linkedin facebook twitter rss

14 May Analyzing Syntax

Analyzing Syntax

Construction BlueprintSyntactic analysis requires some access to morphological features or characteristics of the words in the sentence being analyzed. Tense, gender and number must be accessible in order to establish agreement. There is clearly some interdependency between morphology and syntax. Morphological phenomena can change categories and roles of words. Roles and categories are the essential components of syntactic analysis.

Establishing syntactic agreement requires knowledge of morphological features. This is not two-way dependency. It is one way in that syntactic phenomena do not affect morphological properties or their analysis. This strengthens the theoretical basis of describing morphology as surface-level and assigning the other linguistic strata in order of depth or proximity to conceptual meaning. This approach sometimes places phonology and prosody closer to the symbolic surface than morphology, though prosody may touch both superficial and deep language phenomena.

Understanding Context Cross-Reference
Click on these Links to other posts and glossary/bibliography references

  

Section 6 #40

Language Section Icon

 

Table of Context

Pigeon HolesThe morphological phenomena that affect semantics may be worth retaining for construction of the relational representation. So, perhaps there should be special pigeon holes in parse trees for retaining pertinent morphological data that could be useful in lower levels of analysis. This will certainly compound the complexity of parse trees, but if it captures linguistic phenomena, perhaps the added complexity is justified. Some grammar formalisms see only two linguistic levels: syntax and semantics. Under that view, morphology could be treated as part and parcel of syntax. If that view is taken, though, morphological structures cannot be treated lightly if robust interpretation is the goal.

Syntactic Patterns

The taxonomy of the grammar includes sentences, structure markers, functions and attributes in a hierarchy. The previous section included a list of some morphological attributes. The following (incomplete) lists of examples structure markers and functions apply to the syntax of any language and represent the types of features 3-DG recognizes:

STRUCTURE MARKERS

S = Sentence NP = Noun Phrase
VP = Verb Phrase ES = Embedded Sentence
A = Adjunct PP = Prepositional Phrase
TP = Tensed Verb Phrase RC = Relative Clause

FUNCTIONS

N = Noun PR = Pronoun RP = Relative Pronoun
PN = Proper Noun Q = Question AJ = Adjective
AV = Adverb P = Preposition V = Verb
M = Modal BV = Be Verb C = Connective
D = Determiner B = Binder PT = Participle
X = Auxiliary

The potential list of attributes is virtually unlimited, so attributes are handled at the pragmatic level using conceptual graph processing as proposed by John Sowa. By storing the syntax rules, identifying the patterns that identify these structure and function markers in an ontology, bound to words as needed, we can sue a single model for capturing phenomena at each stratum of language, and across the boundaries from one stratum to another. Phonology and morphology are closer to the symbolic surface of language understanding. Semantics and pragmatics are “deep” strata closer to the conceptual foundations of knowledge. As we move deeper into the stack of strata, I’ll attempt to show how the same idea of pattern-based rules can deliver robust means of fully automated interpretation.

Click below to look in each Understanding Context section


 

 

Comments are closed.