V.N. B~l~OOZ~l~OV - A. A. POLIKARI'OV 
A MECHANISM OF THE ANALYTICITY DEVELOPMENT IN A 
LANGUAGE, ELEMENTS OF THE THEORY AND THE 
COMPUTER~ EXPERIMENT 
1. From the point of view of the general cybernetic theory we 
consider language to be a device, which regulates the information streams 
in social systems. 1 On this basis it is possible to come to the conclusion 
of bilateral dependency of the language organization level (the level 
of its complexity) on (1) complexity value of the language mechanism 
input information stream, and (2) complexity value of the output in- 
formation stream which is to be produced by the language mecha- 
nism.~ 
The relation between mentioned values 
the following formula: 
L----- S, 
may be represented by 
where L is complexity (volume) of the language (in the elementary 
language units); 8 S is complexity (volume) of the input information 
stream (in elementary sense units, which are supposed to be encoded); 
k is complexity (volume) of the output information stream (in elemen- 
tary sign units, in which sense units are coded). 
In the given paper we consider certain moments of dependency 
of general level of human language complexity on some values of the 
1 The analysis of systems, in which communication processes are carried out and in 
which the necessity in communicative regulators arises, is made in our papers. It should 
be noted that our analysis is based on the ontological model of the communicative pro- 
cess, which is represented in the series of papers on systems linguistics by G. P. Melni- 
kov, published in 1967-1972 (G. P. M~L'NmOV, 1967, 1969, 1970, 1972 a). 
2 It follows from the fundamental cybernetic law of the necessary variety (W. tk. 
ASHBY, 1958), which determines the functioning of any regulator. A more detailed ana- 
lysis of the given dependency is made in the paper of A. A. PoLn~vov (in press). 
8 Every language unit consists of a model of a sign coupled with its meaning. On the 
principles of the inner organisation of language one may consult the papers ofG. P. M~'- 
Nmov (1969, 1970). 
24 V. N. BELOOZEROV- A. A. POLIKARPOV 
output information stream; internal typological reconstructions of 
language, wherein the change of its complexity is expressed, and some 
fundamental reasons of the changing in the values of the output infor- 
mation stream optimum redundancy are also treated. 
As is going to be shown below, the changing in the values of 
that parameter depends on some specific processes in the social systems 
of communication. That is why the main problem of the given paper 
refers to the sphere of socio-linguistic problematics and constitutes a 
fragment of the more complex socio-linguistic model, which is being 
worked out at present. 
2. Factors that lead to the changes in the parameter of language 
complexity are, in our opinion, connected with increasing or decreas- 
ing of specific "noise level" in language communication channels. 
Such "noise" appears, for instance, in the case of long-time and intense 
language contacts, when the community includes different language 
bearing communicants amalgamated. If the income of foreign- 
language communicants is weak and occurs from time to time, the lan- 
guage contact does not effect considerably the communication situa- 
tion (the foreign incomers accept gradually the common (original) 
language). But in the case of stable income of such communicants into 
the communicative system, and for that reason of stable low average 
value of language message comprehension in the community, the 
communicative system has to change the values of its parameters in 
order to increase the e~ciency of communication. One of the results of 
such a process is the reorganisation of language, its transition to the 
more simple organisation, which is described in the terms of linguistic 
typology as analyticisation. 
The interrelation between the phenomenon of language contacts and 
language analyticisation was already indicated more than once (L. 
T~SNI~.m~, 1939; U. Wr.INR~CH, 1953; U. W~m~ICH, 1972; G. 
C~ARDT, 1950, V. Ju. I~OSnNZW~.IG, 1972) but the mechanism and tran- 
sitional stages of the development of the process remains undiscovered 
up to the present time. The communicative and intralanguage mecha- 
nisms of analyticity developing in language become to our mind, ap- 
parent, if we consider the following hypothetical model. 
3. We take as the starting point of building the model the fact 
that the average level of mutual comprehension of the interlocutors 
decreases at the beginning of interlanguage contacts. Due to the vital 
A MECHANISM OF THE ANALITICITY DEVELOPMENT IN A LANGUAGE 25 
importance of achieving mutual comprehension the interlocutors na- 
turally try to make their speech explicit to each other. This is achieved 
by the increase of volume, complexity of messages, which as regards 
the preceding form of organisation of sign messages, can be defined 
as the increase of message redundancy. After a certain period of time 
the repeated attempts of the interlocutors to establish comlnunication 
result in discovering by the communicants a certain part of the common 
language vocabulary as the most accessible for the typical interlocutor 
under given circumstances, i.e. they result in the narrowing of the 
vocabulary in comparison with the vocabulary of the language at the 
previous stage. It turns out that the most suitable and productive 
part of the vocabulary is, as a rule, that part in which the most general, 
most universal, most frequently used words are concentrated. This very 
part of the vocabulary allows, in a general way, to "hint " each 
phenomenon of reality. Besides, mastering of the language system 
begins only with this most frequently used, most universal part, even 
at the lowest stage of one's /earning the language. 
As a result of the universalization of a definite part of meanings or 
escaping of some of them, all the system of semantic oppositions in 
the language is being reconstructed, i.e. the most peripheral oppositions 
fall out and the central ones are strengthened. This process in the lan- 
guage is well examplified in the escaping semantic oppositions be- 
tween different grammatical " words" (grammatical morphemes and 
pronoun-adverbial indicators). As a result of this, many grammatical 
indicators stop "working" and, being useless, are reduced, and words 
shortened till one syllable only. The stabilising level of an optimum re- 
dundancy of the sign messages in these conditions, beyond doubt, is 
higher than during the period, which preceded the reconstructions of 
the language system, because in order to reach the optimum (i.e. pre- 
ceding) level of understanding with the help of the most general in 
their meaning sign-" hints ", it is necessary to describe the object of 
a message from several various points. Aims, which are reached by 
a synthetic language with the help of a single but complex "hint ", 
in analytical languages are reached with the help of several, actually 
grouped, "hints ", which are not necessarily fixed beforehand in the 
speaker's memory to be used together. Meanwhile, the synthetic lan- 
guage organization presumes that the "hints" are fixed in the long- 
time memory of a typical narrator. In the latter case, the language in- 
tensively spends the memory volume and spares the operating efforts 
for message producing. On the contrary, the analytical language forces 
26 V. N. BELOOZERO'V- A. A. POLIKARPOV 
to economize the communicants' long-time memory and to exploit 
their efforts when encoding the communicative intentions, to a greater 
extent. 
Thus we may assume that the synthetic type of language organi- 
sation is preferable because of communicative causes. This preference 
arises from the fact that the circumstances accompanying this lan- 
guage type allow the fixing in the language system of many of the most 
successful means of denotating objects of reality. Analytical type of 
language organisation, on the contrary, makes the average language 
speaker fix in the language only a certain minimum of language 
experiences, it makes him improvise the actual formation of the mes- 
sage. This, apparently, does not promote the precision, optimality 
in denoting objects. 
4. We undertook a mathematical check-up of the connection be- 
tween the starting point of the process of analyticity increase (i.e. the 
fact of decrease of the level (probability) of comprehension of the sign 
messages generated with the help of the language) and one of the ter- 
minal moments of the process (the increase of the optimum redun- 
dancy level of messages built on the basis of the preserved part of the 
vocabulary). 
For the calculation, the following model of communication was 
designed. A speaker must "convey" to a listener the sense, which can 
be "transferred" in the most favourable, so-called "noiseless" case, 
with the help of the utterance containing N signs. The average proba- 
bility of comprehending the signs (with definite "interferences" pre- 
sent) is P. From the whole number of signs, which to some extent make 
a "hint" to the sense, the speaker chooses m items (m >1 N), and then 
he produces the utterance, consisting of the signs chosen before, to the 
listener. Judging by the reaction of the listener, he makes a conclusion 
on the result of the experiment, whether the sense was caught or not. 
The sense of the utterance is considered to be understood if the number 
of signs, clear for the listener, exceeds N. Then the act of communica- 
tion appears to be accomplished. If the sense of the utterance has not 
been caught, the speaker repeats the random choice of m signs from the 
same set, and build new utterances until the listener comes to understand 
the sense of the utterance at last. In this process the previous utte- 
rances are considered to have neither effect on the comprehension of 
further "transforms" nor on the new choice of the signs. 
In such a model of conversation, the time necessary for one act of 
A MECHANISM OF THE ANALITICITY DEVELOPMENT IN A LANGUAGE 27 
communication can be evaluated by the expected number of signs (K), 
produced in all the utterances taken together. If the average probability 
of every sign to be understood is P, then the probability of comprehen- 
sion of at least N signs of the whole number (m) is P, nN = ~ C,~P~(1-P) ml. 
The average number of utterances, given in order to reach comprehen- 
sion, is 
oo 
i~l 
Then the expected duration of the communication is 
This value was calculated with the help of the computer for dif- 
ferent magnitudes of m, N, P. The typical diagrams of dependency 
K - m for different P is represented in Fig. 1. 
The optimum utterance redtmdancy minimizing the expected time 
of reaching comprehension appears to exist at any volume of in- 
formation, which is to be transferred, and at any probability of compre- 
hension of the signs transferred. The diagram showing the depen- 
m-N dency of the optimum redundancy Aopt = rain ~ oll P is repre- 
sented in Fig. 2. 
The data given above show that, for the communicational situation 
characterized by the partial comprehension of signs, the more re- 
dundancy the message contains the more is the probability of mis- 
comprehension. The speaker's intuitional search for the optimum 
strategy of communication, leading to the rise of sign saturation of 
utterances, evokes other phenomena, characterising the development 
of analytic features in language; the higher the degree of these, the 
lower is the level of comprehension. 
To sum up, we must emphasise, that oll the basis of the model giv- 
en above a more complicated model might be constructed. The phe- 
nomenon of communicants' adaptation to each other in the course 
of the attempts to "transfer" the sense information must be taken into 
consideration in further research. 
The further prospects of elaboration and computerized checking 
of the model seem to be the following: the comprehension level 
during the communication can be connected with such parameters 
28 V. N. BELOOZEROV-A. A. POLIKARPOV 
40 
30 
20 
10 
0.9 
P=0.7 
P= 0.8 
P=0.5 
P~-0.6 
0.4 
P= 0.9 
+9 
÷? 
10 
Fig. 1. 
M 
! I, 
20 30 
A MECHANISM OF THE ANALITIClTY DEVELOPMENT IN A LANGUAGE 29 
Aopt 
N----2 
N>4 
i .: # I,~ ,! 
0 02 0.4 0.6 0.8 
~g. 2. 
P 
! 
1.0 
as (1) the degree of divergency between the language units stock of 
contacting languages (the more prominent is the divergency between 
contacting language the lower is the level of comprehension), (2) the 
percentage of nonprinciple language speakers within the community 
(the more numerous the representatives of the language, the lower the 
average level of comprehension), (3) the intensity of contacts between 
the representatives of different languages (the more frequent the con- 
30 V. N. BBLOOZEROV- A. A. POLIKARPOV 
tacts between different languages representatives, the lower the average 
level of comprehension in that communicative system), etc. Further, 
the definite qualitative types of languages 4 can apparently be evaluated 
by means of the definite quantitative index. 5 The efficient shift of the 
quantitative index during language evolution, breaking the definite 
ranges boundaries, seems to define the transition of language into another 
qualitative type. The quantitative characteristics of languages can be 
calculated theoretically and empirically checked up by numeral cha- 
racterisation of communicative situations (percentage of nonprinci- 
ple language representatives, the divergency degree of contacting lan- 
guages, etc.) in which the language is formed. As a result, the matrix 
of language types can be associated with the matrix of communica- 
tive situation types. 
To put it briefly, the given perspectives are those of creating a 
complex-structured sociolinguistic model with many parameters re- 
flecting the data of modern qualitative and quantitative linguistic ty- 
pology as well as quantitative and qualitative sociological typology. 
One may expect that the successive detection, the quantification of 
the values of linguistic and sociological parameters, the revealing of 
their relationships will result in building up the explanatory sociolin- 
guistics theory. This theory can find out the correlation between the 
given states of a language and the given states of society employing it. 
This theory is very likely to predict with a high degree of certainty 
the future development of languages and the human language in gene- 
ral with a view of perspectives of humanity's social development. It 
is possible, having a carefully worked out theory (that is, knowing prin- 
ciple correlations between language and society), to simulate with the 
help of a computer the evolution of the given language as the function 
of the facts given by the history of the people employing this language, 
of the communicative system history, in which the language is formed, 
Turn to one of the most full, in our opinion, classifications of language types, 
represented in the paper of G. P. M~L'~rmov (1972 b). 
5 In the first line we mean the criteria of evaluation of real object complexity based 
on the categorical system of information theory. These problems are being considered, 
for example, in the paper of S. M. DAt~corF and H. QUASTL~ (1953). 

References

W. IL. ASHBY, An introduaion to cyber- 
netics, London 3rd imp., 1958. 

V. N. BELOOZEROV, O fonologi\[eskich 
edinicach v kanalach massovoj kommu- 
nikacii, in Semiotika sredstv massovoj 
kommunikacii, Materialy simpoziuma 
18-20 ijunja 1973 g., MGU. 

S. M. DaNCO~, H. QUASTL~, The Infor- 
mation Content and Error Rate of Living 
Things, in Information Theory in Biology, 
Urbana 1953. 

G. P. M~t'NIKOV, Sistemnaja lingvistika i 
ee otno~enle k strukturnoj, in Problemy 
jazykoznanija, Moskva 1967. 

G. P. M~Z'NIKOV, Sistemnyj analis gelo- 
vdeskich jazykov v svjazi s problemoj 
modelirovanija jazykovych processov i 
postroenija universal'nych mMinnych 
jazykov, Ot~ety LASUP, S MGPI ira. 

V. I. LENINA, Sektor ob~ej teorii 
sistem, za 1968 i 1969 gg., Moskva. 

G. P. M~'NIKOV, Rdevaja dejatel'nost'pri 
bilingvizme, in IV seminar Fo vopro- 
sam teorii i praktiki perevoda nau~noj 
i techni\[eskoj literatury, 2-4 nojabra 
1970, Tezisy dokladov, Moskwa-Tbi- 
lisi 1970. 

G. P. M~'NmOV, Sistemnyj podchod v lin- 
gvistike, in <~ Sistemnye issledovanija ~, 
E{egodnik 1972, Moskva 1972(a). 

G. P. MBL'NIKOV, Determinantnaja klassi- 
fikacija jazykov i jazyki bantu, in 
<~Africana. Africanskij etnografi~eskij 
sbornik,>, IX, Leningrad 1972 (b). 

A. A. POUKARPOV, K objasneniju oso- 
bennostej sintaksisa jazyka nauki kak 
jazyka massovoj kommunikacii, in Se- 
mioti\[eskie problemy jazykov nauki, ter- 
minologii i informatiki, Materialy sim- 
poziuma, t. 2, MGU, 1971. 

A. A. POUKARPOV, O meste sistemy masso- 
voj kommunikacii /MK/ v tipoIogii kom- 
munikativnych sistem, in Semiotika 
sredstv massovoj kommunikacii, Mate- 
rialy simpoziuma 18-20 ijunja 1973 g., 
MGU. 

V. Ju. R.osrNzw~m, Jazykovye kontakty, 
Leningrad 1972. 

G. ~UCI-IARDT, Izbrannye stat'ji po jazy- 
koznaniju, Moskva 1950. 

L. T~SNmRE, Phonologie et melange des 
langues, in ~, Travaux du Cercle linguis- 
tique de Prague,~, VIII (1939). 

U. W~mlu~ICH. Languages in Contact, 
New York 1953. 

U. WEINREICH, Odnojazycie i mnogoja- 
zy~ie, in , Novoe v lingvistike ~), vyp. 
VI, Moskva 1972. 
