Three myths from the language acquisition literature

One model for the learning of language

A major goal of linguistics and cognitive science is to understand what class of learning systems can acquire natural language. Until recently, the computational requirements of language have been used to argue that learning is impossible without a highly constrained hypothesis space. Here, we describe a learning system that is maximally unconstrained, operating over the space of all computations, and is able to acquire many of the key structures present in natural language from positive evidence alone. We demonstrate this by providing the same learning model with data from 74 distinct formal languages which have been argued to capture key features of language, have been studied in experimental work, or come from an interesting complexity class. The model is able to successfully induce the latent system generating the observed strings from small amounts of evidence in almost all cases, including for regular (e.g., an , [Formula: see text], and [Formula: see text]), context-free (e.g., [Formula: see text], and [Formula: see text]), and context-sensitive (e.g., [Formula: see text], and xx) languages, as well as for many languages studied in learning experiments. These results show that relatively small amounts of positive evidence can support learning of rich classes of generative computations over structures. The model provides an idealized learning setup upon which additional cognitive constraints and biases can be formalized.

Further evidence of automatic reinforcement effects on verbal form

The form of a verbal response allows for reinforcement mediation and language transmission across cultures. Reinforcement, in turn, plays a decisive role in learning verbal forms. The present work addresses methodological limitations of previous studies, providing further evidence of the role of automatic reinforcement in achieving parity with vocal models. In the first experiment, 4 preschool-age children heard the experimenter describe drawings of different actions in the passive voice. Participants were then asked to describe analogous drawings. They used the passive voice after the model was presented and continued to do so even when preferred explicit consequences followed diverging descriptions (i.e., in the active voice). To further investigate the effects of explicit reinforcement and of the passive-voice model, in Experiment 2, we altered the number of trials with explicit reinforcement and with the model. Three of four participants used the passive voice to describe the drawings, despite greater exposure to explicit consequences following descriptions diverging from the model.

The Importance of Form in Skinner's Analysis of Verbal Behavior and a Further Step

A series of quotes from B. F. Skinner illustrates the importance of form in his analysis of verbal behavior. In that analysis, form plays an important part in contingency control. Form and function complement each other. Function, the array of variables that control a verbal utterance, dictates the meaning of a specified form; form, as stipulated by a verbal community, indicates that meaning. The mediational actions that shape verbal utterances do not necessarily encounter their controlling variables. These are inferred from the form of the verbal utterance. Form carries the burden of implied meaning and underscores the importance of the verbal community in the expression of all the forms of language. Skinner's analysis of verbal behavior and the importance of form within that analysis provides the foundation by which to investigate language. But a further step needs to be undertaken to examine and to explain the abstractions of language as an outcome of action at an aggregate level.