4 Complex operational structures
With a few exceptions (clichés, idioms, quotations, borrowed phrases,
fragments seared into one’s memory) information conveyed in
communication is every time created from picked ad hoc meanings even
though this is made easier by the fact that for each of the words there
are many readily available examples of its use in combination with other
words. Complex operational structures control this process together with
the other well-known factors, such as grammar usage, semantics or
stylistic and pragmatic factors. According to our research of the
Russian language, complex structures cover the main ways of grammatical
sentence formation, which we described with 37 groups of operational
structures. The list of simple structures of a language and the modes of
combining them to create complex structures form a closed system, which
represents language distinctive characteristic. Relying on this
research, we think that the analysis with operational strictures is
applicable to other languages too. Our supposition is based on the
premise that the noun, verb and interjection are, according to typology
data, universal from isolating to synthetic languages and so are their
structures while the other operational structures are their derivatives.
There are two main ways of forming complex operational structures: by
means of single (noncoupled) quantitative or elective operations,
establishing a corresponding relationship between the operational
structure of an information unit and a state of the operation, and by
connecting simple structures. Although the single operations are carried
out by non-semantic information, the necessity of their performing is
mostly expressed by signs and therefore they are detectable in speech.
One of the single elective operations is negation. The negationEx → Ex → of avi x means that it gets an elective
state alone (without semantic information) that substitutes its
derivative elective state. Coupled with the remaining quantitative
state, the elective state of the operation characterizes a new unitx → with operational information , i.e.,
an information unit in which instead of semantic information only its
quantity is noted. Because of the asymmetric relation of the elective
states, the quantitative states of the vi and
of the new unit are equivalent: Qx = Qx →(see the preceding section). In the sentence Trust funds are not
admissible , for example, the lack of information of the elective state
introduced by negation signifies cancellation of “are admissible” and
introduction of operational information with a quantitative state
equivalent to it. The presence of operational information is confirmed
by the possibility of introducing the same quantity of semantic
information (e.g., “are prohibited”) instead of it without violating
the grammatical correctness of the sentence.
One of the single quantitative operations is hyponymization .
There is no special symbol to express it in English, Russian and other
European languages (cf. izafat in Persion or Turkic languages), but it
plays an important role in forming complex structures and corresponding
phrases.
The hyponymization of an information unit x is determined by the
relation Qx < Qx <,
meaning that x gets an introduced quantitative state which is
including with respect to its derivative quantitative state and
characterizes missing information. Its hyponymization introduces a
quantitative state without semantic information while the elective state
remains the same (Ex ⊥ Ex <).
So the hyponymized unit gets operational information with a quantitative
state superior to the one of the semantic information. We will say that
it gets an operational information increment .
It is important to emphasize that there is no question of real
information resulting the hyponymization, it’s just about operational,
non-semantic information. Let us clarify our view in contrast to the
available researches of quantitative relations in language. John Lyons
(1977: 45) notes: “Signal-information content, as measured by the
mathematical theory of communication, has frequently been referred to as
surprise-value; and it is this aspect of the theory, if any, which links
the two senses of ’information’, which we are distinguishing by means of
the terms ’signal-information’ and ’semantic information’. … ”Man
bites dog”, as they say, is a more significant item of news than ”Dog
bites man””. In his Axiomatic Semantics , which “is a theory for
the description of the wholly fixed-conventional information values of
linguistic signs (in their capacity of being paradigmatic entities in
grammar as well as fully fledged signs)” (p. XXVII), Sandor G.J. Hervey
(1979) writes: “The relations of hypero-hyponymy, synonymy and paronymy
are defined in ”axiomatic semantics” in terms of set-theoretical
relations holding between classes of entities — denotata (131)”. And
further: “Thus, giving the information that a given sign belongs to one
another of these classes proves to be an economical way of
distinguishing the ”meaning” of that sign from that of many other
signs” (146). Then he inquires (see, for example, p. 162, 164) whether
the word is used as a hyponym of another word. So, information quantity
can be viewed either in regard to preservation of conveyed signal or as
to degree of message novelty or else when comparing the meanings of
different words. The relations between word meanings are numerous and
vary from one part of speech to another. As Gabriella Vigliocco and
David P. Vinson (2005) notice: “Huttenlocher and Lui proposed that
these two content domains are differentially organized: words referring
to objects would be organized hierarchically, whereas words referring to
events would have a matrix-like organization without well-defined levels
of structure” (19). The authors cite Miller, G. A. & Fellbaum, C.
view: “In Wordnet, ʻnouns, adjectives and verbs each have their own
semantic relations and their own organisation determined by the role
they must play in the construction of linguistic messagesʼ (p.197).
These relations and organisation are constructed by hand based on the
relations that are believed to be relevant within a given class of
words. For nouns the most important roles are typically played by
relations including synonymy, hierarchical relations and part-whole
relations. For verbs, instead, dominant are troponymy (hierarchical
relations related to specificity in manner), entailment, causation and
antonymy. Some evidence compatible with a different role of relations
such as cohyponymy and antonymy for nouns and verbs comes from
spontaneously occurring semantic substitution errors” (22). In
contrast, we do not explore relationships that exist in language
understood as Langue and are merely found in speech, but we do explore
relationships that are formed only in speech.
Hyponimization can be caused by the presence of a subordinate word
information. In this case its derivative state includes that of the
principal word information and is equivalent to the state of its
operational information. We will say that the principal word information
is a hypernymic base of its increment with operational information. Only
binary relations exist between operational states. Accordingly,
operational information cannot be measured by a numerical value like the
parameters of material objects.
In the sentence
(5) With a higher soil pH the element phosphorus and others
becomes more available for absorption by plant roots
(borrowed from Meyer 1992: 20) it’s the noun informationni element that serves as a hypernymic
base for the ni phosphorus and is
hyponymized. In case of two information units which structures form a
complex structure, their operational relation characterizes a syntactic
link between them. In the phrases an oak tree, a fruit tree, a
forest tree the ni tree is hyponymized.
In all these phrases the subordinate word information derivative state
includes that of the principal word information and the subordinate word
information has the principal word information (element andtree , respectively) as its hypernymic base. In the phrasegood boy the adjective information
(adji ) Δn (which, unlikeni , does not have its own hypernymic base)
serves as the ni ’s increment equivalent:n Δn . The adji ’s
hypernymic base is determined by the ni . An
increment information equivalent has an influence on the hypernymic baseni and can leave it without change (e.g.,a white mouse : the base ʻmouseʼ) or provoke its change and even
its replacement by the increment information (a mouse
photographed , drawn or invented by a writer : The base is
ʻphotoʼ, ʻdrawingʼ, ʻimageʼ, but not ʻmouse as a living creatureʼ).
Accordingly, the object recognized by ni can
change with the base: a fake gun (the example borrowed from
Lakoff 1982: 37) is a forgery of the gun (resembling it but not a gun);a wooden leg is a device that compensates the lost leg (but is
not a human leg). Hyponymization takes place if theni remains the base of the noun phrase
information, i.e., if the ni remains intact
like in white mouse , or at least if the noun phrase information
can’t be determined without the initial ni ,
like in fake gun , wooden leg – cf.: a hot dog ,
where dog can’t be determined as a hypernymic base and therefore
hyponymization is not determined either.
So, the noncoupled single operations (we do not consider the rest of
them in this article) establish a quantitative or elective relationship
between the operational structure of an information unit and a single
state by creating a new unit based on the substitution (negation) or
inclusion (hyponymization) of states.
The other way of forming complex operational structures is the
connection of simple ones.
One of complex operational structures formed by unification of simple
structures characterizes the information of common noun-verb sentences.
The ni ’s elective state is unified, first, with
the verb information initial elective state and then with its derivative
elective state. It means that the ni has been
substituted while the verb information (vi )
remains the same, denoting the object substitution (by itself, in
particular). We will call a complexvi -ni structure
forming the sentence information integration :
(6) En ⊥ e 0v ,
En ⊥ Ev :
(En ⊥ e 0v ) → (En ⊥ Ev );
Qn ≠ Qv ,
which implies that the ni ’s elective state
unified with the vi ’s substituted state is
substituted by the one unified with the vi ’s
substituting state and which results so in a sequence of the unified
states. Whereas in the case of hyponymization, one of the information
units was principal, the vi andni forming sentences are equitable. So, each in
turn serves as a hypernymic base for the other (we have denoted it by
their non-equivalence: Qn ≠ Qv ). Thevi -ni integration is
described by the same structural formula independently of their
information (see the sentences [1]).
Due to the non-equivalence of the units’ including states, theni in the sentence A ball rolls , for
example, gets an increment information equivalent asvi :Qn < = Qnv(ʻa rolling ballʼ) and the vi gets an increment
information equivalent as ni :Qv < = Qvn(ʻrolls like a ballʼ). An increment information adopts the structure of
the hyponymized item, namely, (v ) innv adopts the nominal structure ofadji :Q (v ) = Q Δn , and
(n ) in vn , the
structure of a parametric (gradable) adverb informationadvi :Q (n ) = Q Δv . The resulting
integral information recognizes a situation which represents the process
as limited by the properties of the object, and the object, as passing
succeeding moments of the lasting process. So, the increment information
of ni with vi causes
its increment also with nominal informationn Δn (v ): The nominal information
constituting the ni ’s increment recognizes
qualities lent to the object by the process (i.e.,ni recognizes an object identified in process).
Cf.: ʻa man standingʼ (The man stands ) and ʻa man sittingʼ
(The man sits ), where different attributes of the man are
recognized – his posture, holding on different body parts, different
muscle tension, readiness for mobility, etc. The increment information
of vi with ni causes
its increment with verbal informationv Δv (n ): The verbal information
constituting the vi ’s increment recognizes
qualities lent to the process by the object (how does the object do
it?). Cf.: ʻstands like a manʼ (One player stands in the middle of
the room ) and ʻstands like a treeʼ (The tree stands there ),
where different attributes of the process are recognized – staying legs
on the floor or roots in the ground.