Theory, Model, Method
Levels of theoretical adequacy
Observations describe specific facts. A theory is considered adequate on an observation level if it covers a given set of data in a theory of common scientific standards. It describes the data in a theoretical way. This may include the construction of an order or even a system of these observations (observed objects).
The components of a theory according to Chomsky's generative grammar
Source of the picture: ilit.umbc.edu
Listing observations as a catalogue of data or phenomena also is observational, but as long as it is not formulated in theoretical standards (i.e. it is not constructed as a theory), it is a pre-scientific observation. Traditional grammars are often compilations of such observations on language. Just collecting words to build up a dictionary of a certain language is an observational task; gathering mispronunciations, speech errors, or wrong usages of words are other examples of linguistic observational tasks.
Hypotheses do not directly describe specific facts; they are general statements expressing relationships among observed phenomena. Observations are made for the purpose of testing hypotheses as a descriptive generalization of observations.
If a hypothesis is confirmed by all the scientist’s observations and not falsified by any, then it may achieve the status of a law.
Descriptively adequate theories are constructed in order to cover (describe) given data and they are also supposed to account for new data. They are, among other things, expected to give theory-bound classifications of the phenomena (theoretical constructs). Modern descriptive grammars are descriptive in this sense. We expect that their rules are theory-based (meaning: they are derived / deduced from a general theory of language or grammar) and will be true for sentences, words, or pronunciations we will come across in future talk situations, although they have not yet been written down in the grammars.
If statements are intended to explain the phenomena, they are formulated in more abstract terms. These terms are theoretical constructs in that they do not correspond to any observed phenomena. Theoretical statements are a crucial part of explanatory theories.
The explanatory power of theories extends to answering the questions “why”, true theories of a common scientific standard are expected to explain a given set of phenomena. Depending on the object under examination, answers may be a) causal (giving explanations by reference to causal laws like those in the natural sciences), or b) final (or teleological) by accounting for the aims, goals, purposes that “explain” the object (e.g. a certain verbal behaviour, a certain form of a text or a talk, etc.). Verbal communicative acts are object to final explanations, the core set of linguistic structures is said to underlie causal explanations.
If theories make statements that are built up to predict certain instances of the phenomena in the future, that is which are not yet existent or can at least be observed, then these theories have predictive power (weather forecasts are supposed to be predictive).
Modern linguistics intends to be explanatory. Therefore modern theories of languages usually found in linguistics appear to students to be “abstract” and “far from language”. As linguistics tends to explain linguistic phenomena, this perceived abstractness and remoteness from “real language” (meaning: observable language) is the completely logical result from using theoretical statements and constructs.
Exercises on Levels of theoretical adequacy