Is the ability of identifying a given [a] sound as phoneme /a/ necessary for spoken language competence?

Developmental psychology tells that infants acquire language through the vocal imitation but no infants try to imitate the voices of their parents. It is known that myna birds imitate the voices and sounds of their keepers. Why don’t infants imitate the voices and sounds? Since the timbral characteristics of sounds are completely controlled by the shape of the sound generator, the voice imitation requires the same shape of the vocal tube that the parents have. Considering this reason, it is impossible for infants to imitate the voices of their parents. Then, why don’t they try to imitate them and why do they perceive the identity between the two different sound streams, e.g, mother’s “Good morning” and father’s “Good morning”? Some readers may reply that infants decode the input streams into two sequences of phonemes and perceive the identity between the two phonemic sequences. Developmental psychology, however, denies this proposal because it claims that the segments and the phonemic awareness are learned later than the vocal imitation. In this work, taking the above question as one of the questions in mathematics and physics, it is answered by introducing a new concept of relative timbre. The proposed framework claims that the ability of identifying a given linguistic sound as phoneme is not required for spoken language competence. As far as the authors know, the cases are easily found in developmental dyslexics.