International Journal of Pedagogics
199
https://theusajournals.com/index.php/ijp
VOLUME
Vol.05 Issue06 2025
PAGE NO.
199-202
10.37547/ijp/Volume05Issue06-55
Definition Of Basic Concepts and Typology Of Linguistic
Corpusa
Akhmedjanova Sitora Djurakhodjayevna
Lecturer, Department of Russian Language and Literature Bukhara State University Bukhara, Uzbekistan
Received:
14 April 2025;
Accepted:
15 May 2025;
Published:
19 June 2025
Abstract:
Corpus linguistics is considered one of the most promising and progressive areas in language study. The
relevance of this article lies in the enormous potential of linguistic corpora, which has not yet been fully recognized
by the scientific community, at least because the text
–
the main object of corpus linguistics
–
in its various forms
of implementation is one of the main components of the language system and the mental and speech activity of a
modern native speaker. The article reveals the concept of "corpus", provides a classification of text corpora,
describes in detail each group of text corpora, provides criteria for linguistic corpora, explains the concept of
"markup", and examines the basic concepts of corpus linguistics, methods and areas of its application. The
advantage of text corpora in linguistic research is described. The article also analyzes the emergence and
development of corpus linguistics, provides a typology of corpora, and describes each type of corpus separately.
Keywords:
Corpus, linguistics, corpus linguistics, text, semantic, anaphoric, graphematic, representativeness,
conceptual apparatus, variation.
Introduction
Each study conducted by a linguist should be aimed at
at least at certain stages of activity:
1. Selection of provisions and a basis for categorization
of the objects under study.
2. Division of objects into categories in accordance with
this basis.
3. Understanding and interpreting the results of
dividing objects into categories, interpreting the
grounds for such division.
At the same time, the first stage of this activity
presupposes the existence of objects under study, i.e.
the acquisition of practical data for the creation of a
theory at the final stage.
Nowadays, corpus linguistics in the preparation and
analysis of empirical data is becoming widespread,
thanks to the intensive growth of information
technology.
Methods. Corpus linguistics first became known in the
1960s. Texts were formed mainly on the basis of the
English language, but soon corpora began to appear (in
corpus linguistics, the plural form "corpora" is used.
See: Explanatory Dictionary of the Russian Language
edited by D. N. Ushakov: Corpus, pl. corpora) based on
material from other languages.
At the same time, at Brown University in the USA,
scientists W.N. Francis and G. Kuchera compiled the
first corpus of texts on an electronic medium, which
consisted of 1 million word usages (500 texts with 2,000
words each). It also had appendices in the form of an
index of word frequency in alphabetical order and
certain statistical data.
A corpus is a collection of texts in one or more
languages that are related by certain characteristics. In
their work, L. Lemnitzer and H. Zinsmeister gave the
following definition of a corpus: "A corpus is a
collection of written or oral statements. Corpus data is
usually digitized, i.e. often stored on computers and
machine-readable." [2, 7].
At the same time, the constituent elements of the
corpus
–
texts
–
are made up of materials, metadata
that these materials represent, and linguistic
generalizations that these materials organize.
As a particular section of linguistics, corpus linguistics
was finally formed at the end of the 20th century.
Corpus linguistics as a separate section of linguistics
was finally formed in the first half of the 90s of the 20th
century. At the same time, the conceptual apparatus
began to take shape [5, 37]. In particular, J. Sinclair
International Journal of Pedagogics
200
https://theusajournals.com/index.php/ijp
International Journal of Pedagogics (ISSN: 2771-2281)
defines the concept of "corpus" as follows: "a set of
naturally occurring language texts selected to
characterize the state of diversity of a language" [3,
171].
Here is one of the key provisions for selecting texts for
compiling a corpus - we are talking about unfinished
texts, in other words, the language sounds in the form
in which it was expressed (oral or written speech).
Moreover, the corpus does not offer r
eal “templates”
and “positions” for the correct organization of a
message, but the maximum possible number of
“variations” of the language, although some of them
are not located in the center of the language system.
Further, the concept of “corpus” is incr
easingly
clarified: “A corpus is a collection of texts intended for
some purpose, usually educational or research. [...]A
corpus is not something a speaker says or knows, but
something created by a researcher. It is a record of the
performance of, usually, many different users,
intended to be studied so that we can draw conclusions
about typical language use. Because it provides
methods for observing patterns of the kind that have
long been noticed by literary critics but not revealed
empirically, computer-aided exploration of large
corpora may perhaps offer a way out of the paradoxes
of dualism" [4, 239-240]. (Our translation)
We assume that a more or less complete formulation
of the concept of "corpus" can be found in the works of
V.P. Zakharov. The linguist describes the corpus as a
large, electronically represented, organized and
planned, philologically impressive conglomerate of
linguistic data, designed to resolve specific linguistic
issues and tasks.
This formulation can be qualified as “activity
-based
”,
which, by and large, explains the linguistic tendency of
organized text arrays.
Results and discussion. As a result, in any of the
described formulations of the concept of "corpus" the
following is noted:
1) a large number of texts must be presented in
electronic form (on the Internet or on any medium);
2) the language material must be distributed for
consideration for linguistic purposes;
3) following the review, there should be a method for
variously dividing the obtained language data (by topic,
genre, year of creation, etc.). Considering the first, the
possibility of constant access to texts in electronic form
was noted. A huge number of text corpora can be
classified into three significant groups:
1. Freely available;
2. Partially available;
3. Commercial.
The first group includes a relatively small number of
text corpora available today. The National Corpus of
the Russian Language, which contains more than 500
million words, is considered to be quite substantial.
The next group includes the majority of available
corpora; however, for solving certain linguistic
problems, this partial access is considered quite
sufficient. For example, in the British National Corpus,
the query results are only up to 50 arbitrary examples,
and most of the functions of the search interface, which
is only provided together with the full (and paid)
version of the corpus, are missing.
There is, however, a non-commercial version of this
corpus that is made available through a simple
registration process. This version offers searchable
texts from 1980-1993, with around 100 million words.
The third group includes, for example, the British
National Corpus, which has a free one-month pre-
subscription option to gain access to Collins Wordbanks
Online, which contains about 533 million words, before
you can buy the commercial version of the corpus.
Another significant criterion of a linguistic corpus of
texts is the presence or absence of markup, because
the presence of a simple conglomerate of texts is not
enough to solve linguistic questions and problems.
Markup is the assignment of special marks to texts and
their elements: external, extralinguistic, systemic and
strictly linguistic, which describe various parameters of
text elements. Metamarkup includes not only
information about the text, but also data about the
author. Let's study strictly linguistic types of markup.
One can start with marking the parts of speech that are
frequently encountered in existing corpora, but at the
same time not only morphological indicators are taken
into account, but also grammatical ones.
The marking of parts of speech is carried out with the
participation of special programs of automated
morphoanalysis. For example, in a small part of the
National Corpus of the Russian language (6 million
word
usages)
manual
elimination
of
morphohomonymy and auxiliary correction of the
results of the process of the program of automatic
morphoanalysis were carried out [6, 86].
In the Mannheim Corpus of the German language, the
marking of parts of speech is present mostly in the sub-
corpora of journalistic texts. Among other types of
markup, it is especially necessary to pay attention to
syntactic markup, which is not presented in the entire
conglomerate of the corpus, but only in a small part of
it, because this type of markup, which involves
determining the syntactic structure for any sentence, is
done almost manually and requires significant time
expenditures. Also, the corpus contains other types of
markup, for example, semantic, prosodic, anaphoric,
graphematic, etc. This largely helps to simplify the
procedure of natural data collection by the researcher,
taking into account the correctly specified search
conditions.
International Journal of Pedagogics
201
https://theusajournals.com/index.php/ijp
International Journal of Pedagogics (ISSN: 2771-2281)
However, in order for the developed corpus of texts to
satisfy various kinds of linguistic tasks that a linguist
faces, it must, in turn, have at least two more
indicators.
First of all, we mean the representativeness of the text
corpus. Kibrik A.E., Brykina M.M., Leontyev A.P. and
Khitrov A.N. believe that representativeness can be
assessed by transforming the "relative frequency" of
the fact under study with the growth of the "sample".
If the “relative frequency” of a fact does not change
frequently with the increase in “each subsequent
fragment of text”, then this means that “the corp
us as
a whole is representative”. At the same time, although
it is observed that it is inadmissible to determine the
links with statistics in this formulation of
representativeness, it is emphasized that this
requirement is mandatory, but still incomplete for
establishing the representativeness of the text corpus.
Basically,
the
issue
of
establishing
the
representativeness of different text corpora is still
considered relevant, but, admittedly, it has not been
sufficiently developed. Only representativeness
transforms the usual complex of different texts into a
text corpus suitable for carrying out linguistic research.
At the same time, human speech activity is so diverse
that it is almost impossible to actually convey all the
existing variations of language mentioned above. For
this reason, the question of the representativeness of a
text corpus is considered more of a question of the
impartiality of any scientific research. In this case, it is
advisable to rely on the common sense of the
researcher himself, when we mean a user corpus
(developed by the researcher himself in accordance
with the goals of his research), or a group of
researchers, when we mean the creation of a corpus
that requires a large scale of linguistic phenomena,
styles, genres, etc. (for example, a national corpus of a
specific language).
An important condition when designating a case is also
its ease of use, in other words, the case must be
equipped with a specialized search system, which must
be (ideally) quite understandable to a sufficient degree
and easy to operate. The operation of the National
Corpus of the Russian Language or the British National
Corpus (English Language Bank) presents significant
problems, which cannot be said about the search
system of the Mannheim German CorpusWe believe
that the corpus should not take up a lot of time, which
is necessary to search for a certain phenomenon, and
should not offer a tricky search methodology, since
studying its basic points requires from the researcher
in some cases purely technical and mathematical
knowledge.
Corpus and its types
In some cases it is very difficult to navigate among the
existing diversity of research corpora, because the
goals and tasks set before the linguist are often
identified in general, but in specific fields and areas
they differ. The initial stage carried out by the
researcher in studying the "objects" under study is the
correct choice of the appropriate corpus. The entire
diversity of existing corpora is determined by the
diversity of "research and practical tasks for the
solution of which they are created" [7, 12].
1. Oral, written, mixed.
Oral corpus is a systematized complex of speech
fragments, equipped with software capabilities for
accessing them [1, 71-72]. Oral corpora first began to
function in the 80s of the 20th century on the basis of
American English. Then special coordination centers
appeared that collected, stored, distributed and
created oral corpora. For example, LDC (Linguistic Data
Consorcium), CSLU (Center for Spoken Language
Understanding), ELRA (European Language Resources
Association).
Most of the existing corpora are written or mixed (for
example, the accessible part of the Mannheim Corpus
of the German language), yet the part of linguistically
marked oral texts even in mixed corpora is quite small
relative to the entire conglomerate of the corpus (very
often these are national corpora of a certain language,
for example: Russian, English).
2. Monolingual
–
bilingual/multilingual.
There are two groups of monolingual corpora:
–
corpora covering the entire language,
–
corpora covering only the language for specific
purposes.
For example, the Corpus of Early English Medical
Writing (CEEM) is a corpus of medical texts in English
from 1375 to 1750, the volume of which is
approximately 1.5 million words. It contains theoretical
works, reference books, and poetic texts on medical
topics.
In bilingual and multilingual corpora, texts can be
presented either comparable or parallel. For example,
in 1992, the European Corpus Initiative (ECI) was
established as an international organization that is
engaged in compiling a large multilingual corpus for
research purposes. The present comparable corpus
contains not only texts of European languages, but also
texts in Russian, Turkish, Chinese, Japanese and many
others. Their volume is more than 98 million words.
This type of corpus is considered commercial. Corpora
of parallel texts are intended, first of all, for
comparative analysis of texts in the direction of
"original - translated" for teaching methods,
techniques and methods of translation. For example,
the European Parliament Proceedings Parallel Corpus
1996-2011, which presents parallel texts of the session
of the European Parliament in different European
International Journal of Pedagogics
202
https://theusajournals.com/index.php/ijp
International Journal of Pedagogics (ISSN: 2771-2281)
languages with translation into English.
3. Synchronous
–
diachronous. Synchronous corpora
provide a representation of text data for studying the
systemic state of a language in a specific period of time.
Thus, the non-commercial version of the British
National Corpus only contains texts from the period
from 1980 to 1993.
To study the historical development of a certain
linguistic phenomenon or the entire linguistic system in
general, there are diachronic corpora. For example, the
Thesaurus
Indogermanischer
Text-
und
Sprachmaterialien, which presents Indo-Germanic
texts from different eras.
4. Unmarked
–
marked.
An unmarked corpus is a conglomerate of texts
containing a specific number of mentions of the
required component. At the same time, the search
results offered in unmarked corpora can be used in
linguistic research, but only from a statistical point of
view.
Annotated corpora (morphologically, syntactically,
etc.) are considered multifunctional, as they provide
many more opportunities for linguistic analysis.
CONCLUSIONS
So, a corpus is a representative conglomerate of
unedited texts, presented in electronic form, usually
marked up for linguistic analysis, equipped with a
relatively easy-to-use search system, which represents
as many language variants as possible.
During the years of the emergence of corpus linguistics,
the problems of computerization in this area were not
identified, and “researchers pointed to the possibility”
of ignoring the variability of language, namely
“territorial, social, age, gender,” etc.
linguistic
distinctions. Nowadays, by ignoring it, we deliberately
limit ourselves with various frameworks when studying
texts of a specific language, which calls into question
the objectivity of this kind of research. With the advent
of electronic corpora, the diversity of forms of language
existence has become more indicative, the means and
possibilities for studying language data have increased.
The modern linguistic corpus contains hundreds of
millions of word usages, and the fact that, thanks to the
electronic corpus, the results of word usage examples
can be obtained incredibly quickly makes the task of
linguists much easier. The typology of corpora shown,
without claiming to be large-scale, presents us with a
real diversity of text corpora and allows us to navigate
it for further scientific research.
REFERENCES
Kodirov A. Linguistic and cognitive representation of
the concept" HOPE" IN RUSSIAN //International
Bulletin of Engineering and Technology.
–
2023.
–
Т. 3.
–
№. 5. –
С. 106
-108.
Lemnitzer L., Zinsmeister H. Korpuslinguistik: Eine
Einführung. Tübingen, 2006.
–
С. 7.
Sinclair J. McH. Corpus, Concordance, Collocation.
Describing English language. Oxford: Oxford University
Press.
–
1991.
–
С. 171.
Stubbs M. Words and phrases: corpus studies of lexical
semantics. Oxford, 2001.
–
P. 239-240.
Ахмеджанова С. Д. РУССКИЙ И УЗБЕКСКИЙ ЯЗЫКИ В
СОПОСТАВИТЕЛЬНОМ АСПЕКТЕ //Ответственный
редактор. –
2023.
–
С. 37.
Ахмеджанова,
С.
.
(2023).
СПЕЦИФИКА
ЛАКУНАРНОСТИ
В
РУССКОЙ
ЛИНГВИСТИКЕ.
Евразийский журнал социальных наук, философии и
культуры, 3(3), 84–88. извлечено от https://in
-
academy.uz/index.php/ejsspc/article/view/11205
Коршунова А. В. Зоонимы Как Слова
-
Компоненты В
Русских Фразеологизмах //Miasto Przyszłości. –
2024.
–
Т. 47. –
С. 673
-675.
Кривнова О. Ф. Области применения речевых
корпусов и опыт их разработки // Тр. XVIII Сессии
Российского акустического общества РАО. Таганрог,
2006.
