Human speech and birdsong have numerous parallels. Both humans and songbirds learn their complex vocalizations early in life, exhibiting a strong dependence on hearing the adults they will imitate, as well as themselves as they practice, and a waning of this dependence as they mature. Innate predispositions for perceiving and learning the correct sounds exist in both groups, although more evidence of innate descriptions of species-specific signals exists in songbirds, where numerous species of vocal learners have been compared. Humans also share with songbirds an early phase of learning that is primarily perceptual, which then serves to guide later vocal production. Both humans and songbirds have evolved a complex hierarchy of specialized forebrain areas in which motor and auditory centers interact closely, and which control the lower vocal motor areas also found in nonlearners. In both these vocal learners, however, how auditory feedback of self is processed in these brain areas is surprisingly unclear. Finally, humans and songbirds have similar critical periods for vocal learning, with a much greater ability to learn early in life. In both groups, the capacity for late vocal learning may be decreased by the act of learning itself, as well as by biological factors such as the hormones of puberty. Although some features of birdsong and speech are clearly not analogous, such as the capacity of language for meaning, abstraction, and flexible associations, there are striking similarities in how sensory experience is internalized and used to shape vocal outputs, and how learning is enhanced during a critical period of development. Similar neural mechanisms may therefore be involved.

/pdf/birdsong-and-human-speech-common-themes-and-mechanisms-39gg3m47d4.pdf

BIRDSONG AND HUMAN SPEECH: Common Themes and Mechanisms

In this article the author proposes an episodic theory of spoken word representation, perception, and production. By most theories, idiosyncratic aspects of speech (voice details, ambient noise, etc.) are considered noise and are filtered in perception. However, episodic theories suggest that perceptual details are stored in memory and are integral to later perception. In this research the author tested an episodic model (MINERVA 2; D. L. Hintzman, 1986) against speech production data from a word-shadowing task. The model predicted the shadowing-response-time patterns, and it correctly predicted a tendency for shadowers to spontaneously imitate the acoustic patterns of words and nonwords. It also correctly predicted imitation strength as a function of "abstract" stimulus properties, such as word frequency. Taken together, the data and theory suggest that detailed episodes constitute the basic substrate of the mental lexicon. Early in the 20th century, Semon (1909/1923) described a memory theory that anticipated many aspects of contemporary theories (Schacter, Eich, & Tulving, 1978). In modem parlance, this was an episodic (or exemplar) theory, which assumes that every experience, such as perceiving a spoken word, leaves a unique memory trace. On presentation of a new word, all stored traces are activated, each according to its similarity to the stimulus. The most activated traces connect the new word to stored knowledge, the essence of recognition. The multiple-trace assumption allowed Semon's theory to explain the apparent permanence of specific memories; the challenge was also to create

/pdf/echoes-of-echoes-an-episodic-theory-of-lexical-access-4u3ubhcqbh.pdf

Echoes of echoes? An episodic theory of lexical access.

Infants acquire language with remarkable speed, although little is known about the mechanisms that underlie the acquisition process. Studies of the phonetic units of language have shown that early in life, infants are capable of discerning differences among the phonetic units of all languages, including native- and foreign-language sounds. Between 6 and 12 mo of age, the ability to discriminate foreign-language phonetic units sharply declines. In two studies, we investigate the necessary and sufficient conditions for reversing this decline in foreign-language phonetic perception. In Experiment 1, 9-mo-old American infants were exposed to native Mandarin Chinese speakers in 12 laboratory sessions. A control group also participated in 12 language sessions but heard only English. Subsequent tests of Mandarin speech perception demonstrated that exposure to Mandarin reversed the decline seen in the English control group. In Experiment 2, infants were exposed to the same foreign-language speakers and materials via audiovisual or audio-only recordings. The results demonstrated that exposure to recorded Mandarin, without interpersonal interaction, had no effect. Between 9 and 10 mo of age, infants show phonetic learning from live, but not prerecorded, exposure to a foreign language, suggesting a learning process that does not require long-term listening and is enhanced by social interaction.

https://www.pnas.org/content/pnas/100/15/9096.full.pdf

Foreign-language experience in infancy: Effects of short-term exposure and social interaction on phonetic learning

In the early months of life, infants acquire information about the phonetic properties of their native language simply by listening to adults speak. The acoustic properties of phonetic units in language input to young infants in the United States, Russia, and Sweden were examined. In all three countries, mothers addressing their infants produced acoustically more extreme vowels than they did when addressing adults, resulting in a "stretching" of vowel space. The findings show that language input to infants provides exceptionally well-specified information about the linguistic units that form the building blocks for words.

https://science.sciencemag.org/content/sci/277/5326/684.full.pdf

Cross-language analysis of phonetic units in language addressed to infants.

The automatic conversion of English text to synthetic speech is presently being performed, remarkably well, by a number of laboratory systems and commercial devices. Progress in this area has been made possible by advances in linguistic theory, acoustic-phonetic characterization of English sound patterns, perceptual psychology, mathematical modeling of speech production, structured programming, and computer hardware design. This review traces the early work on the development of speech synthesizers, discovery of minimal acoustic cues for phonetic contrasts, evolution of phonemic rule programs, incorporation of prosodic rules, and formulation of techniques for text analysis. Examples of rules are used liberally to illustrate the state of the art. Many of the examples are taken from Klattalk, a text-to-speech system developed by the author. A number of scientific problems are identified that prevent current systems from achieving the goal of completely human-sounding speech. While the emphasis is on rule programs that drive a format synthesizer, alternatives such as articulatory synthesis and waveform concatenation are also reviewed. An extensive bibliography has been assembled to show both the breadth of synthesis activity and the wealth of phenomena covered by rules in the best of these programs. A recording of selected examples of the historical development of synthetic speech, enclosed as a 33 1/3-rpm record, is described in the Appendix.

Review of text‐to‐speech conversion for English

Native speakers of Japanese learning English generally have difficulty differentiating the phonemes /r/ and /l/, even after years of experience with English. Previous research that attempted to train Japanese listeners to distinguish this contrast using synthetic stimuli reported little success, especially when transfer to natural tokens containing /r/ and /l/ was tested. In the present study, a different training procedure that emphasized variability among stimulus tokens was used. Japanese subjects were trained in a minimal pair identification paradigm using multiple natural exemplars contrasting /r/ and /l/ from a variety of phonetic environments as stimuli. A pretest‐posttest design containing natural tokens was used to assess the effects of training. Results from six subjects showed that the new procedure was more robust than earlier training techniques. Small but reliable differences in performance were obtained between pretest and posttest scores. The results demonstrate the importance of stimulus ...

/pdf/training-japanese-listeners-to-identify-english-r-and-l-a-1z4a30xyrw.pdf

Training Japanese listeners to identify English /r/ and /l/: A first report

Two experiments were carried out to extend Logan et al.’s recent study [J. S. Logan, S. E. Lively, and D. B. Pisoni, J. Acoust. Soc. Am. 89, 874–886 (1991)] on training Japanese listeners to identify English /r/ and /l/. Subjects in experiment 1 were trained in an identification task with multiple talkers who produced English words containing the /r/–/l/ contrast in initial singleton, initial consonant clusters, and intervocalic positions. Moderate, but significant, increases in accuracy and decreases in response latency were observed between pretest and posttest and during training sessions. Subjects also generalized to new words produced by a familiar talker and novel words produced by an unfamiliar talker. In experiment 2, a new group of subjects was trained with tokens from a single talker who produced words containing the /r/–/l/ contrast in five phonetic environments. Although subjects improved during training and showed increases in pretest–posttest performance, they failed to generalize to tokens produced by a new talker. The results of the present experiments suggest that variability plays an important role in perceptual learning and robust category formation. During training, listeners develop talker‐specific, context‐dependent representations for new phonetic categories by selectively shifting attention toward the contrastive dimensions of the non‐native phonetic categories. Phonotactic constraints in the native language, similarity of the new contrast to distinctions in the native language, and the distinctiveness of contrastive cues all appear to mediate category acquisition.

Training Japanese listeners to identify English /r/ and /l/. II: The role of phonetic environment and talker variability in learning new perceptual categories

In a recent study, Martin, Mullennix. Pisoni, and Summers (1989) reported that subjects' accuracy in recalling lists of spoken words was better for words in early list positions when the words were spoken by a single talker than when they were spoken by multiple talkers. The present study was conducted to examine the nature of these effects in further detail. Accuracy of serial-ordered recall was examined for lists of words spoken by either a single talker or by multiple talkers. Half the lists contained easily recognizable words, and half contained more difficult words, according to a combined metric of word frequency, lexical neighborhood density, and neighborhood frequency. Rate of presentation was manipulated to assess the effects of both variables on rehearsal and perceptual encoding. A strong interaction was obtained between talker variability and rate of presentation. Recall of multiple-talker lists was affected much more than single-talker lists by changes in presentation rate. At slow presentation rates, words in early serial positions produced by multiple talkers were actually recalled more accurately than words produced by a single talker. No interaction was observed for word confusability and rate of presentation. The data provide support for the proposal that talker variability affects the accuracy of recall of spoken words not only by increasing the processing demands for early perceptual encoding of the words, but also by affecting the efficiency of the rehearsal process itself. The perception of spoken language is a skill that requires the listener to extract stable linguistic content from a physical signal that is notoriously unstable. The acoustic realization of speech is simultaneously modulated by numerous variable phonetic, prosodic, and semantic characteristics inherent to the particular message. The speech signal is further modulated by variable source characteristics. Indeed, it has frequently been suggested that if the perceptual system is required to extract canonical units of meaning from spoken language, various idiosyncracies related to different talkers must be somehow "normalized" during an early stage of processing the sensory input (e.g., Joos, 1948). Such talker-specific sources of variability include differing individual dialects, vocal tract sizes and shapes, and speaking rates (Mullennix, Pisoni, & Martin 1988; Verbrugge, Strange, Shankweiler, & Edman, 1976). Sources of variability in speech, whether from phonetic context, spoken stress, or talker variations, have typically been considered "perceptual problems" to be solved by listeners, just as they must be solved for the design of adequate speechrecognition systems (e.g., Gerstman, 1968; Shankweiler, Strange, & Verbrugge, 1976). Although the communicative

/pdf/on-the-nature-of-talker-variability-effects-on-recall-of-hxn4xjk9jz.pdf

On the nature of talker variability effects on recall of spoken word lists.

This paper reports the results of an investigation that employed the modified rhyme test (MRT) to measure the segmental intelligibility of synthetic speech generated automatically by rule. Synthetic speech produced by ten text‐to‐speech systems was studied and compared to natural speech. A variation of the standard MRT was also used to study the effects of response set size on perceptual confusions. Results indicated that the segmental intelligibility scores formed a continuum. Several systems displayed very high levels of performance that were close to or equal to scores obtained with natural speech; other systems displayed substantially worse performance compared to natural speech. The overall performance of the best system, DECtalk—Paul, was equivalent to the data obtained with natural speech for consonants in syllable‐initial position. The findings from this study are discussed in terms of the use of a set of standardized procedures for measuring intelligibility of synthetic speech under controlled laboratory conditions. Recent work investigating the perception of synthetic speech under more severe conditions in which greater demands are made on the listener’s processing resources is also considered. The wide range of intelligibility scores obtained in the present study demonstrates important differences in perception and suggests that not all synthetic speech is perceptually equivalent to the listener.

/pdf/segmental-intelligibility-of-synthetic-speech-produced-by-2xtcusye1i.pdf

Segmental intelligibility of synthetic speech produced by rule.

We present the results of studies designed to measure the segmental intelligibility of eight text-to-speech systems and a natural speech control, using the Modified Rhyme Test (MRT). Results indicated that the voices tested could be grouped into four categories: natural speech, high-quality synthetic speech, moderate-quality synthetic speech, and low-quality synthetic speech. The overall performance of the best synthesis system, DECtalk-Paul, was equivalent to natural speech only in terms of performance on initial consonants. The findings are discussed in terms of recent work investigating the perception of synthetic speech under more severe conditions. Suggestions for future research on improving the quality of synthetic speech are also considered.

https://link.springer.com/content/pdf/10.3758%2FBF03201008.pdf

John S. Logan

Papers

Training Japanese listeners to identify English /r/ and /l/: A first report

Training Japanese listeners to identify English /r/ and /l/. II: The role of phonetic environment and talker variability in learning new perceptual categories

On the nature of talker variability effects on recall of spoken word lists.

Segmental intelligibility of synthetic speech produced by rule.

Perception of synthetic speech produced automatically by rule: Intelligibility of eight text-to-speech systems