2. Language Faculty/LAD:
2.1 Universal Grammar:
2.2 Principles and Parameters:
2.3 Maturation and Continuity:
3. Parameter Theory:
3.1 The Initial State of Parameters:
3.3 Possible Settings of Parameter Values:
4. Parameters and Bilingualism
5. Parameters in Computational Linguistics:
This paper is mainly investigating the properties of parameters and their connection to language acquisition. One of the core questions will be, in how far Chomsky´s assumption of parameters being generally equipped with a default value can be validated. In order to obtain a more detailed account of parameters in language acquisition, a chronological analysis of relevant study areas will be made, moving from a broad perspective to more and more detailed descriptions. This analysis contains a brief overview of the Faculty of Language and the LAD as well as of Universal Grammar, simply because the parameters in question derive from these systems. Afterwards a detailed look will be taken on specific features of parameters, taking Meisel´s text `Parameters in Acquisition´ as a basis. To become more specific in order to further question certain features of parameters that have occurred by then, further linguistic fields will be taken into account, namely bilingual studies, neurolinguistics, and computational linguistics.
2. Language Faculty/LAD:
First defined by Chomsky in 1965, the Faculty of Language can be seen as human´s tool to incorporate, process, and produce speech. Containing the Language Acquisition Device, it is furthermore responsible for children´s capacity to learn all possible human languages quasi automatically. According to Chomsky, Fitch and Hauser, the language faculty should be separated into two basic parts, namely the faculty of language in the broad sense (FLB) and in the narrow sense (FLN) (2002: 1569). They define the FLB by including “a sensory-motor system, a conceptual-intentional system, and the computational mechanisms for recursion, providing the capacity to generate an infinite range of expressions from a finite set of elements”, whereas the “FLN only includes recursion” (2002: 1569). In evolutionary terms, it should be mentioned that they assume that the FLN “is the only uniquely human component of the faculty of language”, enabling humans to communicate with theoretically infinite speech systems, thereby distinguishing humans from presumably all other animals (Chomsky, 2002: 1569). FLN is the abstract linguistic computational system alone, independent of the other systems with which it interacts and interfaces. FLN is a component of FLB, and the mechanisms underlying it are some subset of those underlying FLB (Chomsky, 2002: 1570).
From a language acquisitional point of view, “a key component of FLN is a computational system (narrow syntax) that generates internal representations and maps them into the sensory-motor interface by the phonological system, and into the conceptual-intentional interface by the (formal) semantic system” (Chomsky, 2002: 1571). Chomsky, Fitch and Hauser further argue that “according to recent linguistic theory, the computations underlying FLN may be quite limited” (2002: 1573). They “propose […] that FLN comprises only the core computational mechanisms of recursion as they appear in narrow syntax and the mappings to the interfaces” (Chomsky, 2002: 1573). They furthermore state that the “FLN may approximate a kind of `optimal solution´ to the problem of linking the sensory-motor and conceptual-intentional systems”, meaning that the generative processing of FLN satisfies the necessary connections between narrow syntax and sensory-motor as well as conceptual-intentional system in a near optimal fashion (2002: 1574).
It shall be noticed and kept in mind that Chomsky and his colleagues take into consideration a possible outsourcing of crucial language elements (e.g.; subjacency, Wh- movement, the existence of garden-path sentences) from the FLN into the FLB: “Many of the details of language that are the traditional focus of linguistic study […] may represent by-products of this solution [optimal linkage of FLB and FLN], generated automatically by neural/computational constraints and the structure of FLB - components that lie outside of FLN” (Chomsky, 2002: 1574).
Studies of the function of the faculty of language differentiate between external language (E-language) and internal language (I-language) (Chomsky, 2002: 1570). The latter is the primary object when exploring the mechanisms of language acquisition.
Finally, it shall be mentioned that Chomsky takes “possible variation among humans in the language faculty” into consideration, but with respect to UG “[i]t is plausible to suppose that apart from pathology (potentially an important area of inquiry), such variation as there may be is marginal and can be safely ignored across a broad range of linguistic investigation” (Chomsky, 1986: 18-9).
2.1 Universal Grammar:
The theory of UG formulates this a priori knowledge in terms of principles and parameters which determine the set of possible human languages”, “representing the innately specified initial state of the language faculty [and] it is also understood as a crucial component of the language acquisition device (Meisel 1995: 11 & 13).
The studies of Universal Grammar, UG being part of the faculty of language, are based, among other things, on “Plato´s Problem”, i.e. the “discrepancy between experience and knowledge” (Meisel 1995: 11). Since a child´s language input is not sufficient in learning all grammar rules of a given language, it is assumed that the human mind contains an inherited knowledge of foundational rules of all languages, which in combination with language input leads to full proficiency in any of the world´s languages. However, according to the `Principles and Parameter Theory´, “developed by Chomsky (1981) and others”, UG, says Meisel, is not simply conceived as a set of rule:
[E]ven if one were to succeed in writing descriptively adequate grammars consisting of sets of rules, no progress would be made towards explanatory adequacy […]. […] [T]he more elaborate the rules […] became, the less obvious it was how children could acquire such rule systems and how they would evaluate them in comparison to descriptively equally adequate ones (Meisel 1995: 11).
Therefore, UG is first all to be understood as an abstract model and its rules are supposed to be as simple as possible and as flexible as necessary with reference to language´s capacity for change and to constantly adapt to its environment. However, Chomsky states that “whatever system is responsible [for language acquisition] must be biased or constrained in certain ways. Such constraints have historically been termed `innate dispositions´, with those underlying language referred to as `universal grammar´. […] [T]he existence of some such constraints cannot be seriously doubted” (Chomsky, 2002: 1577). Radford highly supports Chomsky´s assumption by arguing that „only a maximally constrained theory of language can lead to the development of an adequate theory of language acquisition. […] [W]e can only explain the phenomenon of language acquisition in children if we assume that grammars contain a highly constrained set of principles […]” (Radford, 1988, 34). He further supports Chomsky´s Structure-Dependence Principle which specifies that all grammatical rules are structure-dependent (Radford, 1988: 35): “The reliance on structure-dependent operations must be predetermined for the language-learner by restrictive initial schematism of some sort that directs his attempts to acquire linguistic competence” (Chomsky, 1972: 63).
In other words, certain initial constrains and biases must be applied to a system of language acquisition not only in order to define it, but especially because it must consist of certain regulation to be able to perform the immense task of acquiring a language almost effortless. The notions of simplicity and constraints are obviously of crucial importance for UG, but it also has to be remarked here that the notion of flexibility within this system does not inevitably contradict with biases and constrains, which will be shown later on.
2.2 Principles and Parameters:
The `Principle and Parameter Theory´ “distinguishes between phenomena related to nonparameterized universal principles and those that depend on parameters of UG or on language-specific properties of grammar” (Meisel 1995: 13). A principle is true for all human languages, e.g. that all sentences must contain a subject constituent, even if it is not overtly pronounced, or the x-bar principle. Parameters must be set according to the requirements of the language being acquired, e.g. if a specific language contains the “null-subject” or “pro-drop” parameter. Hence, parameters allow variations in children´s language acquisition, i.e. different developments of grammars of different languages.
The principles itself are divided into absolute universals and relative universals (Radford, 1988: 38-9). Absolute universals, like the word says, are of absolute validity. Again, subject obligation and the x-bar principle can serve as an example here. Relative universals on the other hand tend to be predominantly applicable to all languages, but might have “numerous exceptions” (Radford, 1988: 39). Radford is giving the Consistent Serialisation Principle, the fact that “[l]anguages tend to place modifying elements either consistently before or consistently after elements (or heads)”, as an example for such a relative universal, because there are a few of the world´s languages like Samoan in which Modifiers follow their heads (Radford, 1988: 38).
Radford also discusses the “Theory of Core Grammar”, the idea of a “universal `core´ of linguistic principles which characterise the full range of unmarked linguistic phenomena found in natural language”, resulting in a “rule of Core Grammar” (“core rule”) (Radford, 1988: 41). According to Chomsky, mechanisms that are situated within a core grammar can be described as unmarked, meaning they represent a dominant default value (1986: 146). “Unmarked”, in its simplest meaning, could define a “normal” linguistic unit (Chomsky, 1986: 146). Opposite to these unmarked values, one finds the “Periphery of marked rules and structures” (peripheral rule), defining rather irregular values (Radford, 1988: 41). Chomsky summed up the characteristics of core/unmarked and peripheral/marked values as follows:
The distinction between core and periphery leaves us with three notions of markedness: core versus periphery, internal to the core, and internal to the periphery. The second has to do with the way parameters are set in the absence of evidence. As for the third, there are, no doubt, significant regularities even in departures from the core principles (for example, in irregular verb morphology in English), and it may be that peripheral constructions are related to the core in systematic ways, say by relaxing certain conditions of core grammar (Chomsky, 1986: 147).
Core and periphery should not be seen as mere opposites. Chomsky clearly sees a high degree of interdependency between the two sides. In the end, it will be the flexible interplay between marked and unmarked values that will allow a child to learn language.
So far, having briefly explained the Language Faculty, the LAD, UG, and the theory of principles, with a slight outlook on its parameters, it becomes more or less clear that the immensity of terminologies connected to language acquisition are highly abstract and that theories seem to have a high tendency of overlapping each other. The notion of “relative universals”, for example, shows a high similarity to “parameters”; simply due to the fact that its rules are just not universal. Notions such as “unmarked core grammar” and “principle” as well as “marked peripheral grammar” and “parameters” also contain much similar information. Therefore, it appears reasonable to focus mainly on the ideas of “principles and parameters”, simply to simplify, but without declaring all other terminologies unimportant or even negating them, not at all.
Defining principles, even though essential for the LAD and the outcome of parameters, has the disadvantage that they are hard to grasp, one-dimensional, invariant (at least in the long run), and consequentially cannot account for differences in language acquisition. Parameters, on the other hand, are numerous, vary in their specific values, are more likely to be definable, and finally differentiate one language from another. Parameters should therefore be of an undeniable value in explaining phenomena of language acquisition. Meisel confirms this view, stating that from an acquisitional perspective the parameterized principles of UG are of much more interest because one is actually able to analyse which parameter values are getting set during a child´s speech development (1995: 14). These possible choices are not predetermined a priori but depending “on the basis of information available in the input” (Meisel, 1995: 14). The importance of parameters becomes even more obvious if one is confronted with the paradox that children and adults are supposed to work with the same rules of UG, but their grammars differ (Pinker, 1984: 7; Meisel, 1995: 13). Therefore, principles alone have very few to say about language specific properties.