Connectionism is an important theoretical framework for the study of human cognition and behavior. Also known as Parallel Distributed Processing (PDP) or Artificial Neural Networks (ANN), connectionism advocates that learning, representation, and processing of information in mind are parallel, distributed, and interactive in nature. It argues for the emergence of human cognition as the outcome of large networks of interactive processing units operating simultaneously. Inspired by findings from neural science and artificial intelligence, connectionism is a powerful computational tool, and it has had profound impact on many areas of research, including linguistics. Since the beginning of connectionism, many connectionist models have been developed to account for a wide range of important linguistic phenomena observed in monolingual research, such as speech perception, speech production, semantic representation, and early lexical development in children. Recently, the application of connectionism to bilingual research has also gathered momentum. Connectionist models are often precise in the specification of modeling parameters and flexible in the manipulation of relevant variables in the model to address relevant theoretical questions, therefore they can provide significant advantages in testing mechanisms underlying language processes.
Blocking can be defined as the non-occurrence of some linguistic form, whose existence could be expected on general grounds, due to the existence of a rival form. *Oxes, for example, is blocked by oxen, *stealer by thief. Although blocking is closely associated with morphology, in reality the competing “forms” can not only be morphemes or words, but can also be syntactic units. In German, for example, the compound Rotwein ‘red wine’ blocks the phrasal unit *roter Wein (in the relevant sense), just as the phrasal unit rote Rübe ‘beetroot; lit. red beet’ blocks the compound *Rotrübe. In these examples, one crucial factor determining blocking is synonymy; speakers apparently have a deep-rooted presumption against synonyms. Whether homonymy can also lead to a similar avoidance strategy, is still controversial. But even if homonymy blocking exists, it certainly is much less systematic than synonymy blocking. In all the examples mentioned above, it is a word stored in the mental lexicon that blocks a rival formation. However, besides such cases of lexical blocking, one can observe blocking among productive patterns. Dutch has three suffixes for deriving agent nouns from verbal bases, -er, -der, and -aar. Of these three suffixes, the first one is the default choice, while -der and -aar are chosen in very specific phonological environments: as Geert Booij describes in The Morphology of Dutch (2002), “the suffix -aar occurs after stems ending in a coronal sonorant consonant preceded by schwa, and -der occurs after stems ending in /r/” (p. 122). Contrary to lexical blocking, the effect of this kind of pattern blocking does not depend on words stored in the mental lexicon and their token frequency but on abstract features (in the case at hand, phonological features). Blocking was first recognized by the Indian grammarian Pāṇini in the 5th or 4th century bc, when he stated that of two competing rules, the more restricted one had precedence. In the 1960s, this insight was revived by generative grammarians under the name “Elsewhere Principle,” which is still used in several grammatical theories (Distributed Morphology and Paradigm Function Morphology, among others). Alternatively, other theories, which go back to the German linguist Hermann Paul, have tackled the phenomenon on the basis of the mental lexicon. The great advantage of this latter approach is that it can account, in a natural way, for the crucial role played by frequency. Frequency is also crucial in the most promising theory, so-called statistical pre-emption, of how blocking can be learned.