Insights From Animals

Peter Marler
"Origins of Music and Speech: Insights from Animals"
in Wallin, Merker and Brown, eds.,
The Origins of Music (2000)

[32] Some fifteen years or so ago, the thinking of zoologists about the semantics of calls of animals, especially the vocalizations of monkeys and apes, underwent something of a revolution. Not long ago, speculations about how best to interpret animal calls were all based on what Donald Griffin (1992) aptly described as the "groans of pain" (GOP) concept of animal communication. This approach assumed that vocalizations of monkeys and other animals are displays of emotion or affect, much like our own facial expressions. Only humans are thought to have progressed beyond this condition and to have achieved symbolic signaling. Premack (1975) stated the prevailing view clearly and succinctly: "Man has both affective and symbolic communication. All other species, except when tutored by man, have only the affective form." Symbolic signals are taken to be those that have identifiable referents that the signal can be said to connote in an abstract, noniconic fashion. For an animal communication system to qualify as symbolic, information about one or more referents has to be both encoded noniconically by signalers and decoded in equivalent form by receivers.

Note that this is not a discussion about whether animal signals are meaningful or meaningless. Both affective and symbolic animal signals are meaningful and are often rich in information content; both serve important and diverse functions, some communicative to other individuals, some with repercussions for the physiological and mental states of the signaler. At issue here is not the presence of meaning but the kind of meaning that affective and symbolic signals convey. This is a complex subject with many dimensions. Some view the contrasts as differences in degree rather than kind. In some circumstances signals traditionally thought of as affective, such as human facial expressions, can assume a symbolic function. Complex signals may contain within them intimately blended components in which the balance between affective and symbolic content can vary dramatically from one the another. Speech is an obvious case. Anonymous computerized speech, lacking individual iden-
[33]
tity, gender, and emotion, is a sadly impoverished vehicle for social communication. We must not fall into the trap of assuming that signal systems that are not languagelike are necessarily impoverished as vehicles for social communication.

If you say so. But to apply the encoding-decoding criterion outside the realm of language seems merely to establish quite the impoverished conception of social communication.

p. 36:
Phonological Syntax

Recombinations of sound components (e.g. phonemes) in different sequences (e.g. words), where the components themselves are not meaningful. I call this "phonocoding."

Lexical Syntax

Recombinations of component sequences (e.g. words in the lexicon) into different strings (sentences). Here there is meaning at two levels, the word and the sentence. The meaning of the string is a product of the assembled meanings of its components. I call this "lexicoding."

Peter Cariani
"Life’s journey through the semiosphere"

Hoffmeyer follows the contemporary tendency to think of sensory systems as being highly specialized for particular ecological niches, and hence to interpret Uexküll’s umwelts as mostly incommensurable perspectives. However, despite the spectacular adaptations that are sometimes observed (e.g. the sonar of the bat), these particular sensory enhancements are invariably built out of ancient body-designs that have been conserved over huge phylogenetic spans. The same evolutionary conservatism may hold for the neural coding strategies that are used in representing and processing sensory information. While the particular experiential textures of things, their qualia, undoubtedly vary across different vertebrates, the basic body-plans, sensory organs and neural representations are roughly similar. We may see in different colors, hear in different frequency registers, and smell different odors, but the basic relational organizations of our percept-spaces in the end may not be so radically different. Birdcalls are almost certainly interpreted by other birds in a manner that is very different from how we interpret them, but there is enough commonality to what we hear to enable us to imitate birdcalls well enough to fool the birds themselves. The same goes for birds listening to and imitating human speech. Such cross-species invariants are not possible without general-purpose sensory and effector mechanisms for both analyzing and producing wide ranges of sounds. It is thus possible for more generalist evolutionary solutions to prevail, especially in the realm of the senses, where appearances change rapidly. For predator and prey alike, one needs general purpose sensory systems that reliably recognize other animals under widely varying conditions.