2. Here I explain the existence of phones in terms of a revised motor theory of speech (Liberman, Cooper, Shankweiler, & Studdert-Kennedy, 1967; Liberman & Mattingly, 1985; 1989). According to Alvin Liberman and Ignatius Mattingly 'the objects of speech perception are the intended phonetic gestures of the speaker, represented in the brain as invariant motor commands' (Liberman & Mattingly, 1985, p. 2). And speech perception 'processes the acoustic signal so as to recover the coarticulated gestures that produced it. These gestures are the primitives that the mechanisms of speech production translate into actual articulator movements, and they are also the primitives that the specialised mechanisms of speech perception recover from the signal' (Liberman & Mattingly, 1989, p. 491). Alvin Liberman implies this arises after an imitation linkage in one of his first formulations: 'In its extreme and old-fashioned form, this view [that articulatory movements mediate between acoustic stimulus and word perception] says that we overtly mimic the incoming speech sounds and then respond to the proprioceptive and tactile stimuli that are produced by our own articulatory movements' (Liberman, 1957, p. 122).
3. The motor theory proposed here argues, in contrast to the formation made by Liberman and Mattingly, that the motor information in speech that enables the 'recovery of articulated gestures' exists simply for enabling the vocal imitiation required for infants to learn spoken vocabulary. Since this function is essential for the continued existence of language there is no need to suggest its use in a subsequent process such as word identification.
4. This new motor theory is proposed from a unified perspective that encompasses speech as an informational, developmental, evolutionary and neurobiological phenomena.
6. The continued existence of language therefore depends upon words being imitable. However, in spite of the importance of this, the theoretical and experimental concern of speech scientists is nearly exclusively with speech as spoken and heard. This perhaps is due to imitation being a transient and nonlinguistic activity, while word perception and word production are both constantly occurring and obviously the reason of why people speak. However, this is a mistake: as noted above, imitation, in spite of playing a mostly transient role in language acquisition, is a necessary process for the existence of speech.
7. Indeed, from an information processing perspective imitation is more fundamental as imitation rather than hearing must shape the nature of information that spoken words carry. The argument is this: the speech stream contains information for two information processes: (i) the identification of words and (ii) their imitation. Though it must carry information for both functions, it does not follow that each has equally shaped speech since it might be that (a) carrying one kind of information will always provide the other, but not vice versa; and (b) that carrying one kind of information requires that speech is highly tailored, while carrying the other requires no special adaptation. This is the situation in regard to the needs of information which enables imitation and word identification.
8. Let us briefly, review the special requirements if speech is to carry the information needed for vocal imitation.
(a). It must be auditory. (Though vision aids imitation, we can readily imitate speech when blind-folded).
(b). This auditory information must not only map into motor movements but movements which can recreate the auditory information.
(c). The imitation which is critical is done by infants. Such information must therefore enable movements to be copied from by adults with vocal apparatuses of very different shape and size.
(d). Such information must enable an extraordinary range of potential articulations. From the International Phonetic Association Alphabet we can see that the world's languages differ in making speech using thirteen vocal tract 'places' (from the lips to the space between the vocal cords, the glottis) and do so through no less than eleven types of movements (nasals to lateral clicks).
In sum, the need for speech to be imitable places considerable constraints upon the information that must be carried in speech.
9. In contrast, fewer, if any, informational contrasts apply to word identification. Unlike imitation this does not map from articulations to articulations, but from the totality of the speech stream to semantic units. Informationally, the totality of the speech stream carries many varied sources of information visual, auditory and contextual that provide a rich informational 'soup' that offers many redundant means by which words can be identified. Moreover, word identification can take advantage of the information that enables words to be imitated, since such information by its nature identifies a series of articulations and so provides information that can identify words.
10. Thus, on informational grounds, even though imitation is transient and a bystander activity, we would expect speech to have arisen to primarily carry information to aid word imitation. Since, first, the information needed for imitation must be highly specific for this purpose, but not that of the information enabling word identification. Second, while the information enabling imitation will always enable word identification, information aiding word identification will not necessary aid that vocal imitation. In other words, information contained in speech for imitation is privileged (speech must be shaped around it) over that enabling its identification.
(a). It is fluent.
(b). It can be done immediately (shadowing and echolalia), or it can be done after being held in short-term and long- term memory.
(c). It is independent of native language, language skills, word comprehension and speaker intelligence. Many autistic and some mentally retarded people engage in the echolalia of overheard words (often their only vocal interaction with others) without understanding what they echo (Fay, 1969; Schneider, 1938; Schuler, 1979; Stengel, 1947).
(d). It is prelinguistic: 18 week-old infants spontaneously copy vocal expressions provided the accompanying voice matches (Kuhl & Meltzoff, 1982). Imitation of vowels has been found as young as 12 weeks (Kuhl & Meltzoff, 1996).
(e). It happens quickly: words can be repeated within 250- 300 milliseconds both in normals (during shadowing) (Marslen-Wilson, 1973) and during echolalia by retarded individuals (Fay & Coleman, 1977).
(f). It happens directly. Indeed, it can be quicker to imitate a syllable than to initiate it. Porter and Lubker (1980) found as they put it 'simply *executing* a shift to [o] upon detection of a second vowel in [ao] takes very little longer than does *interpreting and executing* it as a shadowed response' (p. 599). As they add 'the early phases of speech analysis yield information which is directly convertible to information required for speech production'.
(g). It occurs independent of normal speech. Speech shadowing provides evidence of a 'privileged' input/output speech loop separate to the other components of the speech system (McLeod & Posner, 1984). Neurocognitive research likewise finds evidence of a direct (nonlexical) link between phonological analysis input and motor programming output (Coslett, Roeltgen, Rothi, & Heilman, 1987; McCarthy & Warrington, 1984).
12. In summary, we have a remarkable though still scientifically under-explored skill. It is unlikely we would possess it unless it played an important role in speech. That role is likely to be for theoretical and empirical reasons to enable vocabulary acquisition.
14. Many kinds of research show that this needs and involves direct imitation.
(a). Roughly, between one in twenty and nine in twenty words made by infants at around 24 months are mimicked (Bloom, Hood, & Lichtbown, 1974). These figures are minima since they concern only immediately overheard words. Many words that seem spontaneous are in fact delayed imitations overheard days or weeks previously (Miller, 1977).
(b). At 13 months children who imitate new words (but not ones they already know) show a greater increase in noun vocabulary at four months and nonnoun vocabulary at eight months (Masur, 1995).
(c). A major predictor of vocabulary increase in older children is their skill in repeating nonword phone sequences (a measure of mimicry and storage) (Gathercole & Baddeley, 1989).
(d). Language acquisition problems link to impairments in vocal imitation (Bishop, North, & Donlan, 1996).
15. In addition to aiding vocabulary increase, imitations aid the acquisition of speech in other ways.
(a). Imitation provides the basis for making longer sentences than children could otherwise spontaneously make on their own (Speidel & Herreshoff, 1989).
(b). Children analyse the internal linguistic rules and patterns of speech by repeating to themselves phrases and sentences overheard previously -- 'crib talk' (Kuczaj, 1983).
(c). Many proto-conversations involve children (and parents) repeating what each other has said in order to keep alive social and linguistic interaction.
(d). Repetition enables immigrant monolingual children to learn a second language by taking part in 'conversations' (Fillmore, 1979).
The capacity to imitate overheard words therefore is closely linked to our capacity to learn language.
17. First, not only does gender and age vary the anatomy with which speech is made but anatomy is also at an individual level highly variable. Sinus cavities vary 20-fold in volume, and show different degrees and lateralisation of asymmetry (Williams, 1967, pp. 26-27). Moreover, the shape and size of these cavities vary widely due the swelling of mucous membranes during upper respiratory tract infections 'colds'. Such variability is important in shaping the speech stream as it introduces variable 'antiresonances' in speech, and changes the spectral shape of nasal formants (Dang & Honda, 1996).
18. Second, vocal motor movements often start several speech units before they are pronounced - coarticulation.
19. Third, speech articulation can take place in many ways. Though they greatly effect how it is produced, colds, smoking a pipe, biting a pencil, clinching teeth (such as in ventriloquism), gross oral deformations such as hair-lips, cleft palates or amputations of the tongue tip do not prevent intelligible speech. So effective is articularly adaptation, that it enables us to speak in ways that grossly alter the mechanism of normal speech production. For instance, in buccal-source or 'Donald Duck' speech, we partially block off our vocal tract with the back of our tongues, and use our tongue, teeth and cheek to create vibrations to replace those normally created by our vocal cords (Weinberg, 1971).
20. Fourth, speech can be articulated in ways which diverge considerably in speed, timbre, pitch, loudness and emotion. Speech further exists in different forms such as song, verse, scream and whisper. Intelligible speech can be produced with pragmatic intonation and in regional dialects and foreign accents. Any method of imitating speech must be compatible with speech being produced in all these various ways.
22. This is not a radical idea: it is, after all, accepted that spoken words are sequences of motor movements organised around motor targets (Shaffer, 1984). If a lip articulating a consonant is knocked, the vocal apparatus makes a target related correction of movement (Gracco & Lvfqvist, 1994) -- much as perturbing an arm movement leads to an automatic correction to fulfil its spatial-temporal target (Haggard & Wing, 1995). There is no intrinsic reason why vocalisations should not be imitated in terms of the goals to which they aim.
23. How might motor goals be used to code imitation? Here I suggest a number of tricks are used.
24. First, the goals exploited are not spatial. The above observations that many different articulations can achieve them rules that out. Further, as James Abbs (1986, p. 206) has observed 'For speech motor actions, the individual articulatory movements would not appear to be controlled with regard to three- dimensional spatial targets, but rather with regard to their contribution to complex vocal tract goals such as resonance properties (e.g., shape, degree of constriction) and or aerodynamically significant variables'. Moreover, the actual physical positions of speech articulators are difficult to observe even when a speaker is clearly visible. Complex auditory goals however would be detectable merely from the speech stream in which they are produced.
25. Second, those goals that are used are done so categorically in contrasts. It is easier to communicate one of two states (present and nonpresent) than a degree in a continuum between them. For instance, consider imitating by sight the degree of quickness of a hand movement compared to imitating a contrast between a present 'quick' and a nonpresent state 'not quick' movement. In the latter case, all you need is imitate (and do so consistently) some appropriation to its quickness and slowness that preserves the category difference. Indeed, the use ofcontrast, allows the communication to variably widely (in speed or pitch) provided whatever form it takes preserves an identifiable contrast.
26. Third, infants face the problem of spotting from the 'blooming and buzzing confusion' of the speech stream which aspects of it need to be reused as the goals for a vocal imitation. This problem does not however arise if vocal goals use only invariants (particularly categorical invariants) for which the brain in its auditory processing is already sensitive. In this case, a child would overhear speech in terms of the innate perceptual sensitivities with which it is born, then let its motor imitation areas reuse them as motor goals in its attempt to mimic the units it has overheard. The only problem here is that an infant will initially hear speech in terms of all those invariants with which it is born. Thus, a child will need to go through a period of learning which subset (used by its surrounding language) to focus upon. As will be noted below, evidence exists that children in fact do go through such a stage (Aslin, Pisoni, Hennessy & Perey, 1981). 27. Fourth, only innate sensitivities that deal with high-level properties are used. For example, the manner in which second formants change, rather than low-level properties such as their pitch. The reason is that while low level acoustic traits are likely to be specific to individual anatomical productions, high- level ones can be aimed at independently of anatomical variation.
28. In summary, I conjecture that speech is imitable because it arose from coopting innate categorical invariant sensitives for its motor goals. This has several implications: (a) speech will be based around an imitation unit, (b) speech will arise both evolutionally and developmentally around motor imitation circuits in cooperation with those brain areas processing auditory invariants. (c) The processes behind speech will be amodal enabling nonauditory based forms of language.
(a). It will be made up of contrasts (let us call them features). These features will closely link to physical articulations since goals are usually achieved in one predominate way.
(b). However because motor goals can be achieved in many ways, this will not be a strict one-to-one mapping.
(c). Perception of the categorical invariants behind these features in spite of being associated with speech will be found in all mammals and possibly even in other vertebrates (such as birds) since they coopt invariances already present in the brain.
(d). Likewise, we would expect that newborns will be sensitive to those features found in languages other than that surrounding them.
(e). Since any language uses only a subset of them, we would expect languages to vary widely in which ones are used.
(f). We also would expect that the circuits which enable motor imitation in the nonhuman primate brain will have been coopted in humans to form those that enable speech.
(g). This unit will be present in speech purely to enable its imitation. There is no reason thus why it should also be used in the identification of words and so normally enter conscious awareness unless it is represented by symbols.
(h). Lastly, this unit will already have been identified in speech but not identified as an imitation code but have had attributed to it all manner of other functions. We would expect speech research to find many facts about this unit to be anomalous, and there to be a lack of consensus as to what it does.
30. I identify phones with this unit. This identification it should be noted is based purely from theoretical considerations as to what such an imitation code should look like.
31. Since Panini, phoneticians have known that phones are made by contrasting features such as voiced and unvoiced. The table of the International Association Alphabet consists of rows (method of production) and columns (means of articulation) in which such feature contrasts are noted. If we look at the table we see that these are mostly linked to differences in articulation.
32. However, as noted, they are known not to map one-to-one upon them.
33. Moreover, vocal articulations can like other kinds of motor articulation engage in several goals in parallel even though the goals are themselves sequential. (Consider, for instance, racket sports where a player will be aiming their racket to hit a ball while preparing to position themselves to return it: the movements are done in parallel even though the goals sequentially follow each other).
34. Languages vary widely in the features they use. This trivial fact is difficult to explain in natural selection terms since while evolution might evolve one set of features, it would not select to use only a small subset. This might give work to phoneticians and give a richness to the world's languages, but it does not increase from an evolutionary perspective the utility of speech as a form of communication.
35. As this theory requires, categorical contrasts between features can be perceived by a great variety of mammals (chinchilla: Kuhl & Miller, 1975; dogs: Adams, Molfese & Betz, 1987; macaques: Kuhl & Padden, 1983), and even birds (Japanese Quail: Kluender, Diehl & Killeen, 1987; Parrots: Patterson & Pepperberg, 1994).
36. Moreover, newborns are sensitive to those present in languages other than those spoken around them (Eimas, Miller & Jusczyk, 1987; Streeter, 1976; Trehub, 1976) though this sensitivity is lost when they start to speak (Aslin et al., 1981).
37. Another issue is that the awareness of phones is enhanced greatly in those familiar with their alphabetic symbolic representation (Morais, Cary, Alegria & Bertelson, 1979; Read, Zhang, Nie & Ding, 1986). Thus our familiarity with phones may not be representative of all speakers.
39. First, we would predict that homologous areas to the Wernicke areas in nonhuman primates would process the acoustic parameters which characterise phones such as fundamental frequencies, voice onset times and place of articulation. This has been reported: (Steinscheider, Arezzo, & Vaughan, 1982).
40. Second, we would predict that this area in nonhuman primates is involved in the goal perception of motor actions. This has also been found (Perrett et al. 1989).
41. Third, we would predict that motor imitation is linked to the homologous area of the Broca's area. Again this is established: (Rizzolatti, Fadiga, Gallese & Fogassi, 1996; Gallese, Fadiga, Fogassi & Rizzolatti, 1996).
42. Fourth, we would predict that human speech areas also function as imitation ones. This is the case: electrical simulation studies of the human brain find 81% of areas that show disruption of phone identification disrupted also the imitating of oral movements and vice versa (Ojemann, 1983); and that lesions in the speech areas show a 0.9 correlation between those causing impairments to the copying of oral movements and those impairing phone production and perception (Kimura & Watson, 1989). As this theory requires speech circuitry seems to have evolved out of prior existing motor goal imitation circuity.
43. Nothing in this theory requires that imitable speech should be restricted to the auditory modality. Imitable motor goals could also arise in the visual-spatial domain. Such forms of speech have indeed arisen with the sign languages of the deaf. Words in these languages unlike spoken ones are not made of sequential units but configurations of subword unit arrangements (Poizner, Klima & Bellugi, 1987). In spite of this difference, (which relates to the greater opportunity to visually present concurrent imitable motor features), sign language parallels spoken language. Both are learnt by imitation (indeed, autistics in deaf families echolalia in sign-language: Poizner, Klima, & Bellugi, 1987, p. 68). Both sign and vocal speech neurobiologically use the same areas linked with imitation in the brain, though sign languages use further areas particularly in the right hemisphere. (Hickok, Kritchevsky, Bellugi & Klima, 1995).
44. This theory suggests that speech arose from different evolutionary adaptions to those presently considered to be important. The components discussed above are all present in the primate brain. Indeed, the fact that they can be reemployed to enable sign-based language suggests no particular evolution of them was necessary. Moreover, as noted above, phonology cannot have been evolved specially for speech since any language only uses a small subset of phones. What was needed for speech instead was: (a) the respiratory control needed for producing sequences of vocalisations (an ability which distinguishes us from chimpanzees); (b) minor anatomical changes (development of pharynx and expansion of larynx). (c) Perhaps most importantly, the social need to communicate with a vocabulary-based language.
Adams, C. L., Molfese, D. L., & Betz, J. C. (1987). Electrophysiological correlates of categorical speech perception for voicing contrasts in dogs. Developmental Neuropsychology, 3, 175-189.
Allen, W. S. (1953). Phonetics in ancient India, London: Oxford University Press.
Aslin, R. N., Pisoni, D. B., Hennessy, B. L. & Perey, A. J. (1981). Discrimination of voice onset time by human infants: New findings and implications for the effects of early experience. Child Development, 52, 1135-1145.
Bishop, D. V. M., North, T., and Donlan, C. (1996). Nonword repetition as a behavioural marker for inherited language impairment: Evidence from a twin study. Journal of Child Psychology and Psychiatry, 37, 391-403.
Bloom, L., Hood, L. & Lichtbown, P. (1974). Imitation in language: If, when, and why. Cognitive Psychology, 6, 380-420.
Coslett, H. B., Roeltgen, D. P. Rothi, L. G. & Heilman, K. M. (1987). Transcortical sensory aphasia: Evidence for subtypes. Brain and Language, 32, 362-378.
Dang, J. & Honda, K. (1996). Acoustic characteristics of the human paranasal sinuses derived from transmission characteristic measurement and morphological observation. Journal of the Acoustic Society of America, 100, 3374-3384.
Eimas, P. D., Miller, J. L. & Jusczyk, P. W. (1987). On infant speech perception and the acquisition of language. In S. Harnad, (Ed.), Categorical perception, (pp. 161-195), New York: Cambridge University Press.
Fay, W. H. & Coleman, R. O. (1977). A human sound transduce/reproducer: Temporal capabilities of a profoundly echolalia child. Brain and Language, 4, 396-402.
Fay, W. H. (1969). On the basis of autistic echolalia. Journal of Communication Disorders, 2, 38-47.
Fillmore, L. W. (1979). Individual differences in second language acquisition. In C.J. Fillmore, D. Kempler & W. S-Y. Wang, (Eds.), Individual differences in langauge ability and language behavior. pp. 203-228, New York: Academic Press.
Gallese, V., Fadiga, L., Fogassi, L. & Rizzolatti, G. (1996). Action recognition in the premotor cortex. Brain, 119, 593-609.
Gathercole, S. E. and Baddeley, A. D. (1989). Evaluation of the role of phonological STM in the development of vocabulary in children: A longitudinal study. Journal of Memory and Language, 28, 200-213.
Gracco, V. L. & Lvfqvist, A. (1994). Speech motor coordination and control: Evidence form lip, jaw, and laryngeal movements. Journal Neuroscience, 14, 6585-6597.
Haggard, P. & Wing, A. (1995). Coordinated responses following mechanical perturbation of the arm during prehension. Experimental Brain Research, 102, 483-494.
Hickok, G., Kritchevsky, M., Bellugi, U & Klima, E. S. (1995). The role of the left frontal operculum in sign language aphasia: Clues to the function of Broca's area (Abstract). Brain and Language, 51, 179-181.
Kimura, D. & Watson, N. (1989). The relations between oral movement control and speech. Brain and Language, 37, 565-590.
Kluender, K. R., Diehl, R. L. and Killeen, P. R. (1987). Japanese Quail can learn phonetic categories. Science, 237, 1195-1197.
Kuczaj, S. A. (1983). Crib speech and language practice. New York: Springer-Verlag.
Kuhl, P. K., & Meltzoff, A. N. (1982). The bimodal perception of speech in infancy. Science, 218, 1138-1141.
Kuhl, P., & Miller, J. (1975). Speech perception by the chinchilla: Voice-voiceless distinction in alveolar plosive consonants. Science, 190, 69-72.
Kuhl, P., & Padden, D. (1983). Enhanced discriminability at the phonetic boundaries for the voicing feature in macaques. Perception and Psychophysics, 32, 542-550.
Kuhl, P. K., & Meltzoff, A. N. (1996). Infant vocalizations in response to speech: Vocal imitation and developmental change. Journal of the Acoustic Society of America, 100, 2425-2438.
Liberman, A. M. (1957). Some results of research on speech perception. Journal of the Acoustical Society of America, 29, 117-123.
Liberman, A. M., Cooper, F. S., Shankweiler, D. P. & M. Studdert- Kennedy. (1967). Perception of the speech code. Psychological Review, 74, 431-461.
Liberman, A. M. & Mattingly, I. G. (1985). The motor theory ofspeech perception revised. Cognition, 21, 1-36.
Liberman, A. M. and Mattingly, I. G. (1989). A specialization for speech perception. Science, 243, 489-494.
Marslen-Wilson, W. (1973). Linguistic structure and speech shadowing at very short latencies. Nature, 244, 522-523.
Masur, E. F. (1995) Infants' early verbal imitation and their later lexical development. Merrill-Palmer Quarterly, 41, 286-306.
McCarthy, R & Warrington, E. K. (1984). A two-route model of speech production: Evidence from aphasia. Brain, 107, 463-485.
McLeod, P. & Posner, M. I. (1984). Privileged loops from percept to act. In H. Bouma & D. Bouwhuis, (Eds.), Attention and performance X. pp, 55-66, Hillsdale, NJ: Erlbaum.
Miller, G. A. (1977). Spontaneous apprentices: Children and language. New York: Seabury Press.
Morais, Josi., Cary, Luz., Alegria, Jisus., and Bertelson, Paul. (1979). Does awareness of speech as a sequence of phones arise spontaneously? Cognition, 7, 323-331.
Ojemann, G. A. (1983). Brain organization for language form the perspective of electrical stimulation mapping. Behavioral and Brain Sciences, 6, 189-230.
Patterson, D. K. & Pepperberg, I. M. (1994). A comparative study of human and parrot phonation: Acoustic and articulatory correlates of vowels. Journal of the Acoustical Society of America, 96, 634-648.
Perrett, D. I., Harries, M. H., Bevan, R., Thomas, S., Benson, P. J., Mistlin, A. J., Chitty, A. J., Hietanen, J. K. & Oretega, J. E. (1989). Frameworks of analysis for the neural representation of animate objects and actions. Journal of Experimental Biology, 146, 87-113.
Pollack, I., & Pickett J. M. (1963). The intelligibility of excerpts from conversation. Language and Speech, 6, 165-171.
Porter, R. J. & Lubker, J. F. (1980). Rapid reproduction of vowel-vowel sequences: Evidence for a fast and direct acoustic- motoric linkage in speech. Journal of Speech and Hearing Research, 23, 593-602.
Read, C., Zhang, Y., Nie, H., & Ding, B. (1986). The ability of manipulate speech sounds depends on knowing alphabetic writing. Cognition, 24, 31-44.
Remez, R. E., Rubin, P. E., Pisoni, D. B., & Carrell, T. D. (1981). Speech perception without traditional speech cues.Science, 212, 947-950.
Rizzolatti, G., Fadiga, L., Gallese, V. and Fogassi, L. (1996). Premotor cortex and the recognition of motor actions. Cognitive Brain Research, 3, 131-141.
Rizzolatti, G, & Arbib, M. A. (in press). Language within our grasp. Trends in Neuroscience.
Schneider, D. E. (1938). The clinical syndromes of echolalia, echopraxia, grasping and sucking. Journal of Nervous and Mental Disease, 88, 18-35, 200-216.
Schuler, A. L. (1979). Echolalia: Issues and clinical applications. Journal of Speech and Hearing Disorders, 44, 411- 434.
Shaffer, L. H. (1984). Motor programming in language production. In H. Bouma, & D. G. Bouwhuis, (Eds.), Attention and performance, X. London: Erlbaum.
Shannon, R. V., Zeng, F-G., Kamath, V., Wygonski, J., & Ekelid, M. (1995). Speech recognition with primarily temporal cues. Science, 270, 303-304.
Speidel, G. E. & Herreshoff, M. J. (1989). Imitation and the construction of long utterances. In G. E. Speidel & K. E. Nelson, (Eds.), The many faces of imitation in langauge learning. pp.181- 197, New York: Springer-Verlag.
Stengel, E. (1947). A clinical and psychological study of echo- reactions. Journal of Mental Science, 93, 598-612.
Streeter, L. A. (1976). Language perception of 2-month-old infants shows effects of both innate mechanisms and experience. Nature, 259, 39-41.
Trehub, S. E. (1976). The discrimination of foreign speech contrasts by infants and adults. Child Development, 47, 466-472.
Weinberg, B. (1971). A study of buccal-speech. Journal of Speech and Hearing Research, 14, 657-658.
Williams, R. J. (1967). You are extra-ordinary. New York: Random House.
Zatorre, R. J., Meyer, E., Gjedde, A., & Evans, A. C. (1996). PET studies of phonetic processing of speech: Review, replication, and reanalysis. Cerebral Cortex, 6, 21-30.