The Distributional method

Distributional method describes environments of linguistic units (morphemes, words, phrases), representing them symbolically: N – noun, A– adjective, T – article,V – verb, D – adverb.

Distribution is the total of all the environments in which an element can occur; all occurrences can be symbolized. The distribution of the verb to make can be presented in the following way: He makes me do it (NVNVtoN), He makes up for smth (Nvup for N), I make a present (NVTN), I make a bed (NVTN), etc.

In each particular environment an element develops peculiar qualities. For example a morpheme modifies its qualities getting into various environments, it’s variants are called allomorphs. Phenomena, genii, teeth, children, tables, cats, brushes are concrete allomorphic representationss of the morpheme of plurality, which is abstract and can be likened to the unobservable deep structure.

Using the essentials of Distributional method scholars distinguish morphemes standing in contrastive distribution jump:: jumped, uncontrastive distribution learnt::learned and complementary distribution bed:: beds, house::houses,child:: children, etc. (plural).

In syntax they compare distributional formulas or coded structures, i.e. the distribution of a verb, for instance, can be represented by a set of distributional formulas (symbolic representations of concrete structures).

There are a number of postulates here to be observed:

if two or more distributional formulas are identical their meanings are identical;

if two or more distributional formulas are different their meanings are different.

But in actual usage these postulates can be disproved. This method turns out to be too formalized, as one and the same distributional formula conceals different meanings. Semantically different structures I make a bed, I make a basket, I make a road, I make a promise are symbolized by one and the same formula NVTN. The Distributional method doesn’t reveal any difference between the structures Napoleon’s victory and Napoleon’s defeat, though we feel intuitively that they are semantically different. Semantically different structures John’s eager to please (NV_beAV_to), John’s easy to please (NV_beAV_to) are identically coded.This method is mechanistic, form-oriented, it ignores meanings. It is ineffective when analyzing polysemy, homonymy, ambiguity, implicit syntactic relations, syncretism. The syntactico-semantic syncretism of the element greyly in the sentence The rain falls greyly remains unrevealed.

The IC Method (method of immediate constituents)

This method was elaborated by the head of American Descriptive Linguistics Leonard Bloomfield.The IC method aims at describing any complex form ranging from long sentences to multi-element words in terms of their constituents. The form is divided into two parts, the remaining parts are also divided into parts until ultimate indivisible pieces are arrived at:

un][gent]le][man][ly. The main requirement on the morphological level is that ultimate constituents (or at least one of them) should be recognizable as morphemes: book||let; let is a diminutive suffix. The word ham||let (a small village) can also be divided into 2 parts, though we do not know what ham here means.

Proceeding from the intuition of a native speaker, L.Bloomfield analyzed the sentence Poor John ran away in the following way Po or ][ John// ran ][ away.

The main requirement of the method on the syntactical level is that ultimate constituents should be words.

There are several varieties of diagramming of this analysis. We can represent the candelabra division (1) and the derivation tree division (2).

Poor John ran away (1) (candelabra diagram)

└---------┘ └------┘

└--------┘

S

/ \

NP VP (2) This is a derivation tree division.

/ \ / \

A N V D

Poor John runs away

S

/ \

NP VP

/ \ / \

T N V D

The rain falls greyly

The word greyly semantically refers to the noun rain, but the diagram doesn’t show it..

The method shows the derivation of a sentence, but it’s formalized, mechanistic, it disregards meanings and can’t be employed to analyze polysemy, homonymy, ambiguity, implicit syntactic relations, syncretism.

Дата добавления: 2014-01-07