaus et al., 2005 10c HxxxxRange et al., 2012 34c Hx x xRiemer et al., 2014 Elephant (Elephantidae sp.) 1c DU x Hobhouse, 1915 Grey wolf (Canis lupus lupus) 4c HxFrank & Frank, 1985 1c Hx Grzimek, 1942 9c HxxxxRange et al., 2012 Quokka (Setonix brachyurus) 2c Hx Bonney & Wynne, 2002 Raccoon (Procyon lotor) 3c Hx Johnson & Michels, 1958 1c Hx McDougall & McDougall, 1931 3c Hxx xxx Michels et al., 1961 Red fox (Vulpes vulpes) Ն4c H x Schmid, 1936 Rock squirrel (Spermophilus variegatus) 4c Hxxxxx xKing & Witt, 1966 Tayra (Eira barbara) 1c H x Schmid, 1936 Corvids (Corvidae) Common raven (Corvus corax) 4c Hx x Bagotskaya et al., 2012 1c V x Capener, 2011 36c, > 50w H, V x x x Heinrich, 1995 10c DU x Heinrich, 2000 12c DU x Heinrich & Bugnyar, 2005 7c Vx Pfuhl, 2012 Green jay (Cyanocorax yncas) 5c V x x x H. Manrique, personal communication Hooded crow (Corvus corone cornix) 10c Hxxx x xBagotskaya et al., 2012 Jackdaw (Corvus monedula) 14c Vx Cimadom, 2013 1c Hx Dücker & Rensch, 1977 New Caledonian crow (Corvus moneduloides) 12c Vx x x xTaylor, Medina, et al., 2010 11c HxTaylor et al., 2012 Rook (Corvus frugilegus) 12c V x x x x I.F. Jacobs, unpublished data Parrots (Psittaciformes) African grey parrot (Psittacus erithacus) 1c Vx x xKrasheninnikova, 2014 1c Vx x xKrasheninnikova et al., 2012 This document is copyrighted by the American Psychological Association or one of its allied publishers. 4c Vx Pepperberg, 2004 This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 3c V x C. Schloegl, personal communication Blue-and-gold macaw (Ara ararauna) 2c Vx Fischel, 1936 Blue-fronted amazon (Amazona aestiva) 2c Vx x xSchuck-Paim et al., 2009 Blue-headed macaw (Ara couloni) 7c Vx x xKrasheninnikova et al., 2012 Blue-throated macaw (Ara glaucogularis) 2c Vx x xKrasheninnikova et al., 2012 Budgerigar (Melopsittacus undulatus) 3c Hxx xxxxDücker & Rensch, 1977 5c Vx x xKrasheninnikova, 2014 13c Vx x xKrasheninnikova et al., 2012 Cockatiel (Nymphicus hollandicus) 10c Vx x xKrasheninnikova, 2013 Cuban amazon (Amazona leucocephala) 1c V x C. Schloegl, personal communication Eclectus parrot (Eclectus roratus) 10c Vx x xKrasheninnikova, 2014 1c V x C. Schloegl, personal communication Galah (Eulophus roseicapilla) 16c Vx x xKrasheninnikova, 2013 Golden conure (Guaruba guarouba) 8c Vx x xKrasheninnikova et al., 2012 2c Vx Schuck-Paim et al., 2009 (table continues) 104 JACOBS AND OSVATH

Table 4 (continued)

Species N Or Pe Co Sl Ps Cr Do Ot Reference

Greater Patagonian conure (Cyanoliseus patagonus) 3c Vx x xKrasheninnikova, 2014 7c Vx x xKrasheninnikova et al., 2012 Greater Vasa parrot (Coracopsis vasa) 10c Vx x xKrasheninnikova, 2014 Green-winged macaw (Ara chloroptera) 4c Vx x xKrasheninnikova et al., 2013 2c V x C. Schloegl, personal communication Hyacinth macaw (Anodorhynchus hyacinthius) 4c Vx x xSchuck-Paim et al., 2009 Illiger’s macaw (Ara maracana) 2c Vx x xKrasheninnikova, 2014 2c Vx x xKrasheninnikova et al., 2012 Kea (Nestor notabilis) 7c Vx xx Werdenich & Huber, 2006 Lear’s macaw (Anodorhynchus leari) 4c Vx x xSchuck-Paim et al., 2009 Monk parakeet (Myiopsitta monachus) 7c Vx x xKrasheninnikova, 2014 Orange-winged Amazon (Amazona amazonica) 45c Vx x xKrasheninnikova & Schneider, 2014 Peach-faced lovebird (Agapornis roseicollis) 15c Vx x xKrasheninnikova, 2014 12c Vx x xKrasheninnikova et al., 2012 Rainbow lorikeet (Trichoglossus haematodus) 10c Vx x xKrasheninnikova et al., 2013 Red-lored Amazon (Amazona autumnalis) 1c V x C. Schloegl, personal communication Red-spectacled Amazon (Amazona pretrei) 2c Vx x xKrasheninnikova, 2014 7c Vx x xKrasheninnikova et al., 2012 Senegal parrot (Poicephalus senegalus) 12c Vx x xKrasheninnikova, 2014 1c V x C. Schloegl, personal communication Slender-billed cockatoo (Cacatua tenuirostris) 6c Vx x xKrasheninnikova, 2014 6c Vx x xKrasheninnikova et al., 2012 Spectacled parrotlet (Forpus conspicillatus) 22c Vx x xKrasheninnikova et al., 2013 Sulphur-crested cockatoo (Cacatua galerita) 3c Vx x xKrasheninnikova et al., 2013 Sun parakeet (Aratinga solstitialis) 2c Vx x xKrasheninnikova et al., 2012 Thick-billed parrot (Rhynchopsitta 3c Vx x xKrasheninnikova, 2014 pachyrhyncha) 7c Vx x xKrasheninnikova et al., 2012 Non-passerine non-psittacine birds Great grey owl (Strix nebulosa) 12c HxObozova & Zorina, 2013 Harris hawk (Parabuteo unicinctus) 1c Vx Colbert-White, McCord, Sharpe, & Fragaszy, 2013 Rock pigeon (Columba livia) 4c Hx xSchmidt & Cook, 2006 Other passerine birds Blue tit (Parus caeruleus) 7c Vx x x xObozova et al., 2014 Common crossbill (Loxia curvirostra) 12c Vx x x xObozova et al., 2014 Common myna (Acridotheres tristis) 1c Hx xx Dücker & Rensch, 1977 Great tit (Parus major) 9c H x Kawamori & Matsushima, 2012 Marsh tit (Poecile palustris) 8c H x Kawamori & Matsushima, 2012 Varied tit (Parus varius) 9c H x Kawamori & Matsushima, 2012 Note. A total of 49 mammal and 46 bird species were tested on patterned string conditions in 112 studies. N ϭ maximum number of individuals tested (c: captive, w: wild). Bold indicates studies that were at least partially dedicated to the string-pulling problem with presented results, thereby excluding anecdotes, personal communications, or very brief or unclear results. Italics indicate studies where it is unclear whether this species was tested. Or ϭ orientation down (D), down-up (DU), horizontal (H), vertical (V). Pe ϭ perpendicular; Co ϭ converging; Sl ϭ slanted; Ps ϭ pseudo-crossed; Cr ϭ crossed; Do ϭ double crossed; Ot ϭ other (including conditions with only one string as long as it was not straight as in the regular single string condition). a Although this species was not tested in this reference, they were previously tested by the same author and extensively reported here, which allows for inclusion in the table.

Improvement in string pulling over a number of sessions shows The order in which an animal is tested under different con- This document is copyrighted by the American Psychological Association or one of its allied publishers. the positive effects of learning and experience, with a correspond- ditions might help indicate how it solves problems. Mason and This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. ing decrease in exploratory behavior, solution time, dropping Harlow (1961) explored extensively the role of experience in errors, and incorrect choices (e.g., Balasch, Sabater-Pi, & Padrosa, string pulling. Rhesus macaques first tested on the crossed 1974; Beck, 1967; Davis & McDowell, 1953; Dücker & Rensch, condition had more difficulty solving it than those first tested 1977; Ellison et al., 2015; Finch, 1941b; Fischel, 1930; Fischer & on the perpendicular condition. When conditions were shifted, Kitchener, 1965; Gagne et al., 2012; Harlow, Schlitz, & Settlage, juveniles with experience of the perpendicular condition per- 1955; Heinrich, 1995; Krasheninnikova et al., 2013; Mason & formed more poorly than naïve subjects, but adults performed Harlow, 1961; Mason, Blazek, & Harlow, 1956; Michels, Pustek, better. Under the pseudocrossed condition, juveniles previously & Johnson, 1961; Nissani, 2004; Range et al., 2012; Riemer et al., trained on the crossed condition made more errors than those 2014; Riesen et al., 1953; Schuck-Paim et al., 2009; Taylor, trained on the perpendicular condition, suggesting negative Medina, et al., 2010). Previous experience does not always have a transfer mediated by proximity error. Adolescent monkeys clear directional effect; perhaps surprisingly, cats became worse on found the pseudocrossed condition more difficult than the per- the perpendicular condition with increased experience (Whitt et pendicular one—arguably because they followed the strings al., 2009). visually to some extent and then lost track. Their experience did STRING-PULLING PARADIGM 105

Figure 4. The most commonly used string patterns. Here, the correct strings are always on the left but normally their locations are randomized. The species tested on the perpendicular, converging, slanted, crossed, pseudocrossed, and double crossed patterns are listed in Table 4.

not have any strong effect on subsequent perpendicular perfor- Silva, 2005). To advance understanding of the role of learning mance but did adversely affect performance under the crossed- and experience in animal problem-solving, focus should shift strings condition—again implying that the monkeys made the from whether animals use previous experience to what kind of proximity error. experience they use (Call, 2013). Some animals solve each problem separately through associative trial-and-error learning when given many repetitions of the same condition—but not when the conditions are intermixed, Understanding Contact under which conditions they fail to perform above chance. Squirrel Subjects have to pay attention to the contact between reward and monkeys who have learned to solve the crossed and pseudocrossed string to be successful in the contact/no contact condition, where patterns on repetitive trials require an average of twice as many two rewards are present but only one is in contact with and trials to solve them in an intermixed series (Cha & King, 1969; but connected to a string (see Figure 4). In contrast to the perpendic- see Harris & Meyer, 1971a for contrasting results). The same ular setup, animal subjects will perform at chance level if they effect can be found in budgerigars and rock squirrels. Possibly all only pull the string closest to the reward. For example, crossbills that these subjects learned is spatial discrimination conditional on and blue tits often hang down from their perch and pull the string the location of the reward (Dücker & Rensch, 1977; King & Witt, 1966). Raccoons perform differently; their success on two inter- closest to the food; their performance is, therefore, significantly mixed series is best explained by their following the strings visu- better on the perpendicular and slanted than the contact/no contact This document is copyrighted by the American Psychological Association or one of its allied publishers. ally (Michels et al., 1961). condition, whereas hooded crows score about equally well under This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Alongstandinganddelicateissueinpsychologyishow both conditions—possibly because they pay attention to the strings experience influences problem-solving abilities. A common (Obozova et al., 2014). Generally, correct choices are made more argument is that the problem-solving capacities of animals are often in the perpendicular condition (Krasheninnikova, 2013; only revealed if they have no (or very little) relevant experience Krasheninnikova et al., 2012, 2013; Mason & Harlow, 1961; of the problem components; but that might be too stringent, Schuck-Paim et al., 2009). This indicates that although these especially when comparing their abilities to those of humans, species are goal-directed toward a reward they do not pay close who in most cases have relevant experience. The burden of attention to the configuration of the strings or do not understand proof seems to depend on the species in question. For example, that the proximal end of a string has to be in contact with the it has been argued that animals do not understand the physical reward for it to be obtainable. Note that, even if the same subjects principles involved in the trap-tube when it is made nonfunc- are tested on both conditions, they are either often not statistically tional by turning it upside down (e.g., Povinelli, 2000). Later compared, or there is a lack of counterbalance in the method investigations showed that humans—who possess such under- (Bagotskaya et al., 2012; Klüver, 1961; Obozova & Zorina, 2013; standing—also avoid the nonfunctional trap (Silva, Page, & Povinelli, 2000; Range et al., 2012; Schmidt & Cook, 2006). 106 JACOBS AND OSVATH

Understanding of contact is more frequently addressed using the Researchers have mostly investigated whether or not animals so-called support paradigm, which likewise tests means-end un- understand this difference using the support paradigm or the derstanding but, instead of using a string, uses a reward placed on broken-tool paradigm. Povinelli’s (2000) chimpanzees had diffi- a surface such as a cloth. It is similar in many ways to the culty differentiating between a cloth supporting an apple and one string-pulling paradigm. The main difference is that the connection wrapped around an apple, which suggests that they did not differ- is established through gravity instead of a knot. The unobtainable entiate connection from contact. reward is placed next to or over a cloth or on a broken cloth, to The broken-tool paradigm uses two baited tools, one of which is name the typical conditions. broken. The status of each tool is demonstrated before being Performance using the support and string-pulling paradigms is aligned; they are then partially covered before the subject is often similar (e.g., Amici et al., 2012; Herrmann et al., 2007, 2008; allowed to choose, so that they appear identical. Orangutans and Schmitt et al., 2012). In great apes, a solution to the two tasks chimpanzees have difficulty choosing the correct tool, possibly develops around the same age (Chevalier-Skolnikoff, 1983; Ma- because of poor attention or memory (Mulcahy & Schubiger, thieu et al., 1980; Poti & Spinozzi, 1994; Redshaw, 1978). 2014; Seed, Seddon, Greene, & Call, 2012). Seemingly successful transfer between conditions on the string- Some evidence suggests that apes perform better when broken pulling task could be explained as the result of a change in the strings or cloths are used instead of tools—possibly because the unrewarded string only, and not of any change in the rewarded tools first need to be moved into contact with the food, which string; that is, under those conditions that test for understanding of requires extra attention (Herrmann et al., 2008). The broken-tool contact, one succeeds by always pulling the string that touches the paradigm does not involve tools in the usual sense, because the reward. Changes are made to the unrewarded string—increasing reward is attached to a stick; in this way, the paradigm resembles the width of the gap, placing food over the string, and so forth— the string-pulling and support paradigms more than it might ap- but the rewarded string never changes. In principle, such transfers pear. could be explained by associative learning and, if so, do not serve That animals might not understand the string’s connection with their purpose. The problem is not apparent in the typical patterned the object is supported by several studies in which the subjects string conditions with only one reward, because both strings are tried to run or fly away with the attached reward (e.g., Adams, changed between conditions (see Figure 4). The solution is to 1929; Cross, 1947; Heinrich, 1995; Heubel, 1940; Laidre, 2008; apply similar changes to the rewarded string under the contact/no Vince, 1956). The ravens in Heinrich’s (1995, 1999, 2000) studies contact condition. only flew off with the attached food if they did not themselves pull the food up; it took at least six trials to learn to drop it. Those that Understanding Connection did pull it up themselves never flew off with the string still attached, not even in a thousand trials. In comparison, a cat, a dog, Animals with means-end understanding are not necessarily able and an elephant pulled the loose end of a string tied around a post to distinguish between contact and connection. They understand and not the part behind it which was connected to the reward, that the string is a means to reach a goal, but they might not suggesting a lack of understanding of connection (a chimpanzee understand the mechanism of connectedness. The difficulty in and a rhesus macaque did, however, seem to understand the understanding connection may be based on that is has to be problem) (Hobhouse, 1915). An obvious explanation for the con- inferred in contrast to the directly observable contact state. Pov- fusion is that these animals do not understand how knots function inelli (2000) showed that chimpanzees preferred to pull ropes tied (Povinelli, 2000; Shepherd, 1910; Yerkes, 1927). One would ex- to a banana over ropes that did not touch the banana. However, pect them to untie the knots if they did. There are, however, they did not differentiate between a rope resting on top of the observations of animals biting through an attached string (e.g., banana and one supporting the banana. Therefore, Povinelli con- Krasheninnikova et al., 2012; Yerkes, 1927). Several older studies cluded that chimpanzees do not understand the nature of contact included tests where animal first had to remove a hook from a cage (e.g., connectedness), and that they base their choices solely on bar before they could pull the string. Even if this is a simpler task degree of current or potential contact. Similar results have been than untying or destroying a knot, most subjects initially failed found for other primates, albeit with strong individual differences (e.g., Adams, 1929; Guillaume & Meyerson, 1931; Hobhouse, in poorly described studies (Drescher & Trendelenburg, 1927; 1915; Yerkes, 1927). This document is copyrighted by the American Psychological Association or one of its allied publishers. Fischel, 1930; Guillaume & Meyerson, 1931; Köhler, 1927; Nell- This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. mann & Trendelenburg, 1926). Perceptual Feedback Recently, the understanding of connection was studied in a different fashion. Seven primate species were tested in three to The importance of perceptual feedback is fiercely debated cur- six trials under the contact/no contact condition described in the rently, with potentially major impact on the interpretation of many previous paragraph, as well as on two novel conditions. In one, string-pulling results. If successful string-pulling performance can the two disconnected strings touch; in the other, their ends result solely from associative learning, then direct perceptual feed- overlap (see Figure 4). General performance on these condi- back must be essential; because the pulling results in the reward tions was worse than on the contact/no contact condition, coming closer, it is repeated. This could explain all the classic subjects seem to have had problems distinguishing contact from variations on the string-pulling paradigm. The reward immediately connection (Amici et al., 2012; Herrmann et al., 2008). Adult comes closer, but only when the correct string is pulled. humans are also sensitive to degree of contact, but rely on Taylor and colleagues (2012) tested this hypothesis on New connection between rope and reward as well (Silva, Silva, Caledonian crows by giving them a choice between a coiled string Cover, Leslie, & Rubalcaba, 2008). with an attached reward and another coiled string that was unat- STRING-PULLING PARADIGM 107

tached to the reward (see Figure 5). Most subjects stopped pulling (2005) compared the performance of ravens on the standard ver- the correct string before the reward moved any closer. The authors tical string-pulling problem with a modified setup where the birds argue that the crows did not understand that the food could be must pull the string downward so the reward moves up. Although obtained by pulling the string, but instead that they rely on direct the two conditions require different actions, pulling the string visual feedback to solve such problems. Shown pictures of the always results in the food moving closer, thus establishing a setup, most humans consistently choose the correct string (Taylor perceptual feedback loop. Ravens that solved the standard vertical et al., 2012). In support of this hypothesis, an earlier study found problem also solved the pull-down condition, whereas naïve ra- that both experienced and naïve New Caledonian crows performed vens proved unable to do so. This shows that reinforcing visual more poorly when they did not look down before pulling and when feedback is not sufficient to solve a single-string-pulling problem. they had restricted visual access to the string during pulling (Tay- It is possible that the naïve ravens did not succeed because they lor, Medina, et al., 2010). had difficulty attending simultaneously to the string, the position Animals could rely on visual feedback in other paradigms, too; of the reward, and the string-pulling motion. The experienced for example, dropping stones or spitting water into a tube with a ravens were already adept at string pulling and so succeeded under floating reward results in the water level rising and the reward the pull-down condition. Poor attention might explain many string- moving closer (Taylor & Gray, 2009). Under the trap-tube para- pulling failures (see Table 2). Not surprisingly, seemingly more digm, most animals prefer to pull a reward toward themselves attentive species or individuals perform better (e.g., Beck, 1967; instead of pushing it away, even though they are equally successful Jolly, 1964a, 1964b; Nellmann & Trendelenburg, 1926; Warden et strategies. The reason might be attributable to the reinforcing al., 1940). effects of visual feedback (Seed, Hanus, & Call, 2011). Apes were Previous experience clearly helped the ravens solve the pull- observed as initially only able to solve a task involving the turning down task and might similarly aid neo-tropical parrots in string of a crank—functionally similar to vertical string-pulling—if they pulling under conditions without perceptual feedback (Taylor et could monitor the effects of their actions (Völter & Call, 2012). al., 2012). That said, as we previously suggested, experience may However, results from multiple species might point in the op- hinder solution of tasks without perceptual feedback. An animal posite direction. Some mammals do pull coiled strings (Adams, learns through conditioning that pulling a string results in food. 1929; Frank & Frank, 1985; Guillaume & Meyerson, 1931; Tested under a condition with no perceptual feedback, the condi- Klüver, 1961). Some primates will even pull a string where the tioned response may prove insufficient, more pulls are required, reward moves away before coming closer (Beck, 1967; Bolwig, and the subject sees no effect of the initial pulls on the reward’s 1963; Guillaume & Meyerson, 1931; Klüver, 1961). White-handed position. gibbons always solve this, albeit more slowly than a single straight If solving the single-string task is mediated solely by instrumen- string setup (Beck, 1967). Cotton-top tamarins, tested on pulling a tal conditioning toward strings, regardless of any attached reward, tape measure with attached food, subsequently performed at sim- then experienced animals should pull unbaited strings and respond ilar levels when the tape measure was covered such that no visual at random under the perpendicular condition; for example, expe- feedback was possible (Chapman & Weiss, 2013). Note, however, rienced goldfinches were observed to keep pulling on unbaited that most of these studies lack detailed descriptions and often use strings compulsively (Seibt & Wickler, 2006) and rats to pull the single, horizontally oriented strings, making the results difficult to same string they just obtained a reward from (Ewer, 1971). The interpret. A few studies stand out in contrast; for example, Schuck- issue boils down in the end to the effects of experience on Paim and colleagues (2009) found that certain neotropical parrots, problem-solving abilities; a complex question that requires much after having solved several patterned string problems, prefer to pull consideration. So far as we know, no testing has been carried out connected vertical strings over unconnected ones, even without on string-naïve humans, which would be highly informative, if immediate visual or proprioceptive feedback. Our own personal possible. observation of ravens, chimpanzees, and orangutans is that they Several variations on the string-pulling paradigm involve one reel in rewards attached to coiled strings even when they have no straight string and one longer string with one or multiple angles. previous string-pulling experience. The longer string needs to be pulled further for the reward to move, Visual feedback might be important but does not even appear to or the reward at first moves sideways because the string is guided be sufficient for successful string-pulling. Heinrich and Bugnyar by hooks. Yagi (1964) found that Formosan rock and Japanese 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.

Figure 5. The perpendicular coiled condition when applied to humans, who were asked which of the strings to pull. New Caledonian crows were tested on a similar setup, with food rewards. (Reprinted from Taylor et al., 2012.) 108 JACOBS AND OSVATH

macaques preferentially pull the straight strings in these tasks. 2010; Yagi, 1964). Three out of four parrot species tested failed to Unlike Yagi, many authors unfortunately only report number of solve the crossed condition when the same color was used; one errors and do not mention whether or not they found any signifi- species even showed a preference for the incorrect string. They cant difference in preference in these asymmetrical setups. Human could solve it when two colors were used (Krasheninnikova et al., infants prefer pulling a straight unattached string to an angled 2013). This might imply that they did not follow the strings attached string, though this may be because of general orientation. visually but associated the color of the string at the reward with the That said, once they start pulling the correct string, they do not color of the string at the perch. It could as well be that their visual stop—so continuous perceptual feedback does not appear to be system makes it difficult to separate the strings where they required (Richardson, 1932). crossed; knowing the visual systems of the subjects is, therefore, In some situations, illustrated by ice-fishing crows and ravens, important in the design of the tasks. the reward moves closer even while the animal cannot see it. Given When visual discrimination is made harder by placing string or the length of the fishing lines and the dark water, the corvids are rewards (or both) closer together, performance often deteriorates unable to see the fish coming closer while they pull (Holmberg, (Johnson & Michels, 1958; Mason & Harlow, 1961; Schmidt & 1957; Larsson, 1958). Another version involves the string hanging Cook, 2006; Warden et al., 1940)—though again, not always in a dark space such that its distal end cannot be seen (Capener, (Dücker & Rensch, 1977; King & Witt, 1966; Osthaus et al., 2011). This lack of visual continuity leads some authors to con- 2005). Because visual discrimination is so important for solving clude that—in contrast to standard string pulling—ice fishing is a string problems, it can be difficult to compare species with differ- form of tool use (Boswall, 1977; Lefebvre, Nicolakakis, & Boire, ent visual capacities. Examining four primate species, Harris and 2002; Van Lawick-Goodall, 1970). Whether classified as tool use Meyer (1971b) found that those with better visual acuity per- or not, these cases remain unclear. One does not know how the formed better on string-pulling problems. This did not hold when birds respond on their first trial or whether they rely on the weight the patterns involved multiple crossings, suggesting that some of the fish for perceptual feedback. animals do not follow strings visually when the patterns are too Not only visual but also proprioceptive feedback may contribute complex. to solution of the string-pulling task. Required to choose one of two perpendicular vertical strings based on the weight of the Functional Generalization and Fixedness identical-looking closed containers at their ends, most apes were observed to make correct choices but jackdaws did not, possibly Functional generalization—or the use of affordances—is im- because of the intermixed testing regime (Cimadom, 2013; Schrauf portant for flexible problem solving. By relying on relevant func- & Call, 2011). This mirrors findings with mammals tested on tional or structural aspects of the problem rather than arbitrary weight-discrimination problems involving horizontal string pulling cues, the subject can transfer this to a functionally similar task (e.g., Adams, 1933; Hayes, 1951; Klüver, 1961; McCulloch, 1934; easier. Given their abstractness, structural representations are less Povinelli, 2012). Klüver (1961) tested long-tailed macaques on a likely to be influenced by perceptual features (Call, 2013; Seed et coiled-string-pulling test with identical looking but differentially al., 2011). In the string-pulling paradigm, this can be tested by weighted boxes at the ends. Only the heavier box was rewarded. using strings of different color, texture, material, length, and so on. The monkeys chose a string at random and continued pulling if These changes generally do not affect performance, suggesting they felt that the box was heavy; otherwise they switched to the that most animals generalize the function of strings (Amici et al., other string. In the absence of visual feedback, they were able to 2012; Heinrich, 1995; Herrmann et al., 2008; Herter, 1940; King & use proprioceptive feedback to reach a solution. Likewise, 1-year- Witt, 1966; Klüver, 1961; Krasheninnikova et al., 2012, 2013; old human infants seem to be sensitive to both visual and propri- Whitt et al., 2009, but see Hertz, 1926; Krasheninnikova et al., oceptive feedback (Richardson, 1932). 2012, 2013; Schmidt & Cook, 2006; Vince, 1956). Such percep- On a side note, placing the reward in an opaque closed container tual changes do not interfere with the performance of preschoolers could be problematic. Some animals lack an understanding of either (Brown, 1990; Piaget, 1952). As a counterexample, Dücker object permanence (see Table 2) and might use for example, and Rensch (1977) reported on a common Myna that made almost olfactory cues to solve the task (Bierens de Haan, 1932; Heubel, no mistakes on several different string problems until the slightly 1939). Even for animals with an understanding of object perma- worn correct string was replaced. The bird was no longer able to This document is copyrighted by the American Psychological Association or one of its allied publishers. nence who have learned to associate a particular container with choose the correct string above chance in the perpendicular con- This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. food, motivation appears to be higher when they can observe the dition in a thousand trials. food (Fischel, 1936; Medina, 2012). All these perceptual changes involve thin, elongated objects. For true functional generalization, no such perceptual aspects should Perceiving Visual Continuity matter. As we noted previously, one finds many parallels between the support and string paradigms but they also involve important Another question is whether animals follow the strings visually perceptual differences; notably, the support paradigm uses a wide, when solving string-pulling problems. If they do, then the differ- flat object supporting a nonattached reward. ence in appearance of the strings should not matter much; but Performance under the string and support paradigms appears to solution rate is often higher when the two strings are perceptually be closely related (e.g., Amici et al., 2012; Herrmann et al., 2008; different (Heinrich, 1995; Johnson & Michels, 1958; Krashenin- Schmitt et al., 2012). Previous experience in one paradigm might nikova et al., 2012, 2013; Warden et al., 1940; Werdenich & improve performance in the other through functional generaliza- Huber, 2006)—though not always (Dücker & Rensch, 1977; Ma- tion. That said, dogs with string-pulling experience did not per- son & Harlow, 1961; Osthaus et al., 2005; Taylor, Medina, et al., form better than naïve dogs in the support paradigm (Müller, STRING-PULLING PARADIGM 109

Riemer, Virányi, Huber, & Range, 2014). Even within the string- Huber, 2006). Such variation suggests flexible problem solving pulling paradigm, experience with one orientation is not necessar- and functional generalization in obtaining an out-of-reach reward. ily beneficial for the other (Krasheninnikova & Schneider, 2014). Again, if an animal pulls a string expecting a reward, it need not Some macaques find it more difficult to solve a mixed series of have means-end understanding. In addition to the aforementioned crossed and pseudocrossed problems with different colored strings abilities, this requires the animal to recognize what it is capable of when a board covers the middle area—suggesting that functional pulling in (see Table 1). Adding a large and heavy reward might aspects of the task (whether the lines cross or not) are more test such recognition. Because of its size and value, such a reward important for them than color cues, even though both can result in will elicit more interest; on the other hand, because it is too heavy excellent performance (Yagi, 1964; see also Schmidt et al., 2006; to lift, inhibition is required not to waste effort on what cannot be Su, 1982). gained. Heinrich (1995) and Pfuhl (2012) report that a majority of Chimpanzees, bonobos, capuchin monkeys, and human children ravens never pulled a string with a heavy preferred reward, but alike were observed to perform better under the regular contact/no rather another string with a smaller reward. Some ravens flew at contact condition than when the strings were covered by a board the oversized reward, ripping pieces off, rather than continuing the on which broken and unbroken strings were placed so as to form previously successful behavior of pulling a string. New Caledonian reliable cues. Subjects could have solved the problem by pulling crows (Taylor, Medina, et al., 2010) and keas (Werdenich & the string under the board designated by a continuous string on top Huber, 2006) likewise mostly ignored the overloaded string, sup- of the board. The authors suggested that, because the subjects porting the conclusion that they have means-end understanding. understood the task’s functional properties, they saw the seemingly Complex cognitive mechanisms are often defined by their flex- nonfunctional perceptual cues as less relevant (Mayer et al., 2014). ibility. Basing solutions on functional rather than arbitrary cues is Apes also perform better when they are faced with real paper strips one indicator, but such functional solutions are not always flexible. compared with their painted equivalents in the support paradigm Sometimes a rigid rule is applied that is detrimental to solving (Albiach-Serrano, Bugnyar, & Call, 2012). other problems. This is another example of how previous experi- Because pulling a string typically involves many steps, analyz- ence can be disadvantageous for finding subsequent solutions. In ing the techniques used and errors made can be informative about the string-pulling paradigm, this functional fixedness is tested by how animals might functionally generalize problems. Pulling a presenting experienced animals with a string so long that the string horizontally is normally straightforward, nevertheless, one reward can be obtained directly from the ground or a perch. After finds variation in how it is pulled; for example, hand over hand, in 160 patterned trials, six of seven keas still pulled these long one long haul, or using the feet, tail, or teeth (Klüver, 1961; strings, rather than just taking the reward from the ground Warden et al., 1940). Many more techniques are used for vertical (Werdenich & Huber, 2006). The results for four other parrot string pulling (see Figure 6), up to eight in keas (Werdenich & species are mixed. Some individuals of all species adapted their 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.

Figure 6. The two most commonly used techniques by birds. Top: the side-step technique. Bottom: straight pull-up technique. (Reprinted from Heinrich, 1995.) 110 JACOBS AND OSVATH

behavior and went for the reward directly in the first trial (Krash- an effective, functional reformulation; transforming, and recom- eninnikova et al., 2013). In another experiment, budgerigars—but bining learned features; and suddenly realizing the solution as an none of the nine other previously successful parrot species—would “aha!” moment after a long impasse (Call, 2013; DeYoung, Flan- sometimes move to take the food directly rather than pull the string ders, & Peterson, 2008; Emery, 2013; Heinrich, 1995; Köhler, (Krasheninnikova et al., 2012). 1927; Kounios & Beeman, 2009; Shettleworth, 2012; Thorpe, It is difficult to know which of these negative results are because 1963). of functional fixedness or other factors (see Table 2). Lack of Experiments investigating insight in humans are difficult to motivation or inhibition could cause an animal to pull strings at apply to other animals because such experiments are normally random or to pull the closest one. The string’s intrinsic value may conducted on a verbal, mathematical, or spatial basis (Chu & be higher when the animal is less motivated (e.g., when it is sated) MacGregor, 2011). Thorpe’s (1963, p. 100) definition for animal so that a string with an attached food reward is only slightly more insight learning is the most commonly used, “. . . the sudden appealing than one without. It would be useful to test whether production of a new adaptive response not arrived at by trial subjects consistently pull the string with the closest food reward behavior or the solution of a problem by the sudden adaptive (e.g., Pfuhl, 2012). It would likewise be useful to determine what reorganization of experience.” their favorite food is, so it can be used in the experiments. The question whether one can ever know if an animal experi- Both chimpanzees and rats perform best when moderately food ences an aha! moment remains more philosophical than empirical. deprived (Crutchfield, 1939; Birch, 1945)—something one should By contrast, the role of experience is perhaps the most debated consider when dealing with poorly motivated animals. The pow- aspect of insightful problem solving and is, as mentioned several erful influence of motivation can be easily illustrated. Hamilton times above, highly relevant to the string-pulling paradigm. and Ellis (1933) reported that, after initial training, satiated rats A common drawback of many animal-insight studies is a failure pulled a rewarded string less than hungry rats pulled an unre- to explore why or how a supposedly insightful behavior occurred. warded string. Bird and Emery (2009) argued that their rooks showed insight in Many species will reel in unbaited strings (e.g., Altevogt, 1954; dropping stones into a tube to obtain a food reward. Later inves- Beck, 1967; Bierens de Haan, 1932; Hobhouse, 1915; Laidre, tigations on other corvids showed that factors as training and 2008; Nissani, 2004; Schmidt & Cook, 2006; Schuck-Paim et al., trial-and-error might have played a role (Cheke, Clayton, & Bird, 2009; Seibt & Wickler, 2006). One might hope to discover the 2011; von Bayern, Heathcote, Rutz, & Kacelnik, 2009). Of course, extent to which the strings are intrinsically rewarding by first everything boils down to what one means by insight, making it testing animals with a single unbaited string. Some animals may necessary to run multiple control tests to home in on the mecha- stop pulling as soon as they learn that there is no food reward; nisms of what is claimed to be insightful problem solving (Birch, others may be expected to maintain an interest in the string, as a 1945; Epstein, Kirshnit, Lanza, & Rubin, 1984; Seed & Boogert, form of toy. The negative reinforcing behavior, should the animals 2013; Shettleworth, 2009a, 2012; von Bayern et al., 2009). stop pulling the unrewarded string, becomes important to subse- Naturally, one must have experience with some aspects of a quent tests, where it must be overcome to solve any baited-string problem if one is to solve it. However, a solution with too much problems (e.g., Bierens de Haan, 1932; Taylor, Medina, et al., experience is unlikely to be called insightful. A useful rule of 2010). thumb is that the less experience a subject has when solving a Inhibiting string-pulling behavior with empty strings saves time novel problem, the more appropriately the solution is attributable and energy and so indicates behavioral flexibility. Inexperienced to insight. Where to draw the line though for the amount or type of animals may reel in empty strings as a form of exploration, the experience allowed for something still to be insightful is unclear. string holding an intrinsic value for them. A single piece of string Call (2013) distinguishes reasoning from learning, which differ in or rope evokes frequent—and variable—manipulations in ma- the kind of information acquired and how it is used for problem caques of all ages (Harlow et al., 1956; Mason et al., 1959; solving. Learning requires experience to be proximal in space and Torigoe, 1987). Exposing animals to unbaited strings is also an time, thereby increasing its associative strength. Reasoning does effective way of reducing neophobia, which can be a confounding not, instead relying on structural features. A mixture of both is factor, especially with many corvids (Heinrich, 2000; Heinrich & common, and it should be noted that both require some experience Bugnyar, 2005). in the first place. A quantitative approach identifying how much This document is copyrighted by the American Psychological Association or one of its allied publishers. Experienced animals may pull an unbaited string as the result of and what kind of information animals use for problem solving, This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. operant conditioning—as illustrated by the goldfinches’ compul- whether they solve mostly by learning or reasoning, whether their sive pulling (Seibt & Wickler, 2006). They may also pull an solutions are more specialized or more flexible, and so forth, has unbaited string as a form of play. If so, this is an especially advantages over a purely qualitative one. difficult issue to overcome, because repeated exposure to strings Investigations into animal insight have often relied on the string- will have little effect. Establishing whether or not animal subjects pulling paradigm (e.g., Beck, 1967; Bierens de Haan, 1933; Fis- are highly motivated by the reward in such cases is vital. chel, 1930; Harlow & Settlage, 1934; Heinrich, 1995; McDougall & McDougall, 1931; Pepperberg, 2004; Thorpe, 1943; Werdenich Insight & Huber, 2006). One way to demonstrate insight in the string- pulling paradigm would require an animal to solve new patterns The long history and ubiquitous use of the term “insight” aside, immediately and maintain this performance through an intermixed it remains an unclear cognitive phenomenon in humans as well as series. Others have suggested that overcoming functional fixedness other animals. Definitions typically include seeing into a situation; might be a pivotal aspect (Seed & Boogert, 2013; Shettleworth, viewing a problem from a novel, penetrating perspective leading to 2012). Perhaps the most convincing evidence would be an imme- STRING-PULLING PARADIGM 111

diate solution to the crossed condition, with no prior string-pulling Nonetheless, skills evolved or developed within a particular experience. On the other hand, if insight is defined so that it can ecology may prove useful for solving string-pulling problems. The only appear after an initial impasse, then the string-pulling para- examples of foot use in feeding aside, other psycho-motoric ad- digm is probably unsuitable. On first presentation of a patterned aptations seem to matter; for example, prosimian species perform string problem, the animal subject must show a novel behavior to better on single-string problems if they are manipulative, playful, obtain the reward; that is, it must be able to make the transition and do not feed exclusively on insects (Jolly, 1964a, 19964b). In from direct approaches to standing on the perch and pulling the general, ecological factors requiring persistence and patience and string (Shettleworth, 2012). Concerns about definitions aside, the behaviors like extractive foraging are beneficial to solving these behavior demonstrated still might not be called insightful, as it problems (Millikan & Bowman, 1967). For parrots, social organi- could also have been arrived at through chance, trial and error, zation is a better predictor of string-pulling performance than other visual feedback, innate processes, or a combination. On later ecological factors, or phylogeny or absolute or relative brain size presentations, the behavior is not novel anymore and so cannot be (Krasheninnikova, 2014; Krasheninnikova et al., 2013). The few called insightful. comparative studies into the relationship between ecology and Insight is sometimes seen as homologous to causal understand- string-pulling performance are scarce and provide diverse expla- ing, imagination, simulation, foresight, or mental trial and error. nations; therefore, more research in this area is warranted. Making these often vague terms synonymous obscures the skills Motor systems and anatomy differ greatly between species. For involved in reaching a solution, especially regarding whether or example, gibbons cannot pick up strings from flat surfaces because not the solution is arrived at suddenly (Shettleworth, 2012); and so of their elongated digits; the strings must protrude, as from a board insight loses explanatory power (Call, 2013; Kacelnik, 2009). (Beck, 1967). Fortunately, it is usually fairly obvious when an Causal understanding presents an exception to this rule, and may animal is struggling to grasp the string. This is one reason why best be investigated by testing how animals predict, and intervene measuring observation behavior makes a good addition to most on, different causal networks (Blaisdell & Waldmann, 2012). manipulation tasks. Human infants visually anticipate which string an adult should pull if they have previously been successful Ecology, Evolution, and Anatomy themselves (Rat-Fischer, O’Regan, & Fagard, 2014). Eye-tracking can now also be used on animals, providing a useful technology to The cognition involved in string pulling is sometimes down- study the link between anticipation and action. played by pointing at its close resemblance to natural behaviors. The motor system can be important even in the absence of For example, parids pull caterpillars by their threads (Brooks-King string-pulling difficulties. The superior performance of most pri- & Hurrell, 1958; Dickinson, 1969); various birds pull and step on mates over dogs might be attributable to dogs’ less tactile inter- twigs to get at insects, berries, or pine cones (Altevogt, 1954; actions with their environment, which delimits their embodied Krasheninnikova, 2013; Obozova et al., 2014; Seibt & Wickler, cognition (Holekamp, Swanson, & Van Meter, 2013). Prehension 2006; Thorpe, 1963); jays learn to pull oak seedlings to obtain the and means-end behavior might be less important for species that buried acorn (Bossema, 1979); crows and ravens pull unattended develop locomotion relatively quickly, because they can move fishing lines (Bagotskaya et al., 2010; Boswall, 1977; Burns, 1895; toward objects of interest instead of needing to obtain them Holmberg, 1957; Larsson, 1958; Loftin, 1959; Scott, 1974); chip- through other means (Antinucci, 1989). munks pull grass stems down to obtain their heads (Gordon, 1938); The most difficult part of string pulling for birds, motorically, elephants pull on trees to feed on the top foliage (Van Lawick- appears to be stepping on the string. Many have argued that such Goodall, 1970); baboons pull weavers’ nests to consume their eggs capacity only occurs in species that use their feet in feeding (e.g., (Laidre, 2008); and various primates pull vines or branches to Altevogt, 1954; Newton, 1967; Seibt & Wickler, 2006; Thiene- reach shoots and leaves (Abordo, 1976; Krasheninnikova, 2013; mann, 1933). Out of eight Australian parrot species sampled, the McDougall & McDougall, 1931; Van Lawick-Goodall, 1970), a only two that did not use their feet in feeding also failed a behavior that is responsible for 61% of the plant diet in wild string-pulling task (Magat & Brown, 2009, but see Krashenin- orangutans (Chevalier-Skolnikoff, Galdikas, & Skolnikoff, 1982). nikova, 2013), whereas the best string-pullers were those that were All this suggests that the string-pulling paradigm, or at least its more lateralized; that is, they had a strong preference to use either associated behaviors, has ecological relevance for a number spe- their left or right foot. Because lateralization of limbs reflects This document is copyrighted by the American Psychological Association or one of its allied publishers. cies—a relevance often thought to be lacking in other physical cerebral lateralization, it offers a promising avenue for research This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. cognition studies (Edwards, Rottman, & Santos, 2011). into the relationship between brain and behavior (Magat & Brown, If string-pulling techniques result mainly from hardwired be- 2009)—one that does not require such traditional methods as havior, they ought to be species-specific. This appears not to be the lesioning or radiation (e.g., Christensen & Pribram, 1977; Davis et case; for example, many species that use their feet in feeding are al., 1958; Harlow & Settlage, 1948). reported to use various techniques when pulling strings—both Great gray owls, keas, and Harris hawks—all of which rely between and within individuals (e.g., Bagotskaya et al., 2010; more than passerines on foot use for feeding (Sustaita et al., Colbert-White et al., 2013; Heinrich, 1995; Krasheninnikova & 2013)—have been reported to sometimes only use their feet to pull Wanker, 2010; Laidre, 2008; Medina, 2012; Obozova & Zorina, strings (Colbert-White et al., 2013; Obozova et al., 2014; 2013; Obozova et al., 2014; Seibt & Wickler, 2006; Taylor, Werdenich & Huber, 2006). Against this, some individual birds Medina, et al., 2010; Thorpe, 1963; Werdenich & Huber, 2006). who use their feet for feeding fail to pull the string (e.g., Seibt & The fact that there are considerable inter- and intraindividual Wickler, 2006; Vince, 1956), whereas others who do not use their variation and strong learning effects make it unlikely that string feet for feeding do pull the string (e.g., Krasheninnikova, 2013; pulling is governed completely by innate processes. Thorpe, 1963). Though the use of feet in feeding is useful for 112 JACOBS AND OSVATH

predicting string-pulling performance in birds, clearly it is neither string-pulling abilities remains to be investigated, whether by necessary nor sufficient. meta-analysis or large-scale empirical investigations. Wasserman and colleagues (2013) suggested a virtual, touch- An important conclusion from these meta-analyses is that the screen-based string-pulling task to be used for animals with ana- phylogenetic signal is not necessarily strong. A species does not tomical limitations on string pulling. They used this setup to test always perform more similarly to a closer than a more distant pigeons and got results similar to those in a real setup (see also relative. Although absolute and relative brain size in parrots both Brzykcy, Wasserman, Nagasaka, & Perez-Acevedo, 2014). It is have a high phylogenetic signal—like anatomical traits in too early to tell how close an equivalent this virtual test is to the general—string-pulling performance does not. This suggests that original test. However, the string-pulling paradigm poses a phys- socioecological variables are more prominent than phylogeny ical problem, so excluding its sensorimotor aspect likely trans- (Krasheninnikova, 2014) and adds to growing evidence that cog- forms it into a fundamentally different task. Notably, adult humans nition has reached similar levels in distantly related species, respond differently to physical compared with symbolic represen- through convergent or parallel evolution (e.g., Osvath, Kabadayi, tations of the string-pulling paradigm (Silva et al., 2008). & Jacobs, 2014). Great tits are commonly used as model animals in behavioral ecology and are the most tested bird species on the string-pulling Concluding Remarks and Recommendations for paradigm. One of the largest studies to date related their perfor- Future Research mance to life history and ecological variables (Cole, Cram, & Quinn, 2011). Of 365 individuals tested on the single string We have attempted in this review to clarify what cognitive condition, 93 (25%) solved it within a 60-min trial. Their perfor- mechanisms and other factors are thought to be involved in string mance was repeatable over 1 to 2 years and correlated positively pulling (see Tables 1 and Table 2), as an aid to anyone wishing to with performance on a lever-pulling task, also based on operant further explore the paradigm. In concluding, we wish to offer some conditioning and a perceptual/motor feedback loop (Cole, Morand- recommendations for future research. Ferron, Hinks, & Quinn, 2012). Performance was unrelated to For all the vast number of publications involving string-pulling, neophobia, exploration, sex, body conditions, or motivation to feed surprisingly little is known about the precise cognitive mechanisms behind the solutions that animals find for the various pat- after human disturbance, with only minor effects of age and natal terns. This is generally attributable to a lack of well-controlled, origin (Cole et al., 2011). Great tit cognitive skills, as measured in clearly reported investigations involving large samples, and with the string- and lever-pulling paradigms, are thus independent of sound statistics, well-defined theories, and varieties of string pat- most ecological measures and stable in the individual. terns. Consequently, similar performances have elicited widely This is a good example of how large-scale studies can be more divergent cognitive explanations, and similar explanations have informative than the sum of many smaller ones. Testing conditions been proposed for divergent performances. In studies where string- were consistent, making the results reliably comparable. Many pulling was not the primary focus—such as test batteries or inves- time-consuming variables could be investigated, such as repeat- tigations into sensorimotor development or the effects of brain ability over time and performance on multiple tests. They show as lesions—the detailed results are sometimes not even reported, well how the string-pulling paradigm can be useful to behavioral which is unfortunate as these studies often have the largest sample ecology. Even in captive conditions, it is a valuable tool for sizes. studying the effects of factors such as dominance, sociality, bold- Despite long and widespread use of the string-pulling paradigm, ness, pair bonding, scrounging, and tolerance (e.g., Humle & a solid testing protocol is lacking, undermining its comparative Snowdon, 2008; Jolles et al., 2013; Krasheninnikova & Schneider, strengths (MacLean et al., 2012). Of course, this is true for many 2014; Pfuhl et al., 2014); for example, subordinate rooks make other paradigms as well, whereas the problem of interspecies fewer errors, improve faster, and explore the string less than comparison is inherent to the nature of comparative psychology. dominant rooks (I.F. Jacobs, unpublished data; see also Jolles et The string-pulling paradigm might appear to lend itself well to al., 2013). comparisons between vastly different species, given its simplicity As we have pointed out throughout this article, performance on and easily measured outcomes. However, as we have shown, string-pulling problems can be obscured by a variety of factors several factors need to be taken into consideration (e.g., differ- This document is copyrighted by the American Psychological Association or one of its allied publishers. (see Table 2). Meaningful comparisons can still be made so long ences in anatomy, perception, and attention) allowing that, depend- This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. as these issues are taken into account. One can obtain results of fair ing on one’s views on cognition, such factors may well be intrinsic reliability by comparing performance of different species in a to the agent’s cognition. Trying to find any pure cognition de- single study, or across studies with sufficiently similar methodol- tached from the animal’s predispositions is, at the least, a very ogies—taking into account as many relevant anatomical and psy- difficult task. With this in mind, detailed comparison of string- chological features as possible. pulling performance, with variations in the conditions, might have In a meta-analysis of five primate studies, Deaner, van Schaik, its greatest potential in looking at relatively closely related species and Johnson (2006) found the following ordering of abilities, from sharing anatomical and behavioral similarities. Within a group of least to most string-pulling errors: Pongo, Pan, Ateles, Gorilla, closely related taxa, this could prove a fruitful way for measuring Cercocebus/Cercopithecus/Macaca/Mandrillus, Cebus, and Pa- divergences. pio. This result accords with their overall findings on primate That said, comparisons between distantly related species could physical cognition and indicates that the string-pulling paradigm is also be informative—in particular, when it comes to broader a suitable predictor for the whole domain (cf. Reader, Hager, & questions concerning ecology and adaptations to different envi- Laland, 2011). The nature of the relationship between species and ronments. In embarking on such studies, it is particularly important STRING-PULLING PARADIGM 113

that one first has an explicit idea of the underlying theories on References cognition one is attempting to use. This helps in defining what one Abordo, E. J. (1976). The learning skills of gibbons. In D. M. Rumbaugh is measuring. It is also valuable in fostering scientific debate on (Ed.), Gibbon and Siamang (pp. 106–134). Basel: Karger. when considering results from string-pulling studies that may well Adams, D. K. (1929). Experimental studies of adaptive behaviour in cats. differ in their underlying theoretical assumptions. Comparative Psychology Monographs, 6, 1–128. Species comparisons are likely more reliable when made as part Adams, D. K. (1933). Weight discrimination in rats. Psychological Bulle- of the same research project, following a coherent protocol. For- tin, 30, 703. tunately, this is becoming common practice. Clearer reports also Akert, K., Orth, O. S., Harlow, H. F., & Schiltz, K. A. (1960). Learned make comparisons between different studies more reliable, with behavior of Rhesus monkeys following neonatal bilateral prefrontal the caveat that almost never do these involve exact protocol lobotomy. Science, 132, 1944–1945. http://dx.doi.org/10.1126/science replications, which could be a problem particularly when investi- .132.3444.1944 Albiach-Serrano, A., Bugnyar, T., & Call, J. (2012). Apes (Gorilla gorilla, gating phylogenetic questions. In comparing distantly related taxa, Pan paniscus, P. troglodytes, Pongo abelii) versus corvids (Corvus a general recommendation is that patterns be administered hori- corax, C. corone) in a support task: The effect of pattern and function- zontally, which appears especially important when comparing ality. Journal of Comparative Psychology, 126, 355–367. http://dx.doi mammals and birds. .org/10.1037/a0028050 Another recommendation is that, when studying many different Altevogt, R. (1954). U¨ ber das “schöpfen” einiger vogelarten [About the and distantly related species in one study, one should not include “drawing” of some bird species]. Behaviour, 6, 147–152. http://dx.doi too many patterns, as this increases not only the practical but also .org/10.1163/156853954X00086 the theoretical difficulties. The perpendicular condition on its own Amici, F., Barney, B., Johnson, V. E., Call, J., & Aureli, F. (2012). A modular mind? A test using individual data from seven primate species. is a good option for revealing species differences in goal direct- PLoS ONE, 7, e51918. http://dx.doi.org/10.1371/journal.pone.0051918 edness, which could then be correlated with ecological or other Antinucci, F. (1989). Systematic comparison of early sensorimotor devel- overall differences between the species. opment. In F. Antinucci (Ed.), Cognitive structure and development in If one has the ambition to broaden understanding of how dif- nonhuman primates (pp. 67–85). Hillsdale, NJ: Erlbaum. ferent animal species perform string-pulling tasks, many taxa Antinucci, F., Spinozzi, G., Visalberghi, E., & Volterra, V. (1982). Cog- remain unrepresented or underrepresented (e.g., prosimians, ceta- nitive development in a Japanese macaque (Macaca fuscata). Annali ceans, marsupials, cephalopods, reptiles, and nonpsittacine non- dell’. Istituto Superiore di Santita, 18, 177–184. passerine birds). More research on how human infants (and adults) Atkinson, K. (1994). Fruit stringer for bats. The Shape of Enrichment, 3, 10. solve patterned string problems would further improve compara- Audubon, J. J. (1831). Ornithological biography. Edinburgh, Scotland: bility. Judah Dobson. If one wishes to study the role of experience in string pulling, Auersperg, A. M. I., von Bayern, A. M. P., Gajdon, G. K., Huber, L., & then another if obvious recommendation is that a relatively large Kacelnik, A. (2011). Flexibility in problem solving and tool use of kea sample size is favorable, allowing one to vary the testing order and New Caledonian crows in a multi access box paradigm. PLoS ONE, among individuals and determine the influence of other factors 6, e20231. likely to affect performance (see Table 2). Bagotskaya, M. S., Smirnova, A. A., & Zorina, Z. A. (2010). Comparative Although the single-string condition is the most widely used, it study of the ability to solve a string-pulling task in Corvidae. Zhurnal Vysshei Nervnoi Deiatelnosti Imeni I P Pavlova, 60, 321–329. is also the most debated, because success in it can result from Bagotskaya, M. S., Smirnova, A. A., & Zorina, Z. A. (2012). Corvidae can many different cognitive mechanisms, making it a relatively un- understand logical structure in baited string-pulling tasks. Neuroscience informative test on its own. Patterned string tasks are more reveal- and Behavioral Physiology, 42, 36–42. http://dx.doi.org/10.1007/ ing (see Table 1), with the perpendicular, slanted, crossed, and s11055-011-9529-z contact/no-contact conditions being the most important (see Figure Balasch, J., Sabater-Pi, J., & Padrosa, T. (1974). Perceptual learning ability 4). Testing whether subjects rely on perceptual feedback is clearly in Mandrillus sphinx and Cercopithecus nictitans. Revista Espanola de valuable, as well as the exploring the extent to which they can Fisiologia, 30, 15–20. generalize flexibly. Beck, B. B. (1967). A study of problem solving by gibbons. Behaviour, 28, 95–109. http://dx.doi.org/10.1163/156853967X00190 This document is copyrighted by the American Psychological Association or one of its allied publishers. The string-pulling paradigm is widely used in comparative Bierens de Haan, J. A. (1930). U¨ ber das Suchen nach verstecktem Futter This article is intended solely for the personal use ofpsychology the individual user and is not to be disseminated broadly. for many good reasons and so is often to be preferred bei Affen und Halbaffen [About looking for hidden food in monkeys and over other physical-cognition paradigms. It is one of the easiest prosimians]. Zeitschrift fur Vergleichende Physiologie, 11, 630–655. paradigms to execute, requiring few materials and little training. It Bierens de Haan, J. A. (1932). U¨ ber das Suchen nach verstecktem Futter is a straightforward manipulation-based test for some of the first- bei einigen Procyoniden und einem Eichhörnchen [About looking for to-develop and most basic cognitive mechanisms, making it well hidden food in some procyonids and a squirrel]. Zeitschrift fur Ver- suited for studying cognitive development across a range of spe- gleichende Physiologie, 17, 279–303. cies, which, among other things, is central to understanding the Bierens de Haan, J. A. (1933). Der Stieglitz als Schöpfer [The goldfinch as independent evolution of cognition (Osvath et al., 2014). Its basic a puller]. Journal für Ornithologie, 81, 1–22. http://dx.doi.org/10.1007/ BF01932163 design affords endless variations, offering insights into different Birch, H. G. (1945). The role of motivational factors in insightful problem- cognitive abilities and the role of various ecological and evolu- solving. Journal of Comparative Psychology, 38, 295–317. http://dx.doi tionary factors. It is not surprising that the string-pulling paradigm .org/10.1037/h0059937 has been used for so long and promises to continue being a useful Bird, C. D., & Emery, N. J. (2009). Insightful problem solving and creative tool in the future—at least if used in a considerate manner. tool modification by captive nontool-using rooks. Proceedings of the 114 JACOBS AND OSVATH

National Academy of Sciences of the United States of America, 106, preliminary report. Journal of Human Evolution, 11, 639–652. http:// 10370–10375. http://dx.doi.org/10.1073/pnas.0901008106 dx.doi.org/10.1016/S0047-2484(82)80010-9 Blaisdell, A. P., & Waldmann, M. R. (2012). Rational rats: Causal infer- Christensen, C. A., & Pribram, K. H. (1977). The visual discrimination ence and representation. In T. R. Zentall & E. A. Wasserman (Eds.), The performance of monkeys with foveal prestriate and inferotemporal le- Oxford handbook of comparative cognition (pp. 175–198). Oxford: sions. Physiology & Behavior, 18, 403–407. http://dx.doi.org/10.1016/ Oxford University Press. 0031-9384(77)90251-7 Bolwig, N. (1963). Observations on the mental and manipulative abilities Chu, Y., & MacGregor, J. N. (2011). Human performance on insight of a captive baboon (Papio doguera). Behaviour, 22, 24–40. http://dx problem solving: A review. The Journal of Problem Solving, 3, 119– .doi.org/10.1163/156853963X00293 150. http://dx.doi.org/10.7771/1932-6246.1094 Bonney, K. R., & Wynne, C. D. L. (2002). Quokkas (Setonix brachyurus) Cimadom, A. (2013). String pulling and discrimination learning in jack- demonstrate tactile discrimination learning and serial-reversal learning. daws (Corvus monedula) and New Caledonian crows (Corvus monedu- Journal of Comparative Psychology, 116, 51–54. http://dx.doi.org/ loides) (Master’s thesis). University of Vienna, Vienna, Austria. 10.1037/0735-7036.116.1.51 Clementius, B. (1933). Puttertje [Little goldfinch]. De Levende Natuur, 38, Bossema, I. (1979). Jays and oaks: An eco-ethological study of a symbiosis. 99–100. Behaviour, 70, 1–116. http://dx.doi.org/10.1163/156853979X00016 Colbert-White, E. N., McCord, E. M., Sharpe, D. I., & Fragaszy, D. M. Boswall, J. (1977). Tool-using by birds and related behaviour. Avicultural (2013). String-pulling behaviour in a Harris’s Hawk Parabuteo unicinc- Magazine, 83, 88–97. tus. The Ibis, 155, 611–615. http://dx.doi.org/10.1111/ibi.12040 British Pathé. (1932). Plenty of “pull” [Motion picture]. Retrieved from Cole, E. F., Cram, D. L., & Quinn, J. L. (2011). Individual variation in http://www.britishpathe.com/video/plenty-of-pull spontaneous problem-solving performance among wild great tits. Ani- Brooks-King, M. (1941). Intelligence tests with tits. British Birds, 35, 29–32. mal Behaviour, 81, 491–498. http://dx.doi.org/10.1016/j.anbehav.2010 Brooks-King, M., & Hurrell, H. G. (1958). Intelligence tests with tits. .11.025 British Birds, 51, 514–526. Cole, E. F., Morand-Ferron, J., Hinks, A. E., & Quinn, J. L. (2012). Brown, A. L. (1990). Domain-specific principles affect learning and trans- Cognitive ability influences reproductive life history variation in the fer in children. Cognitive Science, 14, 107–133. wild. Current Biology, 22, 1808–1812. http://dx.doi.org/10.1016/j.cub Brückner, G. H. (1933). Anhang [Appendix]. Zeitschrift fur Psychologie .2012.07.051 mit Zeitschrift fur Angewandte Psychologie, 130, 398–399. Cross, A. (1947). Intelligence of rook. British Birds, 40, 273–274. Brzykcy, S. J., Wasserman, E. A., Nagasaka, Y., & Perez-Acevedo, S. Crozier, G. B. (1910). Pets: Tricks for pet birds. Every woman’s encylo- (2014). Validating the virtual string task with the gap test. Animal paedia. London, England: London S. N. Cognition, 17, 1427–1431. http://dx.doi.org/10.1007/s10071-014- Crutchfield, R. S. (1939). The determiners of energy expenditure in string- 0769-5 pulling by the rat. The Journal of Psychology, 7, 163–178. http://dx.doi Burns, F. L. (1895). The American crow. The Wilson Bulletin, 7, 5–41. .org/10.1080/00223980.1939.9917626 Buttelmann, D., Carpenter, M., Call, J., & Tomasello, M. (2008). Rational Davis, R. T., Lovelace, W. E., & McKenna, V. V. (1964). Size constancy as a tool use and tool choice in human infants and great apes. Child Devel- function of ageing and x-ray irradiation in rhesus monkeys. Animal Behaviour, opment, 79, 609–626. http://dx.doi.org/10.1111/j.1467-8624.2008 12, 16–24. http://dx.doi.org/10.1016/0003-3472(64)90096-X .01146.x Davis, R. T., & McDowell, A. A. (1953). Performance of monkeys on Call, J. (2013). Three ingredients for becoming a creative tool user. In randomly presented string problems. Proceedings of the South Dakota C. M. Sanz, J. Call, & C. Boesch (Eds.), Tool use in animals: Cognition Academy of Science, 32, 147–152. and ecology (pp. 3–20). Cambridge, England: Cambridge University Davis, R. T., McDowell, A. A., Deter, C. W., & Steele, J. P. (1956). Press. http://dx.doi.org/10.1017/CBO9780511894800.002 Performance of rhesus monkeys on selected laboratory tasks presented Callaghan, J. C., McQueen, D. A., Scott, J. W., & Bigelow, W. G. (1954). before and after a large single dose of whole-body x radiation. Journal Cerebral effects of experimental hypothermia. Archives of Surgery, 68, of Comparative and Physiological Psychology, 49, 20–26. http://dx.doi 208–215. http://dx.doi.org/10.1001/archsurg.1954.01260050210008 .org/10.1037/h0047480 Capener, J. (Producer). (2011). Bird brain [Motion picture]. England: Terra Davis, R. T., McDowell, A. A., Grodsky, M. A., & Steele, J. P. (1958). The Mater Factual Studios. performance of x-ray irradiated and non-irradiated rhesus monkeys Cha, J., & King, J. E. (1969). The learning of patterned string problems by before, during, and following chronic barbiturate sedation. The Journal squirrel monkeys. Animal Behaviour, 17, 64–67. http://dx.doi.org/ of Genetic Psychology, 93, 37–51. http://dx.doi.org/10.1080/00221325 10.1016/0003-3472(69)90113-4 .1958.10532402 Chapman, K. M., & Weiss, D. J. (2013). Pulling to scale: Motor planning This document is copyrighted by the American Psychological Association or one of its allied publishers. Davis, R. T., & Steele, J. P. (1963). Performance selection through radia- for sequences of repeated actions by cotton-top tamarins (Saguinus This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. tion death in rhesus monkeys. The Journal of Psychology, 56, 119–136. oedipus). Journal of Experimental Psychology: Animal Behavior Pro- http://dx.doi.org/10.1080/00223980.1963.9923707 cesses, 39, 180–186. http://dx.doi.org/10.1037/a0031775 Cheke, L. G., Bird, C. D., & Clayton, N. S. (2011). Tool-use and instru- Deaner, R. O., van Schaik, C. P., & Johnson, V. (2006). Do some taxa have mental learning in the Eurasian jay (Garrulus glandarius). Animal better domain-general cognition than others? A meta-analysis of nonhu- Cognition, 14, 441–455. http://dx.doi.org/10.1007/s10071-011-0379-4 man primate studies. Evolutionary Psychology, 4, 149–196. Chevalier-Skolnikoff, S. (1982). A cognitive analysis of facial behavior in de Jong, H. (1919). Recherches sur la formation d’idées chez le chien Old World monkeys, apes, and human beings. In C. T. Snowdon, C. H. [Research on the formation of ideas in the dog]. Archives néerlandais de Brown, & M. R. Petersen (Eds.), Primate communication (pp. 303–368). Physiologie, 3, 491–572. Cambridge: Cambridge University Press. DeYoung, C. G., Flanders, J. L., & Peterson, J. B. (2008). Cognitive Chevalier-Skolnikoff, S. (1983). Sensorimotor development in orang-utans abilities involved in insight problem solving: An individual differences and other primates. Journal of Human Evolution, 12, 545–561. http:// model. Creativity Research Journal, 20, 278–290. http://dx.doi.org/ dx.doi.org/10.1016/S0047-2484(83)80034-7 10.1080/10400410802278719 Chevalier-Skolnikoff, S., Galdikas, B. M., & Skolnikoff, A. Z. (1982). The Dickinson, J. C., Jr. (1969). A string-pulling tufted titmouse. The Auk, 86, adaptive significance of higher intelligence in wild orang-utans: A 559. http://dx.doi.org/10.2307/4083422 STRING-PULLING PARADIGM 115

Dillis, C., Humle, T., & Snowdon, C. T. (2010). Socially biased learning Frank, H., & Frank, M. G. (1985). Comparative manipulation-test perfor- among adult cottontop tamarins (Saguinus oedipus). American Journal mance in ten-week-old wolves (Canis lupus) and Alaskan Malamutes of Primatology, 72, 287–295. (Canis familiaris): A Piagetian interpretation. Journal of Comparative Doré, F. Y., & Dumas, C. (1987). Psychology of animal cognition: Piag- Psychology, 99, 266–274. http://dx.doi.org/10.1037/0735-7036.99.3 etian studies. Psychological Bulletin, 102, 219–233. http://dx.doi.org/ .266 10.1037/0033-2909.102.2.219 Funk, M. S. (2002). Problem solving skills in young yellow-crowned Drescher, K., & Trendelenburg, W. (1927). Weiterer Beitrag zur Intelli- parakeets (Cyanoramphus auriceps). Animal Cognition, 5, 167–176. genzprüfung an Affen (einschliesslich Anthropoiden) [Further contribu- http://dx.doi.org/10.1007/s10071-002-0149-4 tion to intelligence examination on monkeys (including anthropoids)]. Gagne, M., Levesque, K., Nutile, L., & Locurto, C. (2012). Performance on Zeitschrift fur Vergleichende Physiologie, 5, 613–642. http://dx.doi.org/ patterned string problems by common marmosets (Callithrix jacchus). 10.1007/BF00298986 Animal Cognition, 15, 1021–1030. http://dx.doi.org/10.1007/s10071- Dücker, G., & Rensch, B. (1977). The solution of patterned string problems 012-0511-0 by birds. Behaviour, 62, 164–173. http://dx.doi.org/10.1163/ Gonzalez, R. C., Gentry, G. V., & Bitterman, M. E. (1954). Relational 156853977X00081 discrimination of intermediate size in the chimpanzee. Journal of Com- Dumas, C., & Doré, F. Y. (1991). Cognitive development in kittens (Felis parative and Physiological Psychology, 47, 385–388. http://dx.doi.org/ catus): An observational study of object permanence and sensorimotor 10.1037/h0058811 intelligence. Journal of Comparative Psychology, 105, 357–365. http:// Gordon, K. (1938). Observations on the behavior of Callospermophilus and dx.doi.org/10.1037/0735-7036.105.4.357 Eutamias. Journal of Mammalogy, 19, 78–84. http://dx.doi.org/10.2307/ Edwards, B. J., Rottman, B. M., & Santos, L. R. (2011). The evolutionary 1374284 origins of causal cognition. In T. McCormack, C. Hoerl, & S. Butterfill Grzimek, B. (1942). Weitere Vergleichsversuche mit Wolf und Hund (Eds.), Tool use and causal cognition (pp. 111–128). Oxford, England: [Further comparative experiments with wolf and dog]. Zeitschrift für Oxford University Press. http://dx.doi.org/10.1093/acprof:oso/ Tierpsychologie, 5, 59–73. 9780199571154.003.0006 Guillaume, P., & Meyerson, I. (1931). L’usage de l’instrument chez les Ellison, A. M., Watson, J., & Demers, E. (2015). Testing problem solving singes. II. L’intermediaire lie a l’objet [Tool use in monkeys. II. The means linked to the goal]. Journal de Psychologie Normale et in turkey vultures (Cathartes aura) using the string-pulling test. Animal Pathologique, 28, 481–555. Cognition, 18, 111–118. http://dx.doi.org/10.1007/s10071-014-0782-8 Hall, K. R. L., & Mayer, B. (1966). Hand preferences and dexterities of Emery, N. J. (2013). Insight, imagination and invention: Tool understand- captive patas monkeys. Folia Primatologica, 4, 169–185. http://dx.doi ing in a non-tool-using corvid. In C. M. Sanz, J. Call, & C. Boesch .org/10.1159/000155050 (Eds.), Tool use in animals: Cognition and ecology (pp. 67–88). Cam- Hallock, M. B., & Worobey, J. (1984). Cognitive development in chim- bridge, England: Cambridge University Press. http://dx.doi.org/10.1017/ panzee infants (Pan troglodytes). Journal of Human Evolution, 13, CBO9780511894800.006 441–447. http://dx.doi.org/10.1016/S0047-2484(84)80056-1 Epstein, R., Kirshnit, C. E., Lanza, R. P., & Rubin, L. C. (1984). ‘Insight’ Halsey, L. G., Bezerra, B. M., & Souto, A. S. (2006). Can wild common in the pigeon: Antecedents and determinants of an intelligent perfor- marmosets (Callithrix jacchus) solve the parallel strings task? Animal mance. Nature, 308, 61–62. http://dx.doi.org/10.1038/308061a0 Cognition, 9, 229–233. http://dx.doi.org/10.1007/s10071-006-0016-9 Erhardt, A. (1933). Kritische Bemerkungen zu der Arbeit von Bierens de Hamilton, J. A., & Ellis, W. D. (1933). Persistence and behavior constancy. Haan “Der Stieglitz als Schöpfer” [Critical comments on the work of The Pedagogical Seminary and Journal of Genetic Psychology, 42, Bierens de Haan “The goldfinch as a puller”]. Zeitschrift fur Psychologie 140–153. http://dx.doi.org/10.1080/08856559.1933.10534233 mit Zeitschrift fur Angewandte Psychologie, 130, 393–398. Harlow, H. (1939). Recovery of pattern discrimination in monkeys follow- Ewer, R. F. (1971). The biology and behaviour of a free-living population ing unilateral occipital lobectomy. Journal of Comparative Psychology, of black rats (Rattus rattus). Animal Behaviour Monographs, 4, 125– 27, 467–489. http://dx.doi.org/10.1037/h0062933 174. http://dx.doi.org/10.1016/S0066-1856(71)80002-X Harlow, H., Blazek, N. C., & McClearn, G. E. (1956). Manipulatory Fife, D. D., & Kamback, M. (1970). The ability of squirrel monkeys to motivation in the infant rhesus monkey. Journal of Comparative and solve randomly presented string problems. Psychonomic Science, 18, Physiological Psychology, 49, 444–448. http://dx.doi.org/10.1037/ 131–132. http://dx.doi.org/10.3758/BF03332337 h0047817 Finch, G. (1941a). Chimpanzee handedness. Science, 94, 117–118. http:// Harlow, H. F., Schiltz, K. A., & Settlage, P. H. (1955). Effect of cortical dx.doi.org/10.1126/science.94.2431.117 implantation of radioactive cobalt on learned behavior of rhesus mon- Finch, G. (1941b). The solution of patterned string problems by chimpan- keys. Journal of Comparative and Physiological Psychology, 48, 432– zees. Journal of Comparative Psychology, 32, 83–90. http://dx.doi.org/ 436. http://dx.doi.org/10.1037/h0044104 This document is copyrighted by the American Psychological Association or one of its allied publishers. 10.1037/h0056785 Harlow, H. F., & Settlage, P. H. (1934). Comparative behavior of primates. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Fischel, W. (1930). Weitere Untersuchung der Ziele der tierischen Hand- VII. Capacity of monkeys to solve patterned string tests. Journal of lung [Further examination of the goals of animal action]. Journal of Comparative Psychology, 18, 423–435. http://dx.doi.org/10.1037/ Comparative Physiology, 11, 523–548. h0073305 Fischel, W. (1933). Das Verhalten von Hunden bei doppelter Zielsetzung Harlow, H. F., & Settlage, P. H. (1948). Effect of extirpation of frontal und doppelter Handlungsmöglichkeit [The behavior of dogs on double areas upon learning performance of monkeys. Research Publications objectives and double action possibilities]. Zeitschrift fur Vergleichende Association for Research in Nervous and Mental Disease, 27, 446–459. Physiologie, 19, 170–182. Harris, D. G., & Meyer, M. E. (1971a). Performance of squirrel monkeys Fischel, W. (1936). Die Gedächtnisleistungen der Vögel [The memory on systematic or random presentation of patterned string problems. performance of birds]. Zeitschrift für Züchtung Reihe B, Tierzüchtung Psychonomic Science, 22, 158–160. http://dx.doi.org/10.3758/ und Züchtungsbiologie einschlie␤lich Tierernährung, 36, 13–38. http:// BF03332545 dx.doi.org/10.1111/j.1439-0388.1936.tb00081.x Harris, D. G., & Meyer, M. E. (1971b). The relationship between visual Fischer, G. J., & Kitchener, S. L. (1965). Comparative learning in young acuity and performance on patterned string problems by infrahuman gorillas and orang-utans. The Journal of Genetic Psychology, 107, primates. Psychonomic Science, 22, 160. http://dx.doi.org/10.3758/ 337–348. http://dx.doi.org/10.1080/00221325.1965.10533672 BF03332546 116 JACOBS AND OSVATH

Hayes, C. (1951). The ape in our house. London: Victor Gollancz Ltd. Johnson, J. I., Jr., & Michels, K. M. (1958). Discrimination of small Hector the ‘mule’ can’t be, but he is. (1950, May 20). The Courier-Mail, intervals and objects by raccoons. Animal Behaviour, 6, 164–170. http:// p. 3. Retrieved from http://trove.nla.gov.au/ndp/del/article/49730069 dx.doi.org/10.1016/0003-3472(58)90047-2 Heinrich, B. (1995). An experimental investigation of insight in common Jolles, J. W., Ostojic,´ L., & Clayton, N. S. (2013). Dominance, pair bonds ravens (Corvus corax). The Auk, 112, 994–1003. http://dx.doi.org/ and boldness determine social-foraging tactics in rooks, Corvus frugile- 10.2307/4089030 gus. Animal Behaviour, 85, 1261–1269. http://dx.doi.org/10.1016/j Heinrich, B. (1999). Mind of the raven: Investigating and adventures with .anbehav.2013.03.013 wolf-birds. New York, NY: Harper Collins. Jolly, A. (1964a). Choice of cue in prosimian learning. Animal Behaviour, Heinrich, B. (2000). Testing insight in ravens. In C. Heyes & L. Huber 12, 571–577. http://dx.doi.org/10.1016/0003-3472(64)90081-8 (Eds.), Evolution of cognition (pp. 289–305). Cambridge, MA: MIT Jolly, A. (1964b). Prosimians’ manipulation of simple object problems. Press. Animal Behaviour, 12, 560–570. http://dx.doi.org/10.1016/0003- Heinrich, B., & Bugnyar, T. (2005). Testing problem solving in ravens: 3472(64)90080-6 Kacelnik, A. (2009). Tools for thought or thoughts for tools? Proceedings String-pulling to reach food. Ethology, 111, 962–976. http://dx.doi.org/ of the National Academy of Sciences of the United States of America, 10.1111/j.1439-0310.2005.01133.x 106, 10071–10072. http://dx.doi.org/10.1073/pnas.0904735106 Herbert, M., & Harsh, C. (1944). Observational learning by cats. Journal Kawamori, A., & Matsushima, T. (2012). Sympatric divergence of risk of Comparative Psychology, 37, 81–95. http://dx.doi.org/10.1037/ sensitivity and diet menus in three species of tit. Animal Behaviour, 84, h0062414 1001–1012. http://dx.doi.org/10.1016/j.anbehav.2012.07.026 Herrmann, E., Call, J., Herna`ndez-Lloreda, M. V., Hare, B., & Tomasello, King, J. E., & Witt, E. D. (1966). The learning of patterned strings M. (2007). Humans have evolved specialized skills of social cognition: problems by rock squirrels. Psychonomic Science, 4, 319–320. http://dx The cultural intelligence hypothesis. Science, 317, 1360–1366. http://dx .doi.org/10.3758/BF03342316 .doi.org/10.1126/science.1146282 Kinnaman, A. (1902). Mental life of two Macacus rhesus monkeys in Herrmann, E., Hare, B., Call, J., & Tomasello, M. (2010). Differences in captivity. I. The American Journal of Psychology, 13, 98–148. http://dx the cognitive skills of bonobos and chimpanzees. PLoS ONE, 5, e12438. .doi.org/10.2307/1412207 http://dx.doi.org/10.1371/journal.pone.0012438 Klüver, H. (1937). Re-examination of implement-using behavior in a cebus Herrmann, E., Wobber, V., & Call, J. (2008). Great apes’ (Pan troglodytes, monkey after an interval of three years. Acta Psychologica, 2, 347–397. Pan paniscus, Gorilla gorilla, Pongo pygmaeus) understanding of tool http://dx.doi.org/10.1016/S0001-6918(37)90014-3 functional properties after limited experience. Journal of Comparative Klüver, H. (1961). Behavior mechanisms in monkeys. Chicago, IL: Uni- Psychology, 122, 220–230. http://dx.doi.org/10.1037/0735-7036.122.2 versity Chicago Press. .220 Klüver, H., & Bucy, P. C. (1938). An analysis of certain effects of bilateral Herter, W. R. (1940). U¨ ber das “Putten” einiger Meisen-Arten [On the temporal lobectomy in the rhesus monkey, with special reference to “pulling” of some parids]. Ornithologische Monatsberichte, 48, 104– “psychic blindness”. The Journal of Psychology: Interdisciplinary and 109. Applied, 5, 33–54. http://dx.doi.org/10.1080/00223980.1938.9917551 Hertz, M. (1926). Beobachtungen an gefangenen Rabenvögeln [Observa- Klüver, H., & Bucy, P. C. (1939). Preliminary analysis of functions of the tions on captive corvids]. Psychologische Forschung, 8, 336–397. http:// temporal lobes in monkeys. Archives of Neurology & Psychiatry, 42, 979– dx.doi.org/10.1007/BF02411525 1000. http://dx.doi.org/10.1001/archneurpsyc.1939.02270240017001 Hertz, M. (1937). La rapport de l’instinct et de l’intelligence dans le règne Köhler, W. (1917/1927). The mentality of apes. London, England: Rout- animal [The report on instinct and intelligence in the animal kingdom]. ledge & Kegan Paul. http://dx.doi.org/10.1037/11338-000 Journal de Psychologie Normale et Pathologique, 134, 324–340. Kolb, B., Cioe, J., & Comeau, W. (2008). Contrasting effects of motor and Heubel, F. (1939). Beobachtungen und Versuche über das Sinnesleben und visual spatial learning tasks on dendritic arborization and spine density die Intelligenz eines Macropus gigantus Zimmermann [Observations and in rats. Neurobiology of Learning and Memory, 90, 295–300. http://dx experiments on the life sense and intelligence of a Macropus gigantus .doi.org/10.1016/j.nlm.2008.04.012 Zimmermann]. Bijdragen tot de Dierkunde, 27, 417–436. Kounios, J., & Beeman, M. (2009). The aha! moment: The cognitive Heubel, F. (1940). Beobachtungen und Versuche über das Sinnesleben und neuroscience of insight. Current Directions in Psychological Science, 18, 210–216. http://dx.doi.org/10.1111/j.1467-8721.2009.01638.x die Intelligenz bei einem Palmenroller (Arctogalidia Stigmatica) [Ob- Krasheninnikova, A. (2013). Patterned-string tasks: Relation between fine servations and experiments on the life sense and intelligence of a palm motor skills and visual-spatial abilities in parrots. PLoS ONE, 8, e85499. civet (Arctogalidia stigmatica)]. Archives Neerlandaises de Zoologie, 4, http://dx.doi.org/10.1371/journal.pone.0085499 369–400. http://dx.doi.org/10.1163/036551640X00163 Krasheninnikova, A. (2014). Physical cognition in parrots: A comparative This document is copyrighted by the American Psychological Association or one of its alliedHobhouse, publishers. L. T. (1915). Mind in evolution. London, England: Macmillan. approach (Doctoral dissertation). University of Hamburg, Hamburg, This article is intended solely for the personal use ofHolekamp, the individual user and is not to be disseminated broadly. K. E., Swanson, E. M., & Van Meter, P. E. (2013). Develop- Germany. mental constraints on behavioural flexibility. Philosophical Transac- Krasheninnikova, A., Bovet, D., Busse, U., & Péron, F. (2012, June). The tions of the Royal Society of London Series B: Biological Sciences, 368, parrot and the string: Means-end understanding in twelve psittacid bird 20120350. http://dx.doi.org/10.1098/rstb.2012.0350 species. Poster session presented at the ASAB Interdisciplinary Work- Holmberg, L. (1957). Fiskande kråkor [Fishing crows]. Fauna och Flora, shop 2012: Physical Cognition & Problem Solving, Birmingham, Eng- 52, 182–185. land. Huber, L., & Gajdon, G. K. (2006). Technical intelligence in animals: The Krasheninnikova, A., Bräger, S., & Wanker, R. (2013). Means-end com- kea model. Animal Cognition, 9, 295–305. http://dx.doi.org/10.1007/ prehension in four parrot species: Explained by social complexity. s10071-006-0033-8 Animal Cognition, 16, 755–764. http://dx.doi.org/10.1007/s10071-013- Humle, T., & Snowdon, C. T. (2008). Socially biased learning in the 0609-z acquisition of a complex foraging task in juvenile cottontop tamarins, Krasheninnikova, A., & Schneider, J. M. (2014). Testing problem-solving Saguinus Oedipus. Animal Behaviour, 75, 267–277. http://dx.doi.org/ capacities: Differences between individual testing and social group 10.1016/j.anbehav.2007.05.021 setting. Animal Cognition, 17, 1227–1232. http://dx.doi.org/10.1007/ Inglet, F. (1967). Rook pulling up bone on a string. British Birds, 60, 371. s10071-014-0744-1 STRING-PULLING PARADIGM 117

Krasheninnikova, A., & Wanker, R. (2010). String-pulling in spectacled Michels, K. M., Pustek, J. J., Jr., & Johnson, J. I., Jr. (1961). The solution parrotlets (Forpus conspicillatus). Behaviour, 147, 725–739. http://dx of patterned-strings problems by raccoons. Journal of Comparative and .doi.org/10.1163/000579510X491072 Physiological Psychology, 54, 439–441. http://dx.doi.org/10.1037/ Kruper, D. C., Patton, R. A., & Koskoff, Y. D. (1971). Visual discrimi- h0043615 nation in hemicerebrectomized monkeys. Physiology & Behavior, 7, Miklósi, A., Kubinyi, E., Topál, J., Gácsi, M., Virányi, Z., & Csányi, V. 173–179. http://dx.doi.org/10.1016/0031-9384(71)90279-4 (2003). A simple reason for a big difference: Wolves do not look back Laidler, K. (1978). Aspects of physical and cognitive development in the at humans, but dogs do. Current Biology, 13, 763–766. http://dx.doi.org/ infant orang-utan (Pongo pygmaeus) during the first fifteen months of 10.1016/S0960-9822(03)00263-X life (Doctoral dissertation). Durham University, Durham, England. Miles, H. L. W. (1990). The cognitive foundations for reference in a Laidre, M. E. (2008). Spontaneous performance of wild baboons on three signing orangutan. In S. T. Parker & K. R. Gibson (Eds.), Language and novel food-access puzzles. Animal Cognition, 11, 223–230. http://dx.doi intelligence in monkeys and apes (pp. 511–539). Cambridge: Cambridge .org/10.1007/s10071-007-0104-5 University Press. http://dx.doi.org/10.1017/CBO9780511665486.021 Larsson, E. (1958). Fiskande kråkor och korpar [Fishing crows and ra- Millikan, G. C., & Bowman, R. I. (1967). Observations on Galápagos vens]. Fauna och Flora, 53, 92–94. tool-using finches in captivity. Living Bird, 6, 23–41. Lashley, K. S. (1948). The mechanism of vision: XVIII. Effects of de- Mulcahy, N. J., & Schubiger, M. N. (2014). Can orangutans (Pongo abelii) stroying the visual associative areas of the monkey. Genetic Psychology infer tool functionality? Animal Cognition, 17, 657–669. http://dx.doi Monographs, 37, 107–166. .org/10.1007/s10071-013-0697-9 Lefebvre, L., Nicolakakis, N., & Boire, D. (2002). Tools and brains in Müller, C. A., Riemer, S., Virányi, Z., Huber, L., & Range, F. (2014). Dogs birds. Behaviour, 139, 939–973. http://dx.doi.org/10.1163/ learn to solve the support problem based on perceptual cues. Animal 156853902320387918 Cognition, 17, 1071–1080. http://dx.doi.org/10.1007/s10071-014- Loftin, H. (1959). Bird brains. The Science News-Letter, 76, 334. http:// 0739-y dx.doi.org/10.2307/3941252 Murray, R. (1882). Warne’s model housekeeper.London,England: MacLean, E. L., Matthews, L. J., Hare, B. A., Nunn, C. L., Anderson, Fredrick Warne. R. C., Aureli, F.,...Wobber, V. (2012). How does cognition evolve? Nellmann, H., & Trendelenburg, W. (1926). Ein Beitrag zur Intelligenz- prüfung niederer Affen [A contribution to intelligence examination in Phylogenetic comparative psychology. Animal Cognition, 15, 223–238. lower monkeys]. Zeitschrift fur Vergleichende Physiologie, 4, 142–200. http://dx.doi.org/10.1007/s10071-011-0448-8 Newton, I. (1967). The adaptive radiation and feeding ecology of some Magat, M., & Brown, C. (2009). Laterality enhances cognition in Austra- British finches. The Ibis, 109, 33–96. http://dx.doi.org/10.1111/j.1474- lian parrots. Proceedings of the Royal Society B: Biological Sciences, 919X.1967.tb00005.x 276, 4155–4162. http://dx.doi.org/10.1098/rspb.2009.1397 Nissani, M. (2004). Theory of mind and insight in chimpanzees, elephants Mason, W. A., Blazek, N. C., & Harlow, H. F. (1956). Learning capacities and other animals? In L. J. Rogers & G. Kaplan (Eds.), Comparative of the infant rhesus monkey. Journal of Comparative and Physiological vertebrate cognition: Are primates superior to non-primates? (pp. 227– Psychology, 49, 449–453. http://dx.doi.org/10.1037/h0040209 261). New York, NY: Kluwer Academic/Plenum Press. http://dx.doi Mason, W. A., & Harlow, H. F. (1961). The effects of age and previous .org/10.1007/978-1-4419-8913-0_7 training on patterned-strings performance of Rhesus monkeys. Journal Obozova, T. A., Bagotskaya, M. S., Smirnova, A. A., & Zorina, Z. A. of Comparative and Physiological Psychology, 54, 704–709. http://dx (2014). A comparative assessment of birds’ ability to solve string- .doi.org/10.1037/h0044686 pulling tasks. Biology Bulletin, 41, 565–574. http://dx.doi.org/ Mason, W. A., Harlow, H. F., & Reuping, R. R. (1959). The development 10.7868/S0044513413060093 of manipulatory responsiveness in the infant rhesus monkey. Journal of Obozova, T., & Zorina, Z. (2013). Do great grey owls comprehend means– Comparative and Physiological Psychology, 52, 555–558. http://dx.doi end relationships? International Journal of Comparative Psychology, .org/10.1037/h0041898 26, 187–201. Mathieu, M., Daudelin, N., Dagenais, Y., & Décarie, T. G. (1980). Piag- Osthaus, B., Lea, S. E. G., & Slater, A. M. (2005). Dogs (Canis lupus etian causality in two house-reared chimpanzees (Pan troglodytes). familiaris) fail to show understanding of means-end connections in a Canadian Journal of Psychology, 34, 179–186. http://dx.doi.org/ string-pulling task. Animal Cognition, 8, 37–47. http://dx.doi.org/ 10.1037/h0081039 10.1007/s10071-004-0230-2 Mayer, C., Call, J., Albiach-Serrano, A., Visalberghi, E., Sabbatini, G., & Osvath, M., Kabadayi, C., & Jacobs, I. F. (2014). Independent evolution of Seed, A. (2014). Abstract knowledge in the broken-string problem: similar complex cognitive skills: The importance of embodied degrees Evidence from nonhuman primates and pre-schoolers. PLoS ONE, 9, of freedom. Animal Behavior and Cognition, 1, 249–264. e108597. Parker, S. T., & McKinney, M. L. (1999). Origins on intelligence: The This document is copyrighted by the American Psychological Association or one of its allied publishers. McCulloch, T. L. (1934). Performance preferentials of the white rat in evolution of cognitive development in monkeys, apes, and humans. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. force-resisting and spatial dimensions. Journal of Comparative Psychol- London, England: The John Hopkins University Press. ogy, 18, 85–111. http://dx.doi.org/10.1037/h0074986 Pepperberg, I. M. (2002). The value of the Piagetian framework for McCulloch, T. L., & Pratt, J. (1934). A study of the pre-solution period in comparative cognitive studies. Animal Cognition, 5, 177–182. http://dx weight discrimination by white rats. Journal of Comparative Psychol- .doi.org/10.1007/s10071-002-0148-5 ogy, 18, 271–290. http://dx.doi.org/10.1037/h0075422 Pepperberg, I. M. (2004). “Insightful” string-pulling in Grey parrots (Psit- McDougall, K. D., & McDougall, W. (1931). Insight and foresight in tacus erithacus) is affected by vocal competence. Animal Cognition, 7, various animals—Monkey, raccoons, rat, and wasp. Journal of Compar- 263–266. http://dx.doi.org/10.1007/s10071-004-0218-y ative Psychology, 11, 237–273. http://dx.doi.org/10.1037/h0073941 Pfuhl, G. (2012). Two strings to choose from: Do ravens pull the easier McDougall, W., & McDougall, K. D. (1927). Notes on instinct and one? Animal Cognition, 15, 549–557. http://dx.doi.org/10.1007/s10071- intelligence in rats and cats. Journal of Comparative Psychology, 7, 012-0483-0 145–175. http://dx.doi.org/10.1037/h0073339 Pfuhl, G., Gattermayr, M., & Bugnyar, T. (2014). Will food-handling time Medina, F. S. M. (2012). Study of the cognition and its neural substrate in influence agonistic behaviour in sub-adult common ravens (Corvus New Caledonian crows (Doctoral dissertation). University of Auckland, corax)? Behavioural Processes, 103, 67–74. http://dx.doi.org/10.1016/j Auckland, New Zealand. .beproc.2013.11.003 118 JACOBS AND OSVATH

Piaget, J. (1952). The origins of intelligence in children. New York, NY: Sarris, E. G. (1937). Die individuellen Unterschiede bei Hunden [Individ- Norton. http://dx.doi.org/10.1037/11494-000 ual differences in dogs]. Zeitschrift für Angewandte Psychologie, 52, Plotnik, J. M., Lair, R., Suphachoksahakun, W., & de Waal, F. B. (2011). 257–309. Elephants know when they need a helping trunk in a cooperative task. Sasväri, L. (1985). Different observational learning capacity in juvenile Proceedings of the National Academy of Sciences of the United States of and adult individuals of congeneric bird species. Zeitschrift für Tierpsy- America, 108, 5116–5121. http://dx.doi.org/10.1073/pnas.1101765108 chologie, 69, 293–304. http://dx.doi.org/10.1111/j.1439-0310.1985 Potí, P. (1989). Early sensorimotor development in macaques. In F. Anti- .tb00154.x nucci (Ed.), Cognitive structure and development in nonhuman primates Schiestl, M. (2013). Individuality and foraging strategies in free ranging (pp. 39–54). Hillsdale, NJ: Erlbaum. crows (Corvus corone corone; C.c. cornix) (Master’s thesis). University Potì, P., & Spinozzi, G. (1994). Early sensorimotor development in chim- of Vienna, Vienna, Austria. panzees (Pan troglodytes). Journal of Comparative Psychology, 108, Schmid, B. (1936). Zur Psychologie der Caniden [On the psychology of 93–103. http://dx.doi.org/10.1037/0735-7036.108.1.93 canids]. Zentralblad für Kleintierkunde und Pelztierkunde, 12, 1–77. Povinelli, D. J. (2000). Folk physics for apes: The chimpanzee’s theory of Schmidt, G. F., & Cook, R. G. (2006). Mind the gap: Means-end discrim- how the world works. New York, NY: Oxford University Press. ination by pigeons. Animal Behaviour, 71, 599–608. http://dx.doi.org/ Povinelli, D. J. (2012). World without weight: Perspectives on an alien 10.1016/j.anbehav.2005.06.010 mind. Oxford: Oxford University Press. Schmitt, V., Pankau, B., & Fischer, J. (2012). Old world monkeys compare Pratt, J. G. (1938). An experimental analysis of the process of solving a to apes in the primate cognition test battery. PLoS ONE, 7, e32024. weight discrimination problem in white rats. Journal of Comparative http://dx.doi.org/10.1371/journal.pone.0032024 Psychology, 25, 291–314. http://dx.doi.org/10.1037/h0054318 Schrauf, C., & Call, J. (2011). Great apes use weight as a cue to find hidden Prescott, M. J., & Buchanan-Smith, H. M. (1999). Intra-and inter-specific food. American Journal of Primatology, 73, 323–334. http://dx.doi.org/ social learning of a novel food task in two species of tamarin. Interna- 10.1002/ajp.20899 tional Journal of Comparative Psychology, 12, 71–92. Schuck-Paim, C., Borsari, A., & Ottoni, E. B. (2009). Means to an end: Rackham, H. (1947). Natural history by pliny the elder. Cambridge: Neotropical parrots manage to pull strings to meet their goals. Animal Harvard University Press. Cognition, 12, 287–301. http://dx.doi.org/10.1007/s10071-008-0190-z Raisler, R. L., & Harlow, H. F. (1965). Learned behavior following lesions Scott, J. D. (1974). Woe to the farmer’s foe, the crow. National Wildlife, 12, 44–47. of posterior association cortex in infant, immature, and preadolescent Seed, A. M., & Boogert, N. J. (2013). Animal cognition: An end to insight? monkeys. Journal of Comparative and Physiological Psychology, 60, Current Biology, 23, R67–R69. http://dx.doi.org/10.1016/j.cub.2012.11 167–174. http://dx.doi.org/10.1037/h0022294 .043 Range, F., Möslinger, H., & Virányi, Z. (2012). Domestication has not Seed, A. M., Hanus, D., & Call, J. (2011). Causal knowledge in corvids, affected the understanding of means-end connections in dogs. Animal primates and children: More than meets the eye? In T. McCormack, C. Cognition, 15, 597–607. http://dx.doi.org/10.1007/s10071-012-0488-8 Hoerl, & S. Butterfill (Eds.), Tool use and causal cognition (pp. 89– Rat-Fischer, L., O’Regan, J. K., & Fagard, J. (2014). Comparison of active 110). Oxford, England: Oxford University Press. http://dx.doi.org/ and purely visual performance in a multiple-string means-end task in 10.1093/acprof:oso/9780199571154.003.0005 infants. Cognition, 133, 304–316. http://dx.doi.org/10.1016/j.cognition Seed, A., Seddon, E., Greene, B., & Call, J. (2012). Chimpanzee ‘folk .2014.06.005 physics’: Bringing failures into focus. Philosophical Transactions of the Ray, J. (1678). The ornithology of Francis Willughby. London, England: Royal Society of London Series B: Biological Sciences, 367, 2743–2752. John Martyn. http://dx.doi.org/10.1098/rstb.2012.0222 Reader, S. M., Hager, Y., & Laland, K. N. (2011). The evolution of primate Seibt, U., & Wickler, W. (2006). Individuality in problem solving: String general and cultural intelligence. Philosophical Transactions of the pulling in two Carduelis species (Aves: Passeriformes). Ethology, 112, Royal Society of London Series B: Biological Sciences, 366, 1017–1027. 493–502. http://dx.doi.org/10.1111/j.1439-0310.2005.01172.x http://dx.doi.org/10.1098/rstb.2010.0342 Settlage, P. H. (1939). The effect of occipital lesions on visually guided Redshaw, M. (1978). Cognitive developments in human and gorilla infants. behavior in the monkey. I. Influence of the lesions on final capacities in Journal of Human Evolution, 7, 133–141. http://dx.doi.org/10.1016/ a variety of problem solving. Journal of Comparative Psychology, 27, S0047-2484(78)80005-0 93–131. http://dx.doi.org/10.1037/h0053952 Richards, M. L. (1973). Rook pulling up food on a string. British Birds, 66, Shepherd, W. (1910). Some mental processes of the rhesus monkey. The 365–366. Psychological Monographs, 12, i–61. http://dx.doi.org/10.1037/ Richardson, H. M. (1932). The growth of adaptive behavior in infants: An h0093038 This document is copyrighted by the American Psychological Association or one of its allied publishers. experimental study at seven age levels. Genetic Psychology Mono- Shepherd, W. (1915). Tests on adaptive intelligence on in dogs and cats, This article is intended solely for the personal use of the individual usergraphs, and is not to be disseminated broadly. 12, 195–359. compared with adaptive intelligence in rhesus monkeys. The American Riemer, S., Müller, C., Range, F., & Huber, L. (2014). Dogs (Canis Journal of Psychology, 26, 211–216. http://dx.doi.org/10.2307/1413250 familiaris) can learn to attend to connectivity in string pulling tasks. Shettleworth, S. J. (2009a). Animal cognition: Deconstructing avian in- Journal of Comparative Psychology, 128, 31–39. sight. Current Biology, 19, R1039–R1040. http://dx.doi.org/10.1016/j Riesen, A. H., Greenberg, B., Granston, A. S., & Fantz, R. L. (1953). .cub.2009.10.022 Solutions of patterned string problems by young gorillas. Journal of Shettleworth, S. J. (2009b). The evolution of comparative cognition: Is the Comparative and Physiological Psychology, 46, 19–22. http://dx.doi snark still a boojum? Behavioural Processes, 80, 210–217. http://dx.doi .org/10.1037/h0057069 .org/10.1016/j.beproc.2008.09.001 Rijk, J. C. (1934). De Koolmees putter [The great tit puller]. De Levende Shettleworth, S. J. (2012). Do animals have insight, and what is insight Natuur, 38, 325. anyway? Canadian Journal of Experimental Psychology, 66, 217–226. Riopelle, A. J., Alper, R. G., Strong, P. N., & Ades, H. W. (1953). Multiple http://dx.doi.org/10.1037/a0030674 discrimination and patterned string performance of normal and Shumaker, R. W., Walkup, K. R., & Beck, B. B. (2011). Animal tool temporal-lobectomized monkeys. Journal of Comparative and Physio- behavior: The use and manufacture of tools by animals. Baltimore, MD: logical Psychology, 46, 145–149. http://dx.doi.org/10.1037/h0058658 JHU Press. STRING-PULLING PARADIGM 119

Silva, F. J., Page, D. M., & Silva, K. M. (2005). Methodological- configural discriminations. Physiology & Behavior, 48, 225–231. http:// conceptual problems in the study of chimpanzees’ folk physics: How dx.doi.org/10.1016/0031-9384(90)90305-N studies with adult humans can help. Learning & Behavior, 33, 47–58. Torigoe, T. (1987). Further report on object manipulation in non-human http://dx.doi.org/10.3758/BF03196049 primates: A comparison within 13 species of the genus Macaca. Pri- Silva, F. J., Silva, K. M., Cover, K. R., Leslie, A. L., & Rubalcaba, M. A. mates, 28, 533–538. http://dx.doi.org/10.1007/BF02380867 (2008). Humans’ folk physics is sensitive to physical connection and Trueblood, C. K., & Smith, K. U. (1934). String-pulling behavior of the contact between a tool and reward. Behavioural Processes, 77, 327–333. cat. The Pedagogical Seminary and Journal of Genetic Psychology, 44, http://dx.doi.org/10.1016/j.beproc.2007.08.001 414–427. http://dx.doi.org/10.1080/08856559.1934.10533696 Singh, S. D. (1966). Effect of human environment on cognitive behavior in Ungerleider, L. G., & Brody, B. A. (1977). Extrapersonal spatial orienta- the rhesus monkey. Journal of Comparative and Physiological Psychol- tion: The role of posterior parietal, anterior frontal, and inferotemporal ogy, 61, 280–283. http://dx.doi.org/10.1037/h0023136 cortex. Experimental Neurology, 56, 265–280. http://dx.doi.org/ Smith, B. P., & Litchfield, C. A. (2013). Looking back at ‘looking back’: 10.1016/0014-4886(77)90346-6 Operationalising referential gaze for dingoes in an unsolvable task. Ungerleider, L. G., & Pribram, K. H. (1977). Inferotemporal versus com- Animal Cognition, 16, 961–971. http://dx.doi.org/10.1007/s10071-013- bined pulvinar-prestriate lesions in the rhesus monkey: Effects on color, 0629-8 object and pattern discrimination. Neuropsychologia, 15, 481–498. Spinozzi, G. (1989). Early sensormotor development in Cebus. In F. http://dx.doi.org/10.1016/0028-3932(77)90052-5 Antinucci (Ed.), Cognitive structure and development in nonhuman Van Lawick-Goodall, J. (1971). Tool-using in primates and other verte- primates (pp. 55–66). Hillsdale, NJ: Erlbaum. brates. Advances in the Study of Behavior, 3, 195–249. http://dx.doi.org/ Spinozzi, G., & Natale, F. (1989). Early sensorimotor development in 10.1016/S0065-3454(08)60157-6 Gorilla. In F. Antinucci (Ed.), Cognitive structure and development in Vauclair, J. (2012). Piaget and the comparative psychology of animal nonhuman primates (pp. 21–38). Hillsdale, NJ: Erlbaum. cognition. In E. Martí & C. Rodríguez (Eds.), After Piaget (pp. 59–70). St Amant, R., & Horton, T. E. (2008). Revisiting the definition of animal New Brunswick, NJ: Transaction. tool use. Animal Behaviour, 75, 1199–1208. http://dx.doi.org/10.1016/ Vince, M. A. (1956). String-pulling in birds. 1. Individual differences in j.anbehav.2007.09.028 wild adult great tits. The British Journal of Animal Behaviour, 4, Su, T-T. (1982). Effects of partially covering string arrays on patterned 111–116. http://dx.doi.org/10.1016/S0950-5601(56)80131-7 string performance of platyrrhine monkeys (Doctoral dissertation). Uni- Vince, M. A. (1958). String-pulling in birds. 2. Differences related to age versity of Arizona, Tucson, AZ. in greenfinches, chaffinches and canaries. Animal Behaviour, 6, 53–59. Sustaita, D., Pouydebat, E., Manzano, A., Abdala, V., Hertel, F., & Herrel, http://dx.doi.org/10.1016/0003-3472(58)90008-3 A. (2013). Getting a grip on tetrapod grasping: Form, function, and Vince, M. A. (1961). “String-pulling” in birds. 3. The successful response evolution. Biological Reviews of the Cambridge Philosophical Society, in greenfinches and canaries. Behaviour, 17, 103–129. http://dx.doi.org/ 88, 380–405. http://dx.doi.org/10.1111/brv.12010 10.1163/156853961X00033 Taylor, A. H., Elliffe, D., Hunt, G. R., & Gray, R. D. (2010). Complex Vince, M. A. (1964). Use of the feet in feeding by the Great Tit Parus cognition and behavioural innovation in New Caledonian crows. Pro- major. The Ibis, 106, 508–529. http://dx.doi.org/10.1111/j.1474-919X ceedings of the Royal Society B: Biological Sciences, 277, 2637–2643. .1964.tb03730.x http://dx.doi.org/10.1098/rspb.2010.0285 Völter, C. J., & Call, J. (2012). Problem solving in great apes (Pan Taylor, A. H., & Gray, R. D. (2009). Animal cognition: Aesop’s fable flies paniscus, Pan troglodytes, Gorilla gorilla, and Pongo abelii): The effect from fiction to fact. Current Biology, 19, R731–R732. http://dx.doi.org/ of visual feedback. Animal Cognition, 15, 923–936. http://dx.doi.org/ 10.1016/j.cub.2009.07.055 10.1007/s10071-012-0519-5 Taylor, A. H., Knaebe, B., & Gray, R. D. (2012). An end to insight? New Caledonian crows can spontaneously solve problems without planning von Bayern, A. M. P., Heathcote, R. J. P., Rutz, C., & Kacelnik, A. (2009). their actions. Proceedings of the Royal Society B: Biological Sciences, The role of experience in problem solving and innovative tool use in 279, 4977–4981. http://dx.doi.org/10.1098/rspb.2012.1998 crows. Current Biology, 19, 1965–1968. http://dx.doi.org/10.1016/j.cub Taylor, A. H., Medina, F. S., Holzhaider, J. C., Hearne, L. J., Hunt, G. R., .2009.10.037 & Gray, R. D. (2010). An investigation into the cognition behind Warden, C. J., Barrera, S., & Galt, W. (1942). The effect of unilateral and spontaneous string pulling in New Caledonian crows. PLoS ONE, 5, bilateral frontal lobe extirpation on the behavior of monkeys. Journal of e9345. http://dx.doi.org/10.1371/journal.pone.0009345 Comparative Psychology, 34, 149–171. http://dx.doi.org/10.1037/ Teyrovský, V. (1930). A study of ideational behavior of a garden warbler, h0062212 Sylvia borin (Bodd.). Publ De la Faculté des Sciences de l’Université Warden, C. J., Koch, A. M., & Fjeld, H. A. (1940). Solution of patterned This document is copyrighted by the American Psychological Association or one of its allied publishers. Masaryk, 122, 1–36. string problems by monkeys. The Journal of Genetic Psychology, 56, This article is intended solely for the personal use ofThienemann, the individual user and is not to be disseminated broadly. J. (1933). Der Stieglitz als Schöpfer [The goldfinch as a 283–295. puller]. Ornithologische Monatsberichte, 41, 92–93. Warren, J. M., Leary, R. W., Harlow, H. F., & French, G. M. (1957). Thorpe, W. H. (1943). A type of insight learning in birds. British Birds, 37, Function of association cortex in monkeys. The British Journal of 29–31. Animal Behaviour, 5, 131–138. http://dx.doi.org/10.1016/S0950- Thorpe, W. H. (1944). Further notes on a type of insight learning in birds. 5601(57)80019-7 British Birds, 38, 46–49. Washington, D. (1974). Rooks feeding on suspended fat. British Birds, 67, Thorpe, W. H. (1959). Learning. The Ibis, 101, 337–353. 213–214. Thorpe, W. H. (1963). Learning and instinct in animals. London, England: Wasserman, E. A., Nagasaka, Y., Castro, L., & Brzykcy, S. J. (2013). Methuen. Pigeons learn virtual patterned-string problems in a computerized touch Tolman, E. C. (1937). The acquisition of string-pulling by rats– screen environment. Animal Cognition, 16, 737–753. http://dx.doi.org/ conditioned response or signgestalt? Psychological Review, 44, 195– 10.1007/s10071-013-0608-0 211. http://dx.doi.org/10.1037/h0059293 Weiner, R., & Harlow, H. F. (1952). The effect of nembutal upon learned Tomie, J. A., & Whishaw, I. Q. (1990). New paradigms for tactile dis- performances of the rhesus monkey. The Journal of General Psychol- crimination studies with the rat: Methods for simple, conditional, and ogy, 46, 43–50. http://dx.doi.org/10.1080/00221309.1952.9710636 120 JACOBS AND OSVATH

Werdenich, D., & Huber, L. (2006). A case of quick problem solving in in monkeys. Journal of Comparative and Physiological Psychology, 52, birds: String pulling in keas, Nestor notabilis. Animal Behaviour, 71, 10–17. http://dx.doi.org/10.1037/h0045542 855–863. http://dx.doi.org/10.1016/j.anbehav.2005.06.018 Wobber, V., & Hare, B. (2011). Psychological health of orphan bonobos Whishaw, I. Q., & Tomie, J.-A. (1991). Acquisition and retention by and chimpanzees in African sanctuaries. PLoS ONE, 6, e17147. http:// hippocampal rats of simple, conditional, and configural tasks using dx.doi.org/10.1371/journal.pone.0017147 tactile and olfactory cues: Implications for hippocampal function. Be- Yagi, B. (1964). Solutions of patterned string problems by Japanese mon- havioral Neuroscience, 105, 787–797. http://dx.doi.org/10.1037/0735- keys (Macaca fuscata Yakui). The Annual of Animal Psychology, 14, 7044.105.6.787 73–81. http://dx.doi.org/10.2502/janip1944.14.73 Whishaw, I. Q., & Tomie, J.-A. (1995). Rats with fimbria-fornix lesions Yerkes, R. M. (1927). The mind of a gorilla. Genetic Psychology Mono- can acquire and retain a visual-tactile transwitching (configural) task. graphs, 2, 1–193. Behavioral Neuroscience, 109, 607–612. http://dx.doi.org/10.1037/ Zorn, J. H. (1743). Petino-Theologie oder Versuch, Die Menschen durch 0735-7044.109.4.607 nähere Betrachtung Der Vögel Zur Bewunderung, Liebe und Verehrung Whishaw, I. Q., Tomie, J., & Kolb, B. (1992). Ventrolateral prefrontal ihres mächtigsten, weissest- und gütigsten Schöpffers aufzumuntern cortex lesions in rats impair the acquisition and retention of a tactile- olfactory configural task. Behavioral Neuroscience, 106, 597–603. [Petino-theology or the attempt to encourage mankind to admiration, http://dx.doi.org/10.1037/0735-7044.106.4.597 love and worship of their most powerful, wisest and kindest creator Whitt, E., Douglas, M., Osthaus, B., & Hocking, I. (2009). Domestic cats through the closer examination of birds]. Schwabach, Germany: Rau. (Felis catus) do not show causal understanding in a string-pulling task. Animal Cognition, 12, 739–743. http://dx.doi.org/10.1007/s10071-009- 0228-x Received January 12, 2014 Wilson, W. A., Jr., & Mishkin, M. (1959). Comparison of the effects of Revision received December 1, 2014 inferotemporal and lateral occipital lesions on visually guided behavior Accepted December 5, 2014 Ⅲ

E-Mail Notification of Your Latest Issue Online!

Would you like to know when the next issue of your favorite APA journal will be available online? This service is now available to you. Sign up at http://notify.apa.org/ and you will be notified by e-mail when issues of interest to you become available! 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.