Joan L. Bybee
The article demonstrates many aspects of grammar that can be derived from domain-general cognitive processes, especially those of neuromotor automation, chunking, categorization, inference-making, and cross-modal association. Construction grammar posits a direct connection between the conventionalized constructions of a language and their meanings. All constructions have some specific lexical or grammatical material in them. Construction grammar emphasizes the interaction of the lexicon with the syntax. The domain-general processes involved in construction formation and use are sequential processing and categorization. Sequential processing or chunking is the process by which repeated sequences of experience (words or other events) come to be grouped together in memory as units that can be accessed directly. Categorization is necessary to the cognitive representations of constructions in several ways. First, categorization is necessary for the recognition that an element or sequence is the same as one previously experienced. Second, categorization is used to develop the schematic slots of constructions. Constructions are created through the repetition and thus conventionalization of useful sequences of elements and their meanings arise from associations with the context and implications that are present. The most pervasive process by which new constructions are created is grammaticalization, in which a new construction is created along with a new grammatical morpheme and the latter evolves from a lexical morpheme or combinations of grammatical and lexical morphemes.
Grammaticalization theory has become an influential theory within historical linguistics. Grammaticalization is the process whereby open-class lexical items develop over time into closed-class items with grammatical functions. It is said to be a uniform series of semantic changes involving metaphorical usage such as spatial terms acquire temporal meanings but not vice versa and “bleaching”. Grammaticalization often leads to “morphologization”, which is an independent marker of tense or where a number becomes an affix rather than remaining a free wordform, and may even ultimately fuse with the root of the lexeme to which it is attached. It is always possible for grammaticalization to stop short of morphologization that applies to most of the languages of East Asia. Even without any syntax there could be phonological processes operating between regularly contiguous “words”, and some of these processes could in due course become opaque. This would give rise to situations where the same meaning was expressed by two or more forms in different contexts that is, to instances of synonymy. Even without syntax, some items could be regularly juxtaposed to express a consistent conventionalized meaning. The capacity for allomorphy and morphophonological alternation could have arisen alongside syntax or even before it, but at any rate independently of it.
Víctor M. Longa, Guillermo Lorenzo, and Juan Uriagereka
This article sketches the sorts of issues that arise within evo-minimalism. Lewontin reminds us that the evolutionary study of cognition, and more particularly the language faculty, is no easy game, given the distance between the proteins genes express and the cognitive functions in whose development or deployment such proteins may be implicated. The article attempts to sketch the sorts of difficulties that arise in relating all the necessary steps leading from one to the other extreme of such a causal chain, which casts some doubt on the viability of constructing true evolutionary explanations in the realm of cognitive novelties such as the ‘Faculty of Language–Narrow Sense’. In addition, several difficulties exist in deciphering whether putative relations of hypothesized capacities in various species are genuine homologies, or instead, weaker analogies.
The basic processes of syntax are categorized into four main categories. These categories include a process for assembling words into hierarchical structures, processes for determining the boundaries of segments within such structures, and processes for moving segments within such structures and lastly, processes for determining the reference of elements that are not phonetically expressed. The syntax characterizes all languages, whether signed or spoken, in a highly developed form but it is entirely absent both from the productions of “language-trained” animals and the natural communication systems of other species. Children first acquire nouns, then a few verbs, and only later begin to add other word classes. The acquisition of grammatical items follows some time after the emergence of recognizable syntactic structures, even if those structures do not normally begin to appear until age two or thereabouts, and the earliest stages of development constitute an example of protolanguage, rather than full human language. The emergence of these structures is typically quite rapid with several types of both simple and complex sentence appearing within a few weeks. Creole languages are really a special case of child language development, representing what the language faculty produces when structured input is severely reduced.
This article discusses the kinds of universal syntactic capacities that prompt many linguists to postulate that syntax is a separate, biologically determined entity. Lexical items are stored in the “mental lexicon”, an inventory of arbitrary form-meaning associations. These include single words and morphemes smaller than words (such as affixes), multi-word idioms, set phrases, and constructions of various kinds. Vocabulary learning in humans is sophisticated, involving a complex mix of grammatical properties, phonology, semantics, and cultural knowledge. Some aspects are universal, others language-specific. Vocabulary items belong to complex, structured semantic categories. Universally, verbs fit into one or more “subcategorization frames”, which specify the number and type of obligatory dependents of the verb. The human lexicon displays at least three further unique characteristics. First, the speakers probably store at least 50,000 entries for each of their native languages. Second, though the learning of syntax, phonology, and morphology is subject to critical period effects, new lexical items are learned throughout life and there is no critical period for vocabulary acquisition. Third, the human lexicon crucially contains two major classes of items that include content words, known as lexical categories, and grammatical elements, or functional categories. The lexical/functional division occurs in all languages, including simple languages, such as Riau Indonesian, although functional categories in particular vary greatly cross-linguistically.