Date of Award

Summer 8-1-2012

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Psychology

First Advisor

Michael J. Owren, PhD

Second Advisor

Rose A. Sevcik, PhD

Third Advisor

Gwen Frishkoff, PhD

Fourth Advisor

Michael J. Beran, PhD

Abstract

There has been much discussion regarding whether the capability to perceive speech is uniquely human. The “Speech is Special” (SiS) view proposes that humans possess a specialized cognitive module for speech perception (Mann & Liberman, 1983). In contrast, the “Auditory Hypothesis” (Kuhl, 1988) suggests spoken-language evolution took advantage of existing auditory-system capabilities. In support of the Auditory Hypothesis, there is evidence that Panzee, a language-trained chimpanzee (Pan troglodytes), perceives speech in synthetic “sine-wave” and “noise-vocoded” forms (Heimbauer, Beran, & Owren, 2011). Human comprehension of these altered forms of speech has been cited as evidence for specialized cognitive capabilities (Davis, Johnsrude, Hervais-Adelman, Taylor, & McGettigan, 2005).

In light of Panzee’s demonstrated abilities, three experiments extended these investigations of the cognitive processes underlying her speech perception. The first experiment investigated the acoustic cues that Panzee and humans use when identifying sine-wave and noise-vocoded speech. The second experiment examined Panzee’s ability to perceive “time-reversed” speech, in which individual segments of the waveform are reversed in time. Humans are able to perceive such speech if these segments do not much exceed average phoneme length. Finally, the third experiment tested Panzee’s ability to generalize across both familiar and novel talkers, a perceptually challenging task that humans seem to perform effortlessly.

Panzee’s performance was similar to that of humans in all experiments. In Experiment 1, results demonstrated that Panzee likely attends to the same “spectro-temporal” cues in sine-wave and noise-vocoded speech that humans are sensitive to. In Experiment 2, Panzee showed a similar intelligibility pattern as a function of reversal-window length as found in human listeners. In Experiment 3, Panzee readily recognized words not only from a variety of familiar adult males and females, but also from unfamiliar adults and children of both sexes. Overall, results suggest that a combination of general auditory processing and sufficient exposure to meaningful spoken language is sufficient to account for speech-perception evidence previously proposed to require specialized, uniquely human mechanisms. These findings in turn suggest that speech-perception capabilities were already present in latent form in the common evolutionary ancestors of modern chimpanzees and humans.

Share

COinS