they come from the sky or the ground (Manser 2001). Vervet monkeys also use different alarm calls for different predators. Research has shown they do not react blindly to the alarm calls of others. When a caller proves to be unreliable—for example, because the caller in question is in fact a recording played back by researchers who want to test responses to different predators—the vervet monkeys stop responding. This demonstrates that they can judge the meaning of the call (Seyfarth et al. 1980). Many species also understand and imitate the alarm calls of other species. Campbell’s monkey alarm calls have syntax; the elements hang together as words in a sentence (Zuberbühler 2001). Diana guenon alarm calls do not have this, but they do understand the meaning of the Campbell’s monkeys’ calls (ibid.). Campbell’s monkeys also use different sounds resembling words in different areas of the world (ibid.). I have already discussed the ability of parrots to imitate others, for example, to scare them away and steal their food. Fork-tailed drongos also use mimicry and can imitate the alarm calls of fifty other species. They use this skill to warn other animals and to steal their food (Flower 2011).
Alarm calls are often accompanied by, or consist solely of, visual signals, such as facial expressions, body movements, and gestures. Smell also plays an important role in the alarm calls of many species. As well as sounds, snails use smells in their slime when they are in danger (Breure 2015). African bees use scents to alarm others and to summon the whole swarm for an attack (Slobodchikoff 2012). They can and sometimes do kill humans in this way when they feel threatened. Research into the role of pheromones and smell in animal communication is still in its early stages, but we do know that some smells used as alarm calls consist of different elements, and that the combination of smells, as well as the ratio, provides meaning. Californian thrips insects, for example, use different alarm pheromones for different threats (De Bruijn 2015). Thrips larvae produce a pheromone that consists of two ingredients: decyl acetate and dodecyl acetate. When the danger intensifies, the quantity and ratio of pheromones produced changes. Larvae who receive the signal change their behavior accordingly, so meaning is transferred adequately.
Alarm calls were long thought to be simple instinctive responses to danger, a pre-scripted form of communication, wired into the genetic makeup of an animal. This refers back to a view of non-human animals who act solely on instinct (see chapter 1 and the previous section) and who are not capable of responding intelligently. As ethological research in which other animals are studied in their own habitats progresses, along with technological developments,11 it has been found that in many species, like the prairie dogs, alarm calls are actually very complex. They should be seen as expressions of an individual animal’s intelligence rather than as simple mechanistic reactions: as language, rather than as communication. A single call from a prairie dog shares a large amount of information with others in a much more precise and efficient way than a human scream or word. The language of the prairie dog seems to have a similar structure to human language, including grammar (Slobodchikoff et al. 2009); it might also have functions we cannot yet understand because we do not recognize or perceive them. We cannot hear the complexities in their calls—to us they all sound more or less the same—so it is logical that humans formerly perceived their language as simple calls. We are only now beginning to find out what they are saying thanks to the use of technology. This example shows us that there is more to animal calls than we think, and that it is important to move beyond a view of animals as acting solely on instinct and only using signals with a fixed meaning when we study alarm calls and animal languages more generally.
Viewing alarm calls as a set of language games in which non-human animals create meaning in ways that are sometimes similar to humans, and sometimes very different, provides us with a new way of studying their languages. This begins with recognizing them as subjects, rather than as objects who simply follow their instincts (see also chapter 3). Studying their languages is closely interconnected with studying their social relations, and in order to map and interpret these language games, we also need to study the practices and relations in which they gain meaning. Understanding the context helps us to understand the meaning of signals while getting a better grasp on the structures of non-human animal languages; for example, the grammar in prairie dog alarm calls can help us to better interpret their behavior: both are needed to gain an insight into their languages.
Grammar
The structures, or grammars, of most animal languages have not been studied in detail. This relates back to the fact that while animal languages are studied in biology and ethology, in most of these studies human language is taken as the blueprint for what language—as opposed to communication—is (Slobodchikoff 2012), which precludes many non-human animal expressions. The fact that other animals do not use human language cannot, however, lead us to conclude that they do not have language. If we do not understand their expressions, we cannot conclude that they are not complex or meaningful—different does not automatically equal less. Even if we cannot use human language as a blueprint for what language is, concepts used in studying human language can function as tools for understanding the languages of other animals, even those very different from us.
An example of this can be found in research into the language of Caribbean reef squid (Moynihan 1991). Caribbean reef squid speak with their skin. Pigment cells in their skin called chromatophores are attached to muscles that can be contracted or relaxed. This either exposes the pigment or makes it invisible, which allows the squid to change the color patterns on their skin very rapidly. In doing so they create complex visual patterns ranging from white to camouflage, which send sophisticated signals to other squid. Males, for example, use specific color patterns to flirt with females, who in turn use other patterns to respond. When there are other males around, the male can use half of his body to signal to the female, and the other half to tell his opponent to back off. Many different patterns are possible, and they can change in the blink of an eye. Because of the subtlety and speed of the changes, most of the patterns have not yet been deciphered. In addition to the color patterns, squid also use body postures to create meaning. Biologist Moynihan (1991) argues that the visual patterns, together with the postures, constitute a proper language, with nouns, verbs, adjectives, and adverbs. Nouns and verbs are most important; they are, for example, used to establish whether the other wants to mate. Adjectives and adverbs are then used to describe the intensity of the desire to mate. Because of the difficulty in understanding all the elements of this visual language, the precise rules of this grammar are still unknown, but the broad meaning can be grasped from the context. Studying these expressions as a language with a grammar helps us to get a better grasp of how they function; seeing their grammar as grammar can also help us to get a richer view of what grammar can entail.
Another example is bird song. The songs and calls of many species of birds have been studied extensively, and it was long assumed that the most important functions of song were to attract females and to defend territory. However, recent research focusing on grammar shows that bird languages are far more complex than was supposed, both with regard to the content of messages and their structure (Slobodchikoff 2012). Many aspects once thought to be unique features of human language have also been found in bird languages. Humans can, for example, produce new sounds that have meaning for speakers of the same language because they follow certain syntactic rules. A recursive, hierarchical embedding, needed for new utterances to make sense, requires a context-free grammar. Recursion has been found in the language of many birds (and other non-human animal species, such as elephants), including chickadees (Kerschenbaum et al. 2014). Chickadee language consists of a variety of different sounds that can be combined creatively. Single units are combined, like words, into patterns and sentences, of which the combinations become more complex as the intensity of the communication increases (Slobodchikoff 2012). For the Carolina chickadee, the meaning of sentences changes when the order of the elements changes (Kerschenbaum et al. 2014). Starlings have recently been found to classify sentences from embedded, context-free grammar (Gentner et al. 2006). The syntax of the black-chinned hummingbird also has an open system similar to that of starlings and humans (Slobodchikoff 2012).
Grammar is also being studied in the languages of other animals. The songs of humpback whales (Suzuki et al. 2006), which sound improvised and chaotic to the human ear, are formed like sentences, consisting of smaller units that are combined to form songs containing up to 400 elements. Mexican free-tailed bats are