Lawler’s Language Learning Micro Theory
Human infants become capable of symbolic thought through a three phase process, where the phases are both interdependent and at least marginally overlapping. The first phase marks the development of a proto-language from innate vocalizations, such as cries. Elements of the protolang are icons, in the sense of Peirce and Deacon. The protolang itself may be considered a suite of iconic tokens, primarily vocal, invoking aid of the universe in satisfying the infant’s wants. When the infant recognizes the power of its vocalization, the desire for satisfaction of needs becomes communicative and is realized as commands more than as expressions of distress or dissatisfaction. In this specific sense, communication and the intention to communicate begins as a sort of advancement effect of associative learning. The protolang is primarily expressive, i.e. communicative in one direction.
An ultimate objective is to construct programmed models representing these phases and their development using computational ideas from Minsky’s Society of Mind. Mechanistically, we would say that infant vocalizations are tokens, icons for a goal: to change the current mental state to a different mental state. These can be represented by Knowledge-lines, the elemental components of Minsky’s memory theory, which are seen as partial residues of previous mental states (in all their multi-modal complexity). When the vocalization becomes salient as an action from which the mental state change is “expected” by an anticipation effect, then the vocalization can be inferred to specify an intention to communicate.
In the second phase, the protolang of the infant is transformed to an infant’s personal language (hereafter called an idiolang) which can be characterized as a suite of vocal signs having indexical reference to desired things in the world, as Peirce would use the term indexical. This implies they can be understood by some care-givers in the infant’s world. [note 1.] The motor of this transformation is that choices available to the care-giver need be translated into choices to be decided by the infant. For example, suppose a mother says to an infant “Do you want this or that?” holding a cookie in one hand and a cup of juice in the other, expecting the infant to indicate what it wants before she provides either; the infant may indicate the choice by pointing at one object or the other or the infant may say /thaet/ — merely repeating the last sound heard. The care-giver will imagine the infant has indicated that it wants the juice, whatever the infant actually wanted. Through such communication and mis-communication, the infant will gradually learn that vocalizations can function as names of objects and actions. Socially, we would say that the broadening of useful indexical reference occurs through oral communication within the infant’s world.
Mechanistically, we would say that the infant develops super-iconic interpretants (processes for determining references between signs and things signified) and discovers the arbitrariness of signs, i.e. their conventionality. This development generates the third phase, the purging of the idiolang through replacement of its infant-personal terms with conventional words.
[1.] If we focus on the generation of vocal signs as indexes to things signified, infants’ idiosyncratic terms for things can be a consequence of their mis-perception of or inadequate mimicking of vocalizations of others (e.g. /cul//di/ as the name of our dog “Scurry”). Similarly, an infant might intend to refer to some thing or action for which she has no term: /vae//vae/ as a signifying of “barking” might serve as an example of onomatopoeia; the corpus may hold other examples of purely arbitrary vocalizations assigned to objects or actions, but this would require investigation to determine.