LITHUANIAN UNIVERSITY OF EDUCATIONAL SCIENCES
FACULTY OF PHILOLOGY
DEPARTMENT OF ENGLISH PHILOLOGY
JULIJA RITČIK
A CORPUS-BASED ANALYSIS OF THAT-DELETION IN
COMPLEMENT CLAUSES AFTER THE VERBS OF SAYING,
THINKING AND DISCOVERING IN ENGLISH AND
LITHUANIAN
MA Paper
Academic Advisor: Assoc. Prof. Dr. Judita Giparaitė
Vilnius, 2014
1
LIETUVOS EDUKOLOGIJOS UNIVERSITETAS
FILOLOGIJOS FAKULTETAS
ANGLŲ FILOLOGIJOS KATEDRA
TEKSTYNAIS PAREMTA JUNGTUKO KAD PRALEIDIMO
ANALIZĖ PRIJUNGIAMUOSIUOSE AIŠKINAMUOSIUOSE
SAKINIUOSE PO VEIKSMAŽODŽIŲ, REIŠKIANČIŲ
KALBĖJIMĄ, MĄSTYMĄ IR SUVOKIMĄ, ANGLŲ IR
LIETUVIŲ KALBOJE
Magistro darbas
Humanitariniai mokslai, filologija (04H)
Magistro darbo autorė Julija Ritčik
Patvirtinu, kad darbas atliktas
savarankiškai, naudojant tik darbe
nurodytus šaltinius
___________________________
(Parašas, data)
Vadovas doc. dr. Judita Giparaitė
___________________________
(Parašas, data)
2
CONTENTS
ABSTRACT..............................................................................................................................5
THE LIST OF ABBREVIATIONS..........................................................................................6
INTRODUCTION....................................................................................................................7
1. CORPUS-BASED APPROACH TO LINGUISTIC STUDIES.........................................10
1.1 General overview of corpus-based linguisic studies..............................................10
1.1.1Corpus linguistics as means of studying the language.............................10
1.1.2 Corpus-based studies of grammar............................................................11
1.1.3 Qualitative approach in corpus-based studies.........................................12
1.2. Corpus-based linguistic studies in Lithuania.........................................................13
2.VERB COMPLEMENT CLAUSES IN ENGLISH AND LITHUANIAN.........................15
2.1. Complement clauses in English.............................................................................15
2.1.1 Definitions and terminology.....................................................................15
2.1.2 Properties of complement that-clauses.....................................................15
2.1.3 Verbs controlling complement that-clauses.............................................18
2.1.4 Properties of the complementizer that..................................................19
2.2 Complement clauses in Lithuanian …..................................................................21
2.2.1 Definitions and terminology..................................................................21
2.2.2 Properties of complement that-clauses..................................................22
2.2.3 Verbs controlling complement that-clauses...........................................26
2.2.4 Properties of the complementizer that...................................................27
3. THE DELETION OF THAT IN COMPLEMENT CLAUSES............................................29
1.1.The deletion of the complementizer that in English...............................................29
1.2.The deletion of the complementizer that in Lithuanian..........................................34
4.CORPUS-BASED CONTRASTIVE ANALYSIS OF THAT-DELETION IN VERBAL
COMPLEMENT CLAUSES IN ENGLISH AND LITHUANIAN..........................................37
4.1 The deletion of the complementizer that in complement clauses after the verbs of
saying in English and Lithuanian..................................................................................38
4.2 The deletion of the complementizer that in complement clauses after the verbs of
thinking in English and Lithuanian.........................................................................59
4.3 The deletion of the complementizer that in complement clauses after the verbs of
discoveringin English and Lithuanian.....................................................................69
CONCLUSIONS......................................................................................................................76
SUMMARY IN LITHUANIAN...............................................................................................78
3
REFERENCES.........................................................................................................................80
APPENDIX I............................................................................................................................85
4
ABSTRACT
The aim of this study is to present a contrastive analysis of verbal complement that-clauses in
English and Lithuanian and to describe, contrast and compare the cases of that deletion in
them on the basis of corpus evidence. Matrix verbs from three semantic domains, viz. saying,
thinking, and discovering, were investigated in the spoken, fiction, newspaper, and magazine
registers of the BNC and LLC corpora. The methods chosen for the study were qualitative
analysis as well as contrastive and content analysis. The corpus evidence demonstrated that
zero-that complement clauses are by far more frequent in English than in Lithuanian in all
registers considered. The deletion of that in Lithuanian appeared to be likely in case the
complement clause is preceded by an evidential marker, such as esą, neva “supposedly”, girdi
“(you) hear”, tarsi, lyg “as if”, or gal “maybe”, adopting the function of the complementizer,
which is not the case in English. The corpus evidence also revealed that in both languages
complement zero-that clauses share a structural peculiarity known as closeness of the clause
juncture and are likely to occur with the most frequent verbs in both languages: say, think, and
know. Another similarity is that in both languages matrix clauses followed by zero-that
complements can be confused with comment clauses. Further quantitative corpus-based
researches of Lithuanian complement clauses must be carried out in order to measure the
influence of the contextual factors favouring that deletion. In addition, parallel corpus-based
studies of the texts in the two languages can be conducted to provide a more detailed account
for similarities and differences between that- and zero-that complement clauses in the two
languages.
5
THE LIST OF ABBREVIATIONS USED IN THE CURRENT WORK:
BNC British National Corpus
CAGEL The Cambridge Grammar of the English Language (Huddleston, R., Pullum, G. K.,
2002)
CGEL A Comprehensive Grammar of the English Language (Quirk, R., Greenbaum S.,
Leech G., Svartvik J., 1995)
CGP Collins COBUILD Grammar Patterns 1: Verbs (Francis, G., Hunston, S., & Manning, E.
(eds.), 1998)
DLKG Dabartinės lietuvių kalbos gramatika (Ambrazas V. (ed.), 2006)
DLKŽ Dabartinės lietuvių kalbos žodynas (Keinys, S. (ed.), 2012)
GCE A Grammar of Contemporary English (Quirk, R., Greenbaum, S., Leech, G., Svartvik,
J., 1972)
LG Lithuanian Grammar (Ambrazas, V. (ed.), 2005)
LGSWE Longman Grammar of Spoken and Written English (Biber, D., Johansson, S., Leech,
G., Conrad, S., Finogan, E., 1999)
LKG III Lietuvių kalbos gramatika Volume III (Ulvydas K. (ed.), 1976)
LKŽ Lietuvių kalbos žodynas (http://www.lkz.lt/startas.htm)
LLC Corpus of the Contemporary Lithuanian Language
NODE The New Oxford Dictionary of English Pearsall, J. (ed.) 1998)
6
INTRODUCTION
The present research is concerned with the issue of verb complement that-clauses in
the English and Lithuanian languages. It is a corpus-based contrastive study focusing on
similarities and differences between the complement that-clauses in English and Lithuanian
and on the factors influencing the omission or retention of the complementizer that in both
languages and aims at answering the following research question: what similarities and
differences in contextual factors can be observed in cases of zero-that complement clauses in
English and Lithuanian? The choice of the subject of investigation was motivated by several
major considerations listed below.
In the first place, corpus-based analysis puts new perspectives on studying grammar.
Though for a long time the primary units of corpus-linguistic investigation were lexical items,
the recent advances in corpora design have enabled linguists to study genuinely grammatical
issues (Gries and Stefanowitsch, 2009: 934). According to McCarthy and O’Keeffe (2012: 7),
corpus linguistics “provides a means for the empirical analysis of language and in so doing
adds to its definition and description”. In particular, as corpus-based research allows to base
the generalizations upon a tremendous amount of empirical data, grammar aspects can be
examined from new angles, which often leads to novel observations about grammatical issues.
Secondly, corpus-based linguistic research has been recognised as a powerful method
of studying language use and numerous researches of linguistic structures have been
conducted on its basis around the world. However, corpus-based linguistic analyses are
relatively recent in Lithuania. Only several researches into grammar issues have applied
corpus-based methodology and hardly any of those so far have been concerned with the
aspects of syntax. Thus, the results of the present study can contribute to corpus-based studies
in Lithuania in general and to grammar studies in particular.
Thirdly, the study was motivated by the fact that no corpus-based contrastive
researches on corresponding English and Lithuanian structures have been conducted yet.
Corpus linguistics methodology has been broadly applied to investigating that-clauses in
English. Modern reference grammars based on corpus evidence, for instance, LGSWE (1999),
provide detailed descriptions of different types of that-clauses as well as factors influencing
the omission or retention of the complementizer that. Numerous linguists have studied the
issue of that-deletion in English, most of the recent researches being based on corpus
findings. On the contrary, though the omission of the complementizer that (kad/jog) is
possible in complement clauses in the Lithuanian language, the research into the factors
7
which influence the choice of the structure with zero complementizer has been insufficient
and hardly any Lithuanian books and publications on grammar can account for this
phenomenon. In this respect the research topic is genuinely novel and the investigation might
be considered a pilot corpus-based study into the issue of zero-that verb complements in
Lithuanian as well as a pilot contrastive study on the issue. The findings of the research can
also be of substantial value in teaching English to Lithuanian speaking learners as well as
teaching Lithuanian to English speaking learners. Last but not least, the results can be of
considerable assistance in translation theory.
As the research question suggests, the aims of the present research are:
to present a contrastive analysis of complement that-clauses in English and
Lithuanian;
to describe, contrast and compare the cases of that deletion in them on the basis of
corpus evidence.
In order to achieve the aims the following objectives were set:
1. to describe and compare the properties of complement that-clauses in English and
Lithuan;
2. to collect English and Lithuanian corpus evidence of zero-that complement clauses
taken by the verbs of saying, thinking and discovering;
3. to describe, contrast and compare the data in the two languages with the focus laid
on the contextual factors of their use, such as the register and mode of
communication, the properties of the matrix verbs and the structural properties of
the clauses, as well as on similarities and differences of the use of these structures
in the two languages.
The research methods incorporated in the study include contrastive analysis, corpus
search for collecting the data as well as qualitative approach in analysing, interpreting, and
describing the evidence, particularly content analysis.
Due to the limited scope of the paper, the present study focuses only on that-clauses
which serve the function of verbal complement and, among these, only object clauses have
been chosen for the investigation. The choice is influenced by the fact that, on the one hand,
this type of that-clauses is reported as the most frequent type of complement clauses in both
English and Lithuanian, and, on the other hand, these structures, as well as the conditions
under which the zero complementizer can be used in them in the English language have been
mostly studied and described in corpus linguistics. The latter argument is of particular
importance, taking into consideration the lack of theoretical background in Lithuanian
research on the deletion of the complementizer in complement that-clauses.
8
Another limitation concerns the verbs under consideration: only the verbs from
semantic domains of saying, thinking, and discovering were explored. This choice was
determined by two considerations. Firstly, these three semantic groups of verbs taking object
that-complements are the largest and three English verbs, namely say, think, and notice,
which are reported in corpus-based researches as those taking zero-that complemets most
often, belong to these semantic groups. Secondly, Lithuanian asyndetic object that-clauses are
also reported to frequently follow the matrix verbs belonging to these semantic domains.
One more limitation to the research is that, due to the different design of the
investigated corpora, namely, the British National Corpus (BNC http://corpus.byu.edu/bnc/)
and the Corpus of the Contemporary Lithuanian Language (LLC http://donelaitis.vdu.lt), only
particular text categories are incorporated into the study. The following subsections of the
BNC and the LLC were searched in the process of collecting the data (as these subsections are
of relative stylistic correspondence in both corpora): spoken, fiction, newspaper, and magazine
of the BNC and sakytinė kalba “spoken”, grožinė literatūra “fiction”, and publicistika
“journalism” of the LLC. Due to the tremendous size of the corpora and absence of specific
software simplifying the search for evidence, no quantitative, but only qualitative research
was conducted.
9
1. CORPUS-BASED APPRROACH TO LINGUISTIC STUDIES
1.1 General overview of corpus-based linguistic studies.
1.1.1 Corpus linguistics as means of studying the language
The rapid advancement of informational technologies has significantly contributed to
scientific researches in all areas of study. The possibility to collect and store tremendous
amounts of information, available access to it, as well as the creation of specific software
which helps to process massive data have provided researchers with absolutely novel
opportunities for scientific studies. Among other fields, linguistics acquired a chance to put
new perspectives on the study of language.
Creation of corpora - extensive collections of texts stored in electronic databases and
most often equipped with specific search software - has enabled researchers to reinforce
descriptive approach to language study. Corpus-based research allows to shift the focus from
making generalizations based on speakers' intuitions about the language to observations about
real language in use as well as about typical and common linguistic choices of speakers
(Anderson and Corbett, 2009: 2). The descriptive approach based on the evidence drawn from
corpora is considered an essential feature of corpus linguistics. For instance, McEnery and
Wilson (1996: 1) define corpus linguistics as “the study of language based on examples of
'real life' language use” and Tognini Bonelli (2010: 15) claims that corpus linguistics aims at
analyzing and describing how language use is realized in texts. However, differently from
earlier theories within the descriptive approach to language, modern corpus linguistics
provides a possibility to base judgments not on limited collections of written texts, which
used to be the case in the times prior to the computer era, but on an ample amount of natural
texts of both written and spoken modes of communication, as well as of different genres and
registers. It is not only the amount of data, but also their relevance, authenticity and
availability online, on the one hand, and the speed that specific computer programs guarantee
in processing them, on the other hand, which turns corpus linguistics into a highly attractive
scientific enterprise, as it makes challenging or confirming linguistic intuitions a much faster
process (Anderson and Corbett, 2009: 2) and can guarantee far more reliable results than
those of the researches based on collecting and handling data in a manual way. As McEnery
and Hardie (2012: 2) put it, corpus linguistics is “the study of language data on a large scale.”
Most linguists agree that corpus linguistics is rather a methodology than a separate
branch of linguistics, as linguistics typically studies specific aspects of language. For instance,
McEnery and Wilson (1996: 1) claim that corpus linguistics is “a methodology rather than an
aspect of language requiring explanation or description”. McEnery and Hardie (2012: 1) share
10
a similar view and consider corpus linguistics “an area which focuses upon a set of
procedures, or methods, for studying language.” However, some scholars argue that there is
more to corpus linguistics than merely a methodology, as dealing with such tremendous
amounts of evidence, which were never available before, leads to absolutely new quantitative
researches into language, and the findings of the quantitative studies inevitably influence
qualitative insights. In other words, the methodology based on computations and statistics has
resulted in revealing genuinely new patterns of language and emergence of new linguistic
hypotheses (Tognini Bonelli, 2010: 18). Thus, corpus linguistics appears to be a discipline
which due to its specific methodological principles contributes to making novel discoveries
about the language system as well as describing it at a new, higher level. As McCarthy and
O’Keeffe (2012: 7) put it, corpus linguistics “provides a means for the empirical analysis of
language and in so doing adds to its definition and description.”
1.1.2 Corpus-based studies of grammar
In regard to objects of linguistic studies, the first to benefit from the opportunities
offered by corpora and corpus linguistic techniques were the scholars within the branch of
lexicography. The primary unit of corpus-based analysis has been the word and the majority
of corpus-based linguistic researches have been centered around it. The findings of these
studies have led to novel insights into the nature of word collocations and factors influencing
semantic components of lexical items (Gries and Stefanowitch, 2009: 934).
However, advances in corpus searching techniques, such as tagging and parsing, have
enabled linguists to initiate research into truly grammatical aspects of language. McEnery and
Wilson (1996: 109) estimate that nowadays corpus-based grammatical studies are as frequent
as lexical. Corpus linguistics technologies allow to investigate a wide range of grammar
related issues. For instance, Gries and Stefanovich (2009: 934-940) distinguish the following
directions in corpus-based grammatical studies which focus on the relation of lexis and
grammar: studying structure-sensitive collocates (for instance, the most frequent adjective +
noun structure combinations), studying collocational frameworks and grammar patterns (for
instance, identifying the most frequent nouns in the framework a +[noun] + of and studying
their features or investigating abstract grammatical frames like [verb + noun phrase]),
studying colligates (the tendency of particular lexical items to co-occur with particular
grammatical structures (for instance identifying verbs taking that-clausal or infinitival
complements and studying their features), and collostructional analysis (which deals with
grammatical research into phraseological units of language). Conrad (2010) distinguishes the
11
following 4 types of patterns most common in corpus-based grammar analyses: vocabularygrammar (or lexico-grammar), grammatical co-text, discourse-level factors and the context of
the situation (Conrad, 2010: 229). These examples illustrate the great variety and scope of
corpus-based investigations into grammar.
Besides the diversity of research objects and topics, corpus linguistics allows to view
aspects of grammar from new angles. Corpus-based studies of linguistic structures have
switched the focus from traditional dichotomous perspective, which used to describe grammar
in terms of accurate versus inaccurate choices. Instead, corpus-based studies, based on
processing tremendous amounts of data, allow to investigate and determine common and
uncommon choices and to see the patterns revealing what is typical or untypical in particular
contexts. These patterns show the correspondence between the use of a grammatical feature
and other factors in the discourse or situational context, such as another grammatical feature,
a social relationship, or the mode of communication. In addition, corpus-based methods allow
to reveal significantly more about the factors influencing the choice of linguistic structures
which cannot be made on the basis of appropriate versus inappropriate. For instance, corpusbased analyses have shed much light upon the factors influencing the choice of synonymous
syntactical structures, such as that-clausal, infinitival or -ing verbal complements or the use of
that versus zero complementizer in complement verbal and adjectival clauses. Corpus
annotation as well as corpus-based evidence have led to describing grammar not just in
structural terms, but describing the typical social and discourse circumstances associated with
the use of particular grammatical features and, what is more, to discover more about the
multiple factors that simultaneously have an impact on grammatical choices (Conrad, 2010:
234).
Last, but not least, the development of corpus linguistics at multilingual level offers a
novel and ample opportunity for contrastive language studies. McEnery and Wilson (1996:
188) name the growing variety of corpora in different languages “the most important
development in corpus linguistics recently” and claim that the creation and advancement of
corpora of as many languages as possible is one of the essential current aims in corpus
linguistics considered internationally. Contrastive language studies based on corpora evidence
are of great importance to comparative and contrastive linguistics in general and to translation
theory and second and foreign language teaching in particular.
1.1.3 Qualitative approach in corpus-based studies
The two main approaches to corpus-based studies are quantitative and qualitative.
12
Biber and James (2009: 1287) claim that corpus linguistics usually employs quantitative
studies for investigation of language use. To a great extent this is due to considerable sizes of
contemporary corpora and the possibility to apply advanced computing technologies to
processing the data obtained in corpus-based researches. Counting and analysing frequencies
of occurrences of linguistic phenomena allows to detect specific linguistic patterns and to
base conclusions about their features and behaviour upon statistical data. However, qualitative
research, which does not attempt to assign frequencies to linguistic features identified in the
data, is of no less importance. McEnery and Hardie (2012: 2) claim that qualitative and
quantitative analysis are equally important in corpus linguistics. McEnery and Wilson (1996:
76) emphasize that qualitative analysis offers “a rich and detailed perspective on the data” and
specify the advantages the qualitative approach can offer: it devotes equal attention to both
rare and frequent phenomena, allows to observe subtle variations in the data, and, therefore, to
draw “very fine distinctions.”
Differently from quantitative corpus-based research, in qualitative research the corpus
data serve merely as a basis for detecting and describing linguistic patterns and aspects of
their usage. What is important, corpus data provide authentic examples of particular
phenomena used in real-life situations. Qualitative researches make an important and valuable
part of linguistic investigations in that they precede quantitative researches. As McEnery and
Wilson (1996: 76) put it, “before linguistic phenomena are classified and counted, the
categories for classification must be identified.”
1.2 Corpus-based linguistic studies in Lithuania
Since corpus linguistics is considered to be a rather new undertaking in the linguistic
field of science, which has developed during the last 50 years, it is natural to suppose that in
Lithuania it is much younger. Corpora, software necessary to conduct corpus-based studies as
well as corpus based linguistic researches evolved in Lithuania in the late 90-s of the previous
century. Marcinkevičienė (2010: 22) considers the first Lithuanian corpus to appear in 1997.
Most of the corpora of Lithuanian compiled since then have been small specialized corpora
designed for specific research purposes by linguists conducting their investigations. Currently,
there is only one public Lithuanian corpus available online free of charge – The Corpus of
Contemporary Lithuanian Language (LLC) designed by the Centre of Computational
Linguistics at Vytautas Magnus University in Kaunas.
Lithuanian corpora as a phenomenon are fresh, new and few in number, they still
require much improvement as well as more different specialized corpora of the Lithuanian
13
language are required for carrying out solid and thorough as well as versatile linguistic
researches. For instance, the design and software of LLC, as well as its annotation and the
size and contents of its parts (at least of the spoken section of the corpus) have numerous
drawbacks which make linguistic research on specific topics, such as grammar and syntax,
quite exhausting and rather ineffective.
Another problem is specific software necessary for corpus-based investigation of
peculiar linguistic issues on a large scale, particularly subjects in grammar and syntax. In
corpus linguistics the creation of programs usually lies in the sphere of applied linguistics
rather than linguistics itself, and linguists typically work in close cooperation with
programmers. However, the procedures and algorithms used in certain investigations often are
a matter of intellectual property; thus, they are not shared or widely discussed.
Marcinkevičienė (2000: 14) notes that the means of analysing corpora as well as particular
software and tools are an inseparable, though most often a concealed part of corpora, and
scholars would rather share the procedures and results of their studies than reveal the details
of software. What is more, creating the automatic procedures for identifying specific
linguistic structures is a complicated process, and much software of this kind is still in the
stage of development and improvement (Kirk, 1997: 361).
Thus, due to the novelty of corpora and corpus-based methods of linguistic research, as
well as to the lack and imperfection of the software, so far there have been rather few corpusbased linguistic studies of the Lithuanian language and they have mostly been limited to
lexicographical, lexicological, and morphological studies. There is no account of any corpusbased investigations into linguistic structures of Lithuanian.
14
2. VERB COMPLEMENT CLAUSES IN ENGLISH AND LITHUANIAN
2.1 Complement clauses in English
2.1.1 Definitions and terminology
LGSWE (1999: 658) defines complement clauses of English as “dependent clauses
used to complete the meaning relationship of an associated verb or adjective in a higher
clause.” Greenbaum et al (1996) in their corpus-based study of complement clauses report
them as the most frequent type of subordinate clauses in English (1996: 81). Complement
clauses fall into finite, consisting of wh-clauses and that-clauses, and non-finite, which can be
split into to-infinitive and -ing-clauses (LGSWE, 1999: 658; CAGEL, 2002: 950). Verb
complement that-clauses comprise the biggest subclass of complement clauses of English, as
they are reported as the most frequent type by corpus-based studies (Greenbaum et al, 1996:
81; LGSWE, 1999: 674; Biber, 1999: 133).
The first to introduce the term complement that-clause was Storms, who suggested it as
a more correct one instead of the common then term object that-clause (1966: 254). The term
has been widely accepted and used by scholars in numerous works on related grammatical
issues. However, the term is not absolutely conventional, and different grammarians name the
same phenomenon differently. For instance, CAGEL (2002: 951) refer to them as declarative
content clauses, while CGEL (1995: 1048) term the same category as nominal that-clauses,
since these clauses resemble noun phrases in that they usually function as subjects, objects, or
predicatives. Dixon (2006) in his study of complement clauses and complementation
strategies of eleven languages, names them as complement clauses of the fact type, since they
refer to the fact that something took place (2006: 23).
In this paper, we stick to the terms complement that-clause and complementizer (for
that) adopted from the classification suggested in LGSWE (1999: 658). The decision to adopt
the terminology from this work is due to the fact that LGSWE is considered one of the most
influential modern corpus-based descriptive grammars of English, and a lot of corpus-based
researches on the issues related to the current study cite the work as well as adopt its
terminology.
2.1.2 Properties of complement that-clauses
As both complement that-clauses and wh-clauses belong to the class of the finite
complement clauses, their structure typically resembles that of a main clause, i.e. they must
15
have subjects and include tense or modality (LGSWE, 1999: 658; Dixon, 2006: 24). The
distribution of these clauses in language depends on the mode of communication: corpusbased studies of subordinate clauses in English report that- and wh-clauses as prevailing in
speech, differently from non-finite complements, which are more frequent in writing (Biber,
1999: 134; Kirk, 1997: 354; Greenbaum et al, 1996: 83). Register is also an important factor
influencing the use of complement clauses: for instance, verbal to-infinitive complements are
nearly as frequent in conversation as in academic prose, while that-complement clauses are
common in conversation and considerably rare in academic writing (Biber, 1999: 134).
Another peculiarity of complement that-clauses is the possible omission of the
complementizer that (LGSWE, 1999: 658; Dixon, 2006: 24; CAGEL, 2002: 951), which
means that both sentences illustrated in example (1) are grammatically accurate:
(1) I thought it was a good film.
cf. I thought that it was a good film.
(Biber et al, 1999:658)
Complement that-clauses can perform several syntactic functions typical of noun
phrases. Generally they serve as subjects, directobjects, and subject complements for clausal
complements controlled by verbs, as well as adjectival complements (CGEL, 1995: 1049;
LGSWE, 1999: 658; CAGEL, 2002: 957). These functions are illustrated in the following
example (2a)-(2d) respectively (CGEL, 1995: 1049):
(2)
a. That the invading troops have been withdrawn has not affected our
government's trade sanctions.
b. I noticed that he spoke English with an Australian accent.
c. My assumption is that interest rates will soon fall.
d. We are glad that you are able to join us on our wedding anniversary.
In addition, CGEL (1995: 1049) and Hudston-Ettle (2002: 259) distinguish the function of
appositive:
(3)
Your criticism, that no account has been taken of psychological factors,
is fully justified.
(CGEL, 1995: 1049)
In regards to syntactic functions, LGSWE (1999: 660) report that, according to corpusbased studies, in real language use there is a strong tendency for complement that-clauses to
16
be used in extraposition (with a dummy subject it) rather than in subject (pre-predicate)
position1:
(4) It just never crossed their minds that it might happen.
cf. That it might happen just never crossed their minds.
For the frequency of syntactic functions of complement clauses, numerous corpusbased studies reveal that the most common function of verbal complement that-clauses is that
of the direct object (McDavid, 1964: 108; Elsness, 1882: 1; Greenbaum et al, 1996: 86;
Hudson-Ettle, 2002: 265). The distribution of syntactic functions of complement that-clauses
in general can be illustrated by the example of a corpus-based study on English complement
clauses by Greenbaum et al (1996). They report that there are considerable differences in the
distribution of syntactic functions of that-clauses across spoken and written modes as well as
registers. According to them, the clauses functioning as adjectival complements are more
frequent in writing than in speech; however, in general, this is the least frequent of all
functions2. Complement that-clauses functioning as subjects appear to be more frequent in
formal writing and broadcast discussions, with extraposition being the norm. However, on the
whole, the investigation proves that the function of direct object in verbal complement thatclauses is by far the most frequent in English (op. cit.: 86), which is in complete accordance
with the findings of other corpus-based studies on the issue.
The grammatical function of the complement clause is also a formal factor influencing
the omission or retention of the complementizer that. The complementizer that can generally
be omitted when the that-clause is the direct object or complement, as well as in extraposed
subject constructions (McDavid, 1964: 113; CGEL, 1995: 1049; CAGEL, 2002: 952;
Hudson-Ettle, 2002: 259). However, the complementizer is obligatory when the complement
that-clause is the subject of the sentence or otherwise precedes the matrix predicator
(McDavid, 1964: 107; CAGEL, 2002: 952), or when it serves the function of the adjectival
complement in pre-predicate position, as illustrated in example (5) from (LGSWE, 1999:
662):
(5) That it would be unpopular with colleges or students was obvious.
1
2
The same tendency is reported by Hudson-Ettle (2002: 260)
The same conclusion is made by McDavid (1964: 113) in her corpus based study on alternation of the
complementizer in that-clauses.
17
Another structural factor determining the retention of that is possible ambiguity: in
case the boundaries of the main and the complement clauses are not clear without the
complementizer, it cannot be omitted (McDavid, 1964: 113; Elsness, 1982: 2; CGEL, 1995:
1050). For instance, in the example (6a) and (6b) that omission would make it not clear
whether the adverbial once again belongs to the main or the complement clause:
(6) a. They told us once again that the situation was serious.
b. They told us that once again the situation was serious. (CGEL, 1995: 1050)
Similarly, the absence of the complementizer in the coordinated complement clause in
example (7) would lead to possible interpretation of it as a part of the main clause:
(7) I realize that I'm in charge and that everybody accepts my leadership.
(CGEL, 1995: 1050)
2.1.3 Verbs controlling complement that-clauses
Complement clauses are elements controlled either by an adjectival predicate or a
lexical verb of the main clause.3 Formally, verbs taking that-clauses as complements belong
to the class of transitive (or ditransitive) verbs, as due to their lexico-grammatical pattern they
require an object which completes the meaning of the verb. The lexico-grammatical pattern of
the matrix verb determines the structural type of the complement, and usually a verb can
control only a certain type of complements. As Biber (1999: 136) illustrates it, the verbs
begin, start, want, or like take to-infinitive complements, imagine, mention, conclude, argue,
or guess require that-complements, while think, say, hope, decide, or wish can control both toinfinitive and that-complements. However, such verbs still exhibit strong preference for being
used in language consistently with only one of these types of complements. For instance, the
verbs say, think, and know can grammatically control both complement to-infinitive and thatclauses, but they are exceptionally frequent with the latter, and thus make a strong association
pattern with complement that-clauses (op. cit.: 137).
Semantically, it has been observed that the verbs taking that-clausal complements
belong to very specific domains of meaning. In particular, verbal that-complement clauses
functioning as direct objects usually report human speech, thoughts or attitudes (LGSWE,
3
Henceforth, in the current work they are referred to as matrix verbs, as in CAGEL (2002: 958).
18
1999: 661; CGP, 1996: 103), therefore there are only a few semantic domains the matrix verbs
can be ascribed to. In general, there are two major semantic domains of matrix verbs taking
that-clauses as direct objects, which can be defined as saying and thinking (Biber, 1999:
138), with finer groups within them suggested by different scholars. For instance, Storms
(1966) suggests two main classes of verbs taking that-complements: those expressing the
speaker's attitude to the “truth” of the information conveyed in the content of the complement
clauses (for instance, say, state, or postulate) and those expressing how the speaker is affected
by the information content (for instance, think, suppose, or complain). He also distinguishes
seven subgroups: verbs describing the way of communication (for instance, say), verbs
pointing to the truth or non-truth of the information content (for instance, affirm), objective
verbs implying personal meanings (for instance, notice), subjective verbs implying personal
meanings (for instance, suspect), verbs with affective meanings (for instance, worry), link
verbs (for instance, seem), and verbs governing clauses of purpose with the modal should in
them (for instance, arrange) (op. cit.: 266). LGSWE (1999: 662), as well as Biber (1999:
138), ascribe verbs taking that-complements to semantic domains of “mental verbs of
cognition” (such as think, know, or believe), “speech act verbs” (such as say, admit, or agree),
and “other communication verbs” (such as show or suggest). Dixon (2006:10) divides such
verbs into those of “speaking”, “thinking”, “attention”, and “liking”, while CGP (1996: 97)
suggest nine groups: “say”, “add”, “scream”, “think”, “discover”, “check”, “show”,
“arrange”, and “go” (the last being related to quoting, for instance, The story goes that...).
Still, the verbs listed by different scholars as taking that-complements (McDavid, 1964: 109;
Storms, 1966: 266-267; LGSWE, 1999: 663-666; CAGEL, 2002: 958-963) can be generally
ascribed to groups referring to either a speech act or some kind of mental activity or state.
Certain matrix verbs display an exceptional tendency to be used with complement thatclauses. Corpus-based investigations report the verbs think, say, know, see, believe, find, feel,
suggest, and show as the most common verbs taking that-complements (Biber, 1999: 137;
LGSWE, 1999: 663), with think, say, and know being the most frequent ones (LGSWE, 1999:
13; Rissanen, 1991: 276). Still, there are differences in distribution of these structures in
language. The verb say with that-complement clauses is most frequent in news, while think is
typical of conversation and writing but rather unusual in news and academic prose (LGSWE,
1999: 668).
2.1.4 Properties of the complementizer that
Complementizer that is generally defined as a functional conjunction bearing no
19
semantic meaning. However, Frajzyngier and Jasperson (1990) 4 in their study on the
functional differences between the that-clauses and to-infinitive and -ing complements insist
on the complementizer that possessing semantic meaning due to which this type of
complement has a specific function. They argue that, as it derived from the demonstrative
pronoun that, it can still be semantically related to the anaphorical or cataphorical use of that
in discourse, which refers to items that have been mentioned or are going to be mentioned in
the text. They claim that verbal that-complements express ideas belonging to the semantic
domain of speech and thus are propositions embedded in other propositions expressed by the
whole sentences. Differently, to-infinitive and -ing clauses express ideas belonging to the
domain of reality; they refer to events and do not contain embedded propositions. The idea
can be illustrated by the following example:
(8) a. He said that he likes apples.
b. He wants to eat apples.
c. He was caught eating apples.
(Frajzyngier and Jasperson, 1990: 139)
The embedded clauses in (8b) and (8c) are not fragments of speech and refer to reality, while
the embedded clause in (8a) refers to speech and contains a proposition within another
proposition. Thus, according to the authors, the main function of the complementizer that is to
mark the following clause as belonging to the domain of speech rather than the domain of the
reality (op. cit.: 134). In the authors' opinion, this function of the complementizer that also
accounts for the fact that complement that-clauses are remarkably typical with the verbs of
saying and thinking (op. cit.: 140).
However, most English grammars and works on complement clauses declare that the
complementizer that is different from other conjunctions, such as, for instance, when, though,
or if, because it is not ascribed any inherent semantic meaning anymore and serves a purely
grammatical function of joining two clauses and introducing a complement clause (McDavid,
1964: 103; Storms, 1966: 251; Elseness, 1984: 1; LGSWE, 1999: 135; CAGEL, 2002: 955).
Storms (1966) claims this to be a reason why the omission of that is possible - it does not
affect the semantic relation within the sentence (1966: 251). Similarly, McDavid (1964) and
CAGEL (2002) explain that it does not belong to the boundaries of the complement clause, it
only expands it, and thus, being “simply a syntactic marker of subordination”, can be omitted
(CAGEL, 2002: 955).
4
In this article they follow and develop the ideas expressed by Bolinger in his 1972 work That's that.
20
2.2 Complement clauses in Lithuanian
2.2.1 Definitions and terminology
The syntactic structures of the Lithuanian language that most closely correspond to
English complement clauses are completive clauses 5 defined in LG (2006) as clauses
“subordinated to a verb, a verbal noun, or a neuter adjective or an adverb in a higher clause”
(2006: 725). They comprise a subclass of complex integrated 6 clauses, with relative and
correlative clauses as the two other subclasses (LG, 2006: 725, DLKG, 2005: 665).
Differently from the English language, in this classification completive clauses 7 do not
include non-finite clauses, thus, the class of Lithuanian complement clauses consists only of
two subgroups: embedded questions 8 (introduced by relative pronouns and adverbs) and
completive clauses introduced by conjunctions (LG, 2006: 726-727; DLKG, 2005: 669).
However, the latter subgroup is wider than the category of complement that-clauses in
English, as the Lithuanian complement clauses in this group can be joined not only by the
conjunctions kad/jog (that), but also by the conjunctions lyg and tartum/tarsi (which all can
be translated as if) in case “an unreal, imaginary or possible action is referred to” (LG, 2006:
726), as illustrated in the following example:
(9) Jam pasivaideno, lyg vilkas būtų.
'He fancied he saw something like a wolf.'
Iš tolo girdėjosi, tartum jis ką dainuotų.
'One could hear from afar what sounded
like his singing something.'
(LG, 2006: 726)
Also, sometimes Lithuanian complement clauses are joined by a conjunction kai “when”,
which adds to temporal meaning to the clause (op. cit.: 726):
(10) Aš mėgstu, kai pas mane svečiai.
5
6
7
8
'I like (it) when I have visitors.'
In this work, the English terminology for the issues of the Lithuanian language is mostly adopted from LG
(2006), as this is the most prominent contemporary Lithuanian grammar published in English. For any
specific terms used in other Lithuanian books and publications the translation and the original Lithuanian
terms are provided.
Integrated clauses in this classification represent the functional way of relation between the main and the
subordinate clause, with the subordinate clause being related to one constituent of the main clause. The other
type, non-integrated clauses, represents the semantic way of relation, where the subordinate clause is related
to the main clause as a whole or the predicate group (LG, 2006: 719-720, DLKG, 2005: 660; Labutis, 2002:
157).
Henceforth, in the work completive clauses will be referred to as complement clauses.
They correspond to the class of complement wh-clauses in English.
21
Thus, considering that the complement clauses in Lithuanian can be subordinated to a verb,
noun, or adjective, and that they can be joined to the main clause by conjunctions other than
that, the structure defined in this work as verb complement that-clause (which is the subject of
the current study) comprises a part of Lithuanian complement clauses introduced by
conjunctions.
The classification of Lithuanian complex clauses suggested in DLKG (2005) and LG
(2006) is adopted from the earlier academic Lithuanian grammar LKG III (1976). However,
most of other Lithuanian grammars and works on syntax classify subordinate clauses on the
basis of the way they are joined to the main clauses, on the one hand, or on the basis of their
syntactic function in the sentence, on the other hand. In respect to the means of joining the
main and subordinate clauses, the verb complement that-clauses are generally ascribed to the
group joined by conjunctions, and particularly to those joined by functional conjunctions
kad/jog “that”, while in respect to the syntactic function they can be attributed either to
subject or to object clauses (Balkevičius, 1963: 316, 333; Dambriūnas, 1963: 60, 68;
Masilionis, 1985: 74, 77; Sirtautas and Grenda, 1988: 153, 156, Labutis, 2002: 154, 162;
Holvoet and Judžentis, 2003: 122).
A more recent and detailed classification of Lithuanian complex clauses is suggested
in Holvoet and Judžentis (2003). Besides the distinction on the bases of the means of
connection to the main clause and syntactic function, the verb that-complement clauses,
according to the authors, also fall into the class of nominal (daiktavardinai), as, in the
sentence, they occupy positions typical of nouns9 (2003: 117). Moreover, in respect to the
type of syntactical relation, nominal clauses belong to the class of clauses controlled
(valdomi) by verbs in the main clause (opposed to modifying (modifikuojantys) clauses), while
formally they are finite
(finitiniai) as opposed to non-finite (nefinitiniai) infinitival
(bendraties) and participial (dalyviniai) clauses (op. cit.: 119-120, 135).
2.2.2 Properties of complement that-clauses
Like English complement that-clauses, Lithuanian complement that-clauses are finite,
therefore structurally they also resemble main clauses in the sense that grammatically such a
clause is a simple sentence and thus has a predicative centre (LKG III, 1976: 789;
9
The other two classes are adjectival (būdvardiniai) and adverbial (prieveiksminiai) clauses (Holvoet and
Judžentis, 2003: 117). This classification is similar to that of the English complement clauses in Quirk et al
(1995).
22
Dambriūnas, 1963: 57; Balkevičius, 1963: 312; Masilionis, 1985: 74; Labutis, 2002: 145).
However, differently from English, where a predicative centre includes a subject and a
predicate, Lithuanian subordinate clauses (and complement clauses as a subclass) may consist
of a predicate only 10 (Balkevičius, 1963: 312), as illustrated in the following example:
(11) - Na, šukuoseną! Bifė sukikeno. – Lažinuosi, vis dėlto pasidarysi. (LLC, fiction)
[- Well, a hairstyle! Biffi giggled. - (I) Bet, (you) will have it made yet.]11
The example (11) also illustrates that both the main and the complement clauses can lack
subjects, as Lithuanian is a synthetic language and a finite verb form in the predicate marks
the number and gender of the subject, therefore pronominal subjects are often omitted
(Sirtautas and Grenda, 1988: 22; DLKG, 2005: 572; LG, 2006: 599).
Complement clauses in Lithuanian can also be distinguished according to the type of
modality of the clause. These types are generally consistent with modality types of simple
clauses (Holvoet and Judžentis, 2003: 165). Thus, complement clauses can be statements
(konstatuojamieji), clauses of motivation (skatinamieji), interrogative (klausiamieji), and
exclamatory (šaukiamieji) joined by kaip “how”, kiek “how much”, or koks “what”, as
illustrated in example (12a)-(12d). respectively (op. cit.: 166):
(12) a. Jis žinojo, kad atsitiks nelaimė.
[He knew that an accident would occur.]
b. Jis norėjo, kad visi jo klausytų.
[He wanted everybody to listen to him.]
c. Jis klausė, kada svečiai ateis.
[He asked when the guests are coming.]
d. Stebiuosi, kaip greitai tos
[I'm amazed how soon the holidays
atostogos prabėgo.
have passed.]
The object that-clauses can be either statements or clauses of motivation. The latter typically
are expressed by predicates in subjunctive mood (op. cit.: 166).
Lithuanian complement clauses are subordinated to a verb, a verbal noun, or a neuter
adjective or an adverb of the main clause (DLKG, 2005: 665; LG, 2006: 725). As well as
English, they belong to nominal clauses, thus, similarly to nouns, they are typically controlled
by predicates in the main clause (Holvoet and Judžentis, 2003: 134). The means by which
10
11
Lithuanian simple sentences fall into two-member (dvinariai) sentences (in which the predicative centre
consists of both subject and predicate) and one-member (vienanariai) sentences (in which the predicative
centre consists of the predicate or subject only) (Balkevičius, 1963: 68; Labutis, 2002: 124).
The translations provided in square brackets are made by the author of this paper.
23
verb that-complement clauses are joined to the main clause are functional (asemantic)
conjunctions jog/kad12 “that” (DLKG, 2005: 669; LG, 2006: 726). However, differently from
English, Lithuanian complement clauses are always separated from the main clause by a
comma, which is a general rule for subordinate clauses in Lithuanian (Dambriūnas, 1963: 59;
Masilionis, 1985: 76, Būda, 1986: 6). Thus not only the conjunction that signals the beginning
of the complement clause, but also a comma, which precedes it.
In regards to their
frequency, complement that-clauses are reported as the most frequent type of subordinate
clauses in Lithuanian (Labutis, 2002: 269).
An important property of a complement clause is its syntactic function in the
sentence. Būda (1979) considers the syntactic function of a subordinate clause the most
important criterion for the classification of the subordinate clauses at the highest level (1979:
13). Subordinate clauses in Lithuanian can function as subjects, predicates, objects, attributes,
and adverbials (Balkevičius, 1963: 325; Dambriūnas, 1963: 60; Masilionis, 1985: 77;
Sirtautas and Grenda, 1988: 156; Labutis 2002: 162). The verb complement that-clauses
function as subjects13 and objects (DLKG, 2005: 671-672 LG, 2006: 729-730; Holvoet and
Judžentis, 2003: 118), as illustrated in examples (13a) and (13b) respectively (LG, 2006: 730):
'It seemed to me that it was already late.'
(13) a. Man atrodė, jog jau vėlu.
b. Tėvas patikėjo, kad senis sako teisybę.
'Father believed that the old man was
telling the truth.'
There is a specific difference between the object complement that-clauses in English
and Lithuanian. English complement clauses usually function as direct objects (CGEL, 1995:
1049; LGSWE, 1999: 658; CAGEL, 2002: 957). However, in Lithuanian, which distinguishes
six major cases of nouns, different verbs, depending on their valency patterns, can control
nouns and pronouns in different cases (Ambrazas, V., 2006a: 195). Thus, objects can be
expressed by noun or pronoun forms in genitive, dative14, accusative or instrumental cases, as
well as by a prepositional phrase (Balkevičius, 1963: 346; Masilionis, 1985: 86; Sirtautas and
Grenda, 1988: 161). As a consequence, complement that-clauses “can take the object position
in the superordinate clause interchangeably with oblique case forms (the genitive, accusative,
12
13
14
Both jog and kad are translated into English as that.
The function of an appositive distinguished in CGEL (1995: 1049) in Lithuanian is attributed to the cases of
the subject function (LG, 2006: 729-730)
Nouns in dative complete the object of the main clause, while in genitive, accusative, and instrumental cases
nouns complete predicates of the main clause (Masilionis, 1985: 86).
24
or instrumental) or with a prepositional phrase”15 (LG, 2006: 730), as illustrated in examples
(14a)-(14d) respectively :
(14) a. Aš kartais net gailiuosi, kad tiek daug suprantu. (LLC /fiction)
(cf. gailiuosi (ko?) savo elgesio – genitive)
[I sometimes even regret that I understand so much.]
(cf. regret my behaviour)
b. Tu įtari, kad jis neištikimas? (LLC /fiction)
(cf. įtari (ką?) melą – accusative)
[Do you suspect that he is unfaithful?]
(cf. suspect a lie)
c. Marytė labai džiaugėsi, kad užvažiavom, <…>. (LLC/spoken)
(cf. džiaugėsi (kuo?) pergale – instrumental)
[Marytė rejoiced that we dropped in <…>.]
(cf. rejoiced at the victory)
d. Sutinku, kad kiekvienas įstatymas turi būti kategoriškas.
(LLC /journalism)
(cf. sutinku (su kuo?)su pasiūlymu – a prepositional phrase)
[(I) agree that every law must be categorical.]
(cf. I agree with the proposal)
Last but not least, object that-clauses have a certain position in the sentence. Typically,
they follow the main clause (Balkevičius, 1963: 346). Drotvinas (1961: 180) claims that
object that-clauses have a strictly fixed position, which is always after the main clause. He
also emphasizes that subordinate clauses in conjunctionless (asyndetic) sentenses never
interfere in the main clause (Drotvinas, 1961: 180). Sirtautas and Grenda (1988) mention that
object that-clauses joined by the complementizer jog, as well as subject that-clauses, must
necessarily follow the main clause (1988: 156). However, it is also reported that object thatclauses on certain occasions can precede the main clause, as illustrated in the example (15),
with a particular intention “to emphasize the subordinate clause or to make it the theme of the
15
English object complement clauses can also be interchangeable with prepositional phrases, e.g. agree with
[noun phrase] vs. agree [that-clause]. However, though such prepositional phrases function as prepositional
objects, corresponding complement that-clauses function as direct objects.
25
sentence” (LG 2006 730).
(15) Kad buvo meilė, kas nežino. (DLKG, 2005: 673)
[That there was love who doesn't know.]
Nevertheless, if a complement sentence is joined by jog, it cannot precede the main clause
(Masilionis, 1985: 154). In this respect, it should be mentioned that the typical position of the
English object that-clause is also after the main clause (LGSWE, 1999: 659).
2.2.3 Verbs controlling complement that-clauses
The same way as English verb complement that-clauses, Lithuanian equivalents are
also controlled by verbs in the main clause (Holvoet and Judžentis, 2003: 134). These
Lithuanian verbs are of the same lexico-grammatical pattern as English counterparts: they are
incomplete in meaning and require an object, which can be realized by a noun phrase or a
complement clause (LKG III, 1976: 807; Labutis, 2002: 157). However, only a part of these
verbs in Lithuanian are transitive, namely those which take an object in the genitive case (see
example (14b)). Still, in both languages under consideration the syntactic valency of the verb
in the main clause determines the type of the complement (Biber, 1999: 136; Holvoet and
Judžentis, 2003: 163). For instance, the verb teigti “state” can control a that-clause, but cannot
control an indirect question, while pavyko “managed” takes an infinitival complement but
cannot take a that-clause complement (Holvoet and Judžentis, 2003: 163). Similarly to
English, there are Lithuanian verbs which can control more than one type of complement
clauses. For instance, the verb prašyti “ask” can take that-complement clauses as well as
infinitival complement clauses, as illustrated in example (16a) nad (16b):
(16) a. Jis paprašė, kad atneščiau žurnalą.
[He asked (me) that I should bring the magazine.]
b. Jis paprašė atnešti žurnalą.
[He asked (me) to bring the magazine.]
(Holvoet and Judžentis, 2003: 120)
In respect to the meaning, verbs taking complement that-clauses in Lithuanian can
belong only to certain semantic domains (LKG III, 1976: 812), which are the same semantic
domains their English counterparts belong to. Lithuanian matrix verbs taking complement
that-clauses are generally defined as verbs of saying, thinking, perception, emotion or
26
evaluation (LKG III, 1976: 806; Labutis, 2002: 158; DLKG, 2005: 669; LG, 2006: 726). LKG
III (1976) also provides a more detailed listing of verbs taking object that-clauses
corresponding to different nominal cases. According to it, verbs taking object that-clauses in
the genitive case position are certain verbs of mental states and feelings, e.g. bijoti “fear”or
tikėtis “hope”, certain verbs related to speech acts, e.g. reikalauti “demand” or maldauti
“beg”, as well as all transitive verbs in negative forms; verbs taking object that-clauses in the
accusative case position are certain verbs of perception and feeling, e.g. matyti “see” or
stebėti “observe”, verbs of mental processes such as thinking, discovering, or knowledge, e.g.
galvoti “think”, atsiminti “remember”, or žinoti “know”, or verbs related to various speech
acts, e.g. aiškinti “explain” or sakyti “say”; finally, verbs that can be followed by object thatclauses in the instrumental case position can be related to perception and mental states, e.g.
didžiuotis “be proud” or tikėti “believe”, or certain speech act verbs, e.g. girtis “boast” (op.
cit.: 816-817). Holvoet and Judžentis (2003) suggest grouping the matrix verbs taking
complement clauses on the basis of their modality and distinguish matrix verbs of epistemic
(episteminiai), deontic (deontiniai), dynamic (dinaminis), evaluative (vertinamieji) types, as
well as fearfulness (būgštavimo) type (2003: 168-169).
On the whole it can be observed that in both languages under consideration verbs
taking object that-clauses generally belong to the same major semantic domains of saying and
thinking. Labutis (2002) states that it is even more typical to complete the content of
speaking, thinking and similar activities by a clause than by a nominal object, as such content
usually involves a distinct fact “with its own subject and other actants”, and thus, it is naturally
expressed by a clause (2002: 269). In respect to the verbs of saying, Balkevičius explains the
abundance of verbs of saying with complement object clauses by the fact that such clauses are
typically used to express indirect speech (1963: 348). This function of the complement object
clause is also emphasised in respect to asyndetic complement object clauses in Drotvinas
(1961: 191).
2.2.4 Properties of the complementizer that.
The complementizer that used to join the complement clause to the main clause in
Lithuanian has two forms: kad and jog. The same way as English that, both forms are
considered functional conjunctions. They do not possess any distinct semantic meaning, thus
are often referred to as asemantic or polyfunctional, as they only mark subordination but do
not indicate the type of semantic relation between the clauses (LKG III, 1976: 807; Sirtautas
and Grenda, 1988: 153; Labutis, 2002: 154; DLKG, 2005: 661; LG, 2006: 720), which means
that these conjunctions can generally join different types of subordinate clauses. However,
27
Ambrazas (2006a) observes that in the standard language jog is used only in complement
clauses, while kad, the use of which has considerably expanded over time, has become
polysemic (2006a: 26).
As kad and jog are functional conjunctions, the semantic relation between the main and
the subordinate clauses joined by them is revealed only by “the structure of the clauses and
grammatical and semantic properties of the antecedent and other components” (LG, 2006:
721). As it was mentioned in Chapter 2.1.4, most English works on grammar also consider the
conjunction that asemantic and merelyfunctional. Another similarity between English and
Lithuanian conjunctions is that they join the main and the subordinate clause but perform any
syntactic functions in neither of the clauses (LKG III, 1976: 700; Masilionis, 1985: 74;
Holvoet and Judžentis, 2003: 122).
Considering the two Lithuanian equivalents of the complementizer that used in
complement clauses, it is reported that they are often interchangeable (LKG, 1976: 807;
DLKG, 2005: 669; LG, 2006: 726), and both forms together tend to be used in a sentence with
several hierarchically subordinated complement clauses “in order to avoid monotonous
repetition” (LG, 2006: 726). However, kad is much more widespread in standard Lithuanian
(DLKG, 2005: 669). Also, only the conjunction kad can be used after certain verbs (e.g. bijoti
“fear”, reikalauti “demand”or prašyti “ask”) in complement clauses implying the aim and
containing the predicate in subjunctive mood (DLKG, 2005: 669; LG, 2006: 726), as
illustrated in example (17) taken from DLKG (2005: 669):
[(I) asked that (they) should let me out.]16
(17) Prašiau, kad mane išleistų.
16
In English such complement clauses typically are of to-infinitive type.
28
3. THE DELETION OF THAT IN COMPLEMENT CLAUSES
3.1 The deletion of the complementizer that in English
Complement that-clauses with the complementizer omitted are equally grammatical in
English as those joined by that. The omission of the complementizer is in most cases
considered a free choice and does not change the meaning of the whole sentence (LGSWE,
1999: 680; Biber, 1999: 144), which means that the structures “can be used interchangeably”
(Elsness, 1982: 1). With this in mind, the factors favouring the choice of one structure over
the other have been of considerable interest for modern linguists applying corpus-based
methods, as such methods allow to detect and evaluate different factors influencing the choice
simultaneously. This section of the paper will provide a review of corpus-based studies
conducted on the issue of that deletion in complement clauses.
The criteria for the omission or retention of the complementizer that can be defined as
“partly stylistic and partly grammatical” (McDavid, 1964: 113). The grammatical factors,
such as the syntactic function of the complement that-clause and structural ambiguity, were
described above in 2.1.2. In this section the focus will be laid on stylistic and contextual
factors determining the omission or retention of that in complement clauses revealed by
corpus-based studies.
An early corpus-based study on alternation of that and zero17 in complement clauses
was conducted by McDavid (1964). In her research carried out on a 100,000-word corpus of
American fiction written in 1950-s, she focuses, firstly, on the types of constructions using
that or zero to introduce a complement clause, and, secondly, on the circumstances of that
deletion. 18 Among the constructions under consideration detected in the corpus the author
reports the complement object clause of a transitive verb as the most frequent in general, as
well as with zero that. Among the stylistic factors favouring the deletion of that she names
common matrix verbs (especially know and say) and informality of language (op. cit.: 113).
The influence of style is also considered one of the basic factors influencing the
omission or retention of that in complement clauses by Storms (1966) in his study on
complement that-clauses. According to him, that deletion adds an “element of subjectivity”,
while that retention makes the clause “less personal, less familiar, less warm, less friendly,
less emotive”, but, instead, “objective, factual, formal, official” (op. cit.: 262). In the scholar's
17
The term 'zero' for the absence of the complementizer that is used in McDavid (1964) as well as in numerous
other works on the issue.
18
Other terms used as synonyms for zero that are that deletion, elision or omission.
29
view, it is the amount of subjectivity that distinguishes colloquial spoken language from
“formal, written or elevated speech” (op. cit.: 262).
A thorough examination of factors conditioning that deletion is made by Elseness
(1982; 1984) in his works on the choice of that or zero that in complement clauses and, in
particular, object that-clauses studied in four types of texts in 128,000-word Syntax Data
Corpus, a part of Brown University Corpus of American English published in 1961. Elsness
emphasises that a number of factors can simultaneously have an impact on the choice of the
form of the complementizer (1984: 533). For complement clauses in general, Elsness
identifies the following factors favouring that deletion (1982: 39-41):
1. informality of style;
2. syntactic function of the objectof the matrix verb (with object that-clauses being the
most frequent type of complement clauses with zero that);
3. no potential ambiguity (zero that is more frequent when there is no intervening
adverbial between the matrix verb and the subject of the complement clause);
4. lack of structural complexity near clause boundary (zero that is more likely when
the subject of the object that-clause is realised by a pronoun, not a noun phrase);
5. no deviation from the weight distributional pattern common in English (light
elements at the beginning and heavier elements at the end of both matrix and
complement clauses);
6. closeness of the clause juncture: zero that is more likely when:
•
the subject in the object complement clause is pronominal and coreferential with the
subject in the matrix clause;
•
the subject in the object clause has definite, often anaphoric, reference;
•
subjects in both matrix and object clauses are in the 1st or 2nd person.
In this study, the factors 1.-3. confirm the findings of earlier research, while 4.-6., typical of
object that-clauses, appear to be novel observations (1984: 533). Most of the findings
reported by Elsness were later proved by other corpus-based studies.
Thompson and Mulac (1991) study the discourse conditions of the use of the
complementizer that in spoken American English. Their corpus consists of 240,000 words of
American college students' conversations. The results of this quantitative study reveal that
object that-clauses with zero that are most frequent with the matrix verbs think and guess
(which conforms to the results of earlier studies) in the 1st or 2nd person singular and a
pronominal subject in the complement clause, while the retention of that is more likely in case
the matrix clause includes an adverbial, an auxiliary or an indirect object (which is also in
30
accordance with the results of the previous study by Elsness (1984)). The explanation
suggested by Thompson and Mulac is that certain combinations of the subject and verb in the
main clause like, for instance, I think should be reanalyzed as “unitary epistemic phrases”,
(1991: 237), with the main idea of the sentence being expressed by the complement clause.
According to them, “As this happens, the distinction between 'main' and 'complement' clause
is being eroded <…> with the omission of that a strong concomitant” (1991: 249).
Besides synchronic investigation, there have also been diachronic studies tracing the
historical development of the zero and full forms of the complementier that. Rissanen (1991)
investigates zero and that as object clauses links on the basis of the data from Helsinki Corpus
of English Texts covering the period from Late Middle English (1350-1420) to Early Modern
English (1640-1710). He claims that both zero that and that as object clause links have
existed in English for centuries and represent two variant links (1991: 288). He expresses the
intuition that in spoken mode zero “may <...> have been the unmarked object clause link
throughout the history of English” (op. cit.: 287), and thus, the very term 'omission' of the
complementizer is incorrect. However, the growing frequency of zero in written English,
where that usedto be “the unmarked link”, allows to interpret this novel tendency “as the
'omission' of the expressed conjunction” (op. cit.: 288). In his study, Rissanen also reports a
consistent general increase of the zero that over the period of time under consideration after
the matrix verbs say, tell, know, and think and lists the following factors favouring the choice
of zero that:
•
a pronominal subject in the object clause;
•
the lack of intervening elements between the matrix and the object clauses;
•
the spoken mode of communication (op. cit.: 286).
Another diachronic study on the two forms of complementizer that in complement
object clausesis carried out by Finegan and Biber (1995). They study the use of the
complementizer in three registers of British English in the period from 1650 to 1990 on the
basis of the data from approximately 1,7 million-word ARCHER Corpus and show that the
tendency of the increasing use of zero that in late Middle and early Modern English reported
by Risanen is opposite in some registers, viz. sermon, medical articles, and letters (1995:
247). The findings of this research generally confirm the conditioning factors of the use of
that and zero named in previous studies by McDavid (1964), Elsness (1984) and Thompson
and Mulac (1991), except that the results of this particular study do not show the
coreferentiality between the subjects of the main and the complement clause as influential in
31
the choice of zero that, as it was earlier suggested by Elsness (1984)(op. cit.: 255)19.
While most of the corpus-based studies on the issue of omission and retention of that
in complement clauses concentrate on object that-clauses, Kaltenböck (2004) investigates the
choice of the form of the complementizer in extraposed subject clauses on the basis of the
data from spoken and written texts in the British component of the International Corpus of
English (ICE-GB). His findings also confirm the lack of formality as a significant factor in
the choice of zero over that, though the mode of communication (spoken vs written) is not
identified as a decisive factor (2004: 65). Another result conforming to those obtained in
studies of object clauses is that the absence of intervening adverbials at clause boundary tends
to favour that deletion. Kaltenböck also reports a high percentage of zero that in the registers
of Reportage (press news reports) and Scripted monologue (broadcast news/talks) in
constructions with a passive reporting verb without agent (for instance, It is said). In the
scholar's view, in these structures the complement clause typically presents new information,
while the main clause merely serves as a reporting frame, thus, the omission of that weakens
the hierarchical syntactic difference between the main and the complement clause and “this
results in reanalysis of the complement clause as a non-embedded constituent in
supplementation relation to the reporting frame, which in turn is reinterpreted as
parenthetical” (op. cit.: 65)20.
A recent corpus-based study by Kearns (2007) focuses on the use of zero that in
complement clauses after epistemic verbs. Her study is conducted on the basis of the corpus
of 2,178 tokens of complements to epistemic verbs, collected from eight broadsheet
newspapers from the USA, the UK, Australia, and New Zealand. The study is inspired, on the
one hand, by numerous reports of the epistemic verbs think and guess among the most
frequent verbs with zero-that complements, and, on the other hand, by the hypothesis
advanced in Thompson and Mulac (1991) that combinations like I think should be reanalyzed
as epistemic phrases and the deletion of the complementizer is the result of such reanalysis.
Kearns argues that such syntactic reanalysis is debatable and unnecessary, and aims at proving
that high zero-that rates with epistemic verbs are “triggered by their use with modifier senses,
in contrast to main verb senses” (Kearns, 2007: 489) and that modifying phrases I think and I
guess in constructions with zero-that complements have become “fixed routines, resulting in
19
The results of a more recent corpus-based study by Kearns (2007) also showed that coreferentiality does not
affect zero rates in complement clauses.
20
Kaltenböck adopts the idea of such reanalysis from Thompson and Mulac (1991).
32
higher zero rates than other other modifier uses of epistemic verbs” (op. cit.: 501). The author
arrives at a conclusion that structures like I think can be used in modifier sense with or
without that and function as a “hedged assertion” of the information conveyed by the
complement clause; however, zero-that in such context is more typical because “the raised
status of the embedded clause as the content of the main assertion is more consistent with the
absence of overtly marked subordination” (op. cit.: 501). The results based on the statistical
data obtained in the process of this research correlate with most findings of previous research
on the that-complement clauses, as well as reveal some novel facts about the regional
differences in the use of zero that complements: the rate of zero that appeared to be the
highest in the texts from New Zealand, significantly lower in Australian and American texts,
and the lowest in the texts from the UK (op. cit.: 495).
There have been a considerable amount of current corpus-based studies on
subordinate and complement clauses confirming the majority of the earlier results discussed
above. Greenbaum et al (1996) report formality in both the spoken and written modes as a
factor influencing that retention in complement clauses (1996: 90). Kirk (1997), Hudson-Ettle
(2002), and Kearns (2007) observe that zero that is generally more frequent in the spoken
mode of communication (Kirk, 1997: 360, Hudson-Ettle, 2002: 261, Kearns, 2007: 490).
Biber (1999) distinguishes two types of factors influencing the choice of the form of the
complementizer: stylistic (zero that prevails in conversation while that in academic writing),
and textual (zero that is typical of the most common form of that-clauses: with matrix verbs
think or say21 and coreferential subjects in the matrix and complement clauses, while that is
more frequently chosen in coordinated that-clauses, as well as when the matrix verb is in the
passive form or there is an intervening noun phrase between the matrix verb and the
complement clause ) (1999: 144-147). The same factors are reported in LGSWE (1999: 680681). The data reported by Kearns (2007) also prove that with a particular class of epistemic
verbs intervening elements between the main and the complement clause, such as indirect
objects and matrix adverbials also favour that retention, while pronominal subjects in the
complement clause favour zero that (Kearns, 2007: 491-492).
All in all, there have been different perspectives on studying the form of the
complementizer that in complement clauses as well as attempts to give different explanation
to certain factors influencing the omission or retention of that in complement object clauses.
However, some basic observations have been proved by a number of researches on the issue,
21
Also proved in Rissanen (1991) and Kearns (2007).
33
and the most influential factors favouring the choice of zero that can be summed up as
follows:
•
the spoken mode of communication and informal style;
•
common (highly frequent) matrix verbs (such as say and think);
•
absence of intervening elements between the matrix verb and the subject of the
complement clause;
•
pronominal 1st and 2nd person subjects in the main clause;
• the pronominal subject in the complement clause (coreferential with the subject of the
main clause).
3.2 The deletion of the complementizer that in Lithuanian
Differently from the English language, the term omission or deletion of the
complementizer that is not common in Lithuanian works on linguistics, though occasionally
the absence of the complementizer is referred to as omission (Miliūnaitė, 2005). However,
instead, composite sentences are usually divided into syndetic or conjunctive and asyndetic or
conjunctionless (LG, 2006: 717; DLK, 2005: 658; Sirtautas and Grenda, 1988: 142; Labutis,
2002: 147). Considering complex and compound asyndetic sentences there have been
different opinions on their status. Some scholars, especially in earlier works on grammar,
distinguish complex and compound sentence classes, both of which in turn could fall into
either syndetic or asyndetic (Drotvinas, 1961: 178; Balkevičius, 1963: 321; Dambriūnas,
1963: 69). However, recently there has been a tendency in Lithuanian grammar to distinguish,
at the highest level of composite sentences, the class of asyndetic as opposed to syndetic
composite sentences and within the former to observe subclasses which only resemble or are
comparable to compound and complex sentences in respect to the relation between the
clauses, but are not the same as syndetic complex or compound sentences with conjunctions
omitted (LKG III, 1976 918-919; Sirtautas and Grenda, 1988: 143; Labutis, 2002: 172).
Holvoet (2003) even suggests that asyndetic relation between clauses in composite sentences
is generally of the same nature as in compound sentences, though some asyndetic sentences
functionally rather resemble complex sentences (2003: 112). However, on the whole most
grammarians agree that in a part of asyndetic sentences the relations between clauses are of
the same nature as in complex clauses, and, consequently, a class of clauses resembling
complement clauses as well as complement object clauses, in which the second clause
completes the meaning of an element in the first clause, can be distinguished in asyndetic
34
sentences as well (Drotvinas, 1961: 191; Balkevičius, 1963: 322; Dambriūnas, 1963 69; LKG
III, 1976: 937; Labutis, 2002 146; LKG, 2005: 729; LG, 2006: 777).
There is no account in Lithuanian grammatical studies about the factors influencing the
choice of asyndetic complement structures versus syndetic; only some properties of asyndetic
sentences are described in various works. Drotvinas (1961) in his study on complex asyndetic
sentences mentions that they are much more infrequent than their syndetic equivalents and
often lack explicit criteria for classification (1961: 185) 22 . According to him, asyndetic
complement clauses functioning as objects are most often with matrix verbs of speech acts
and perception with these verbs typically used in present or past tense first or third person
singular form (op. cit.: 191). He also mentions, that such sentences can be used to express
indirect speech (op. cit.: 191). LKG III (1976) reports that the meanings of asyndetic
sentences can be different and it is often difficult to differentiate them, thus, asyndetic
sentences are hardly ever used in academic and formal language and are more typical of the
spoken mode of communication as well as of fiction and folklore (1976: 917). It also lists the
verbs of saying, thinking, and perception as the most frequent matrix verbs with asyndetic
object complement clauses, especially when the verb is positioned at the end of the main
clause and thus emphasises the link between the clauses (op. cit.: 937). It should be mentioned
that these observations generally resemble the stylistic and structural factors favouring the use
of zero that in complement clauses in English.
Interestingly, most grammars do not comment on the punctuation rules for asyndetic
sentences. Dambriūnas (1963) claims that the asyndetic complement sentences are separated
by the comma or the dash (1963: 69). However, the examples provided in numerous works
show that the clauses in asyndetic sentences are most often separated by the colon or the dash,
and less often by the comma. The vast majority of examples used to illustrate asyndetic
complement object clauses employ the colon, as in example (18a). However, there are also
occasional examples that employ the comma, as in example (18b), or the dash, as in example
(18c).
(18) a. Edvardas iš karto suprato: šeimai trūksta pinigų.
(LKG III, 1976: 937)
[Edvardas instantly realised the family needed money.]
b. Žinau, tu atrasi namo kelią.
(Dambriūnas, 1963: 69)
[(I) know you'll find the way home.]
c. Ar tu manai – labai malonu lyginti skaisčią, gražią praeiti su juoda, murzina
22
Sirtautas and Grenda (1988: 143) also consider asyndetic sentences corresponding to complex sentences as a
minor class.
35
dabartimi?
(LKG III, 1976: 937)
[Do you think it is very pleasant to compare the innocent beautiful past to the black
dirty present?]
Differently from the Lithuanian language, in English the main clauses are not separated from
complement clauses by any punctuation marks.
To sum up, strictly speaking, the term 'omission' of complementizer that in
complement clauses is not absolutely correct for the Lithuanian language at least in the
modern grammatical trends. Asyndetic sentences are treated as a separate class; however
within it there are groups of sentences exhibiting the same structural and semantic relations
between the clauses as in complex sentences in general and in complement that-clauses in
particular. Similarly to English zero that complement clauses, Lithuanian asyndetic
complement clauses are considered more typical in spoken and less formal language and are
reported to be frequently used with verbs related to speech acts and mental states or activities.
However, there has been no corpus-based research on Lithuanian syndetic and asyndetic
complement clauses, which could reveal more about the features of these clauses and the
factors favouring the choice of one type of structure over the other.
36
4. CORPUS-BASED CONTRASTIVE ANALYSIS OF THAT-DELETION IN VERBAL
COMPLEMENT CLAUSES IN ENGLISH AND LITHUANIAN
Due to the limited scope of the research the present study is concerned only with a
certain subclass of complement that-clauses of English and Lithuanian, namely, verb
complement that-clauses, which are reported as the most frequent complement clauses in
English (Greenbaum et al, 1996: 81; LGSWE, 1999: 674; Biber, 1999: 133). In addition, only
those functioning as direct objects were investigated, as this type of that-clauses is reported as
the most frequent type of verb complement clauses in English (McDavid, 1964: 108; Elsness,
1882: 1; Greenbaum et al, 1996: 86; Hudson-Ettle, 2002: 265) and the most frequent type of
subordinate clauses in Lithuanian (LKG III, 1976: 815; Labutis, 2002: 269; DLKG, 2005:
672). The choice to investigate the object that-clauses was also influenced by the fact that
these English structures, as well as the conditions under which the complementizer that can
be omitted in them, have been mostly studied and described in corpus linguistics.
The investigation was also limited to the object that-clauses complement to the verbs
belonging to the semantic domains of saying, thinking, and discovering. These three groups
are the major semantic groups of verbs taking that-complements (CGP, 1998: 98-100). Also,
the verbs say, think, and know, which belong to the semantic domains under consideration, are
reported as the most common English verbs taking zero-that complements (Biber, 1999: 137;
LGSWE, 1999: 663; Rissanen, 1991: 276). Lithuanian matrix verbs taking object thatcomplements can also mostly be ascribed to theses semantic domains (LKG III, 1976: 806;
Labutis, 2002: 158; DLKG, 2005: 669; LG, 2006: 726).
The corpus-based part of the study employed the following procedure. Firstly, the lists
of English and Lithuanian verbs of saying, thinking, and discovering taking complement thatclauses were produced23.The initial lists of English verbs were taken from CGP (1998: 98100), as these lists of verbs taking complement that-clauses are the product of a detailed
corpus-based study of verb complements in English. Only the verbs which can take
complement that-clauses in their main (the first and second) meanings were selected. In order
to produce the lists of the Lithuanian verbs of saying, thinking, and discovering the English
verbs were translated into the Lithuanian language 24 and the Lithuanian equivalents were
checked in the verb valency dictionary25 in order to select the verbs which can be followed by
23
24
25
See the lists in Appendix I.
The translation was made using Piesarskas (1998).
Sližienė (1994; 1998; 2004).
37
complement that-clauses in Lithuanian as well as to detect synonyms which were not
included in the list produced by the translation. The verbs which can take complement thatclauses in their main (the first and second) meanings were selected. For both the English and
Lithuanian languages verbs that can be attributed to more than one semantic domain were
included into the domain of their main meaning.26
Secondly, the English and Lithuanian verbs were checked in the corpora employed in
the research, namely, the British National Corpus (BNC http://corpus.byu.edu/bnc/) and the
Corpus of the Contemporary Lithuanian Language (LLC http://donelaitis.vdu.lt), to ensure
that the verbs occur in these corpora with that-complements. On the whole, the final lists of
matrix verbs under study comprised 87 English and 75 Lithuanian verbs of saying, 52 English
and 51 Lithuanian verbs of thinking, and 22 English and 21 Lithuanian verbs of discovering,
which makes the total number of 161 English and 147 Lithuanian verbs. After that, the
corpora were investigated in order to detect and collect the examples of complex sentences
with these matrix verbs followed by zero-that object complements. The following
corresponding text categories of the corpora were explored: Spoken, Fiction, Newspaper and
Magazine in the BNC and Sakytinė kalba [Spoken], Grožinė literatūra [Fiction], and
Publicistika [Journalism] in the LLC27.
Finally, the collected evidence was examined, the cases of zero-that complement
clauses in English and Lithuanian were described and compared; the similarities and
differences in the use of zero-that complements were revealed, described and interpreted on
the basis of the corpus evidence taking a qualitative approach in analyzing and describing the
evidence and employing the methods of contrastive analysis and content analysis. In
analyzing and interpreting the data a special attention was paid to the structural and semantic
peculiarities of English and Lithuanian zero-that complement clause as well as semantic
domains of the matrix verbs and the text categories (or registers).
4.1 The deletion of the complementizer that in complement clauses after the verbs of
saying in English and Lithuanian.
The verbs of saying under consideration “are concerned with speaking writing and
other forms of communication” (CGP, 1998: 98). The corpus evidence collected revealed the
following facts about the omission of the complementizer that in verb complement clauses in
the two languages under consideration.
26
27
For the list of verbs belonging to more than one semantic domain see Appendix I.
Henceforth, in the work the sections of the LLC are refered to as spoken, fiction and journalism.
38
The most obvious observation is that the omission of the complementizer that is by far
more common in the English language than in Lithuanian. From the 87 English verbs of
saying studied, for the following 29 verbs examples of zero-that complements were detected
in all registers investigated28:
admit, allege, announce, claim, complain, confess, declare, decree, deny, emphasize,
explain, hint, imply, indicate, insist, maintain, mention, pray, predict, promise, propose,
recommend, report, reveal, say, stress, suggest, swear, warn.
Structurally, for the most of these verbs the detected examples exhibited a considerable
variety of the tense forms of the matrix verb as well as of the types of subjects in the main and
the complement clauses, which can be illustrated with the following examples of the verb say
with zero-that complements from the BNC29:
(19) a. Sergeant Allen Bovington Cox says specialist teams are involved.
(spoken)
b.And they said it was our car.
(spoken)
c. And Isaid I'm not talking about the council <…>.
(spoken)
d. The hotelier said it was Italy.
(fiction)
e.You said you were thrown out by your landlord.
(fiction)
f. But he could equally well have said one is enthusiasm and the other is cynicism,
<...>.
(fiction)
g. At least one member of the jury said he would not have given a sentence of
death.
(magazine)
h. No one' s saying the Escort 1.4LX is a bad car, <...>.
(magazine)
i. <…>Antoniasaidshe was upset and worried <…>.
(newspaper)
As the examples demonstrate, the subjects in both the main and the complement clauses can
be pronominal as in (19b), (19c), (19e), and (19f), or they can also be expressed by a noun as
in (19i), or a noun phrase as in (19a) and (19g). For some verbs the type of the subjects in the
main and the complement clauses is influenced by semantic peculiarities of the verbs. For
instance, the verbs argue, decree, demand, plead, prophesy, recommend, request, stipulate, or
28
29
The registers of magazine and fiction of the BNC were considered together as one, as there is one section of
journalism in the LLC, which can be comparable with these two.
If not specified, all English examples in the research are taken from the BNC and all Lithuanian examples
from the LLC corpora.
39
submit do not occur with coreferential subjects, as actions expressed by such verbs involve
two or more participants, and are not applied to oneself, as illustrated in example (20):
(20) a. The nature of Alex's profession demanded she kept a certain amount of
medical equipment around.
(fiction)
b. Gascoigne went up to each player <...> and demanded he die for the club and the
cause.
(newspaper)
c. Minister Ronchey decrees they are to stay open ten hours a day.
(magazine)
d. What changes would you recommend I make?
(spoken)
As the present study does not involve counting frequencies, it is difficult to measure
how significant the type of the subject in the main or the complement clause is for the choice
of zero-that complement. However, it should be noted that the corpus evidence suggests that
pronominal subjects in both the main and the complement clauses, as well as coreferential
subjects, as in (19c), (19e), (19g), and (19i), are frequent with matrix verbs of saying followed
by zero-that complements, which is in complete accord with the observations made in corpusbased studies of English zero-that complement clauses (Elsness, 1984: 533; Thompson and
Mulac, 1991: 237; Rissanen, 1991: 288; Finegan and Biber, 1995: 255, Biber, 1999: 144).
In respect to contextual factors, some verbs have showed preference to be used with
zero-that complements in certain registers. For instance, the verbs advise, boast, brag,
concede, remark, signify, stipulate, and wager were not detected with zero-that complements
in the spoken register, but occurred in the fiction and magazine and newspaper registers.
Some cases are illustrated in (21):
(21) a. She was waved on by a sharp-eyed young officer, who boasted he could smell
a smuggler from fifty yards away.
(fiction)
b. <...> we visit his home... still filled with hundreds of beautiful artefacts which
proudly boast they were Hand Made in the Cotswolds.
(newspaper)
c. I remarked it was strange that when a person was in trouble, one forgot at once
any little grudge one had against them, <...>.
(fiction)
d. I remarked it was an experience I would not care to repeat.
(fiction)
e. I'll wager you've forgotten.
(fiction)
f. I'll wager you would need to be pointing almost vertically downhill to achieve
such a mind-numbing speed.
(newspaper)
40
Other verbs, such as assert, beg, proclaim, prophesy, and request were detected with zero-that
complements in the fiction register only. Some cases are illustrated in example (22):
(22) a. I beg you'll send me out of my way to my poor parents.
(fiction)
b. Suppose he needed you there toassert he'd walked into a trap?
(fiction)
c. They had one likeness of Oreste, sent at Christmas when Wilson had requested it
should be taken.
(fiction)
Different distribution of these verbs across the registers can be explained by the fact that they
belong to more formal style and are generally infrequent in spoken language, on the one hand,
and, on the other hand, that some of them are less typical to journalistic articles and more
typical to works of fiction, the language of which is usually not as formal as the language of
newspapers or magazines.
There were also two substantial groups of matrix verbs which either only occasionally
occurred with zero-that complements:
affirm, argue, assert, beg, brag, caution, comment, concede, confide, contend,
demand, dictate, disclose, forecast, grumble, insinuate, joke, moan, ordain, plead, pledge,
preach, proclaim, prophesy, remark, request, signal, signify, specify, stipulate, submit,
threaten, wager,
or were not found in the corpus followed by zero-that complements30:
advocate, attest, aver, command, divulge, enthuse, foretell, grouse, instruct, lament,
opine, order, posit, postulate, profess, quip, remonstrate, sneer, testify, underline,
underscore.
An important fact is that these are the most infrequent verbs of saying in the BNC corpus,
with the general number of occurrences of each of these verbs across the corpus not
exceeding 2,000 and in most cases even being lower than 1,000. This observation generally
confirms the findings of previous corpus-based researches, which claim that rare verbs, which
are typically stylistically marked and more formal, do not tend to be used with zero-that
30
The verbs in bold are the most infrequent in the corpus from the verbs of saying under investigation; thus, it
was possible to check all their occurrences and ensure these verbs have no instances of zero-that
complements in the BNC.
41
complements while common matrix verbs favour the deletion of that (McDavid, 1964: 113;
Biber, 199: 147; LGSWE, 1999: 680).
Last but not least, there is one more distinct feature of zero-that complements with the
matrix verbs of saying in English, which becomes obvious from the corpus evidence. It is
reported by Elsness (1982) and is labeled by him as closeness of the clause juncture (1982:
41). As examples (19)-(22) above illustrate, on the one hand, zero-that clauses typically
immediately follow the matrix verbs, and, on the other hand, pronominal subjects in the
complement clause have definite anaphoric reference.
Regarding the Lithuanian verbs of saying followed by zero-that complements it can be
said that rare verbs also did not show any tendency for the omission of the complementizer
kad/jog “that”. The following verbs occured with zero that in rare cases:
pagrasinti “threaten”, įspėti “warn”, komentuoti “comment”, postuluoti “postulate”,
prisiekti “swear”, susitarti “agree”.
Also, the most infrequent verbs in the LLC listed below, with the general number of
occurrences across the corpus not exceeding 2,000 for each verb, have no instances of zerothat complements in the corpus:
anonsuoti “announce”, bambėti “grouse”, bėdoti “moan”, burbtelėti “grumble”,
dievagotis “swear”, dievažytis “swear”, ginčytis “argue”, įsakyti “command”, pamokslauti
“preach”, propaguoti “advocate”, protestuoti “remonstrate”, sielotis “lament”, signalizuoti
“signal”.
As it was mentioned above, Lithuanian verbs of saying generally are more infrequent
with zero-that complements than their English equivalents. In comparison to 29 out of 87
English verbs of saying detected in the BNC with zero-that complements in all registers, only
6 out of 75 Lithuanian verbs of saying could be found in the LLC followed by complement
clauses with the complementizer that omitted in all three registers under consideration:
pripažinti “admit”, prisipažinti “confess”, pažadėti “promise”, kalbėti “speak, say”,
šnekėti “speak, say”, and sakyti “say”,
and only one of them, namely sakyti, resembled the cases of zero-that complements in
English in respect to the variety of the tense forms of the verb and the types of subjects, as
42
illustrated in example (23):
(23) a. - Ai, gerai tu sakai, ta mergaitė kažkaip tai mirs kaip ir reklamava, <...>.
(spoken)
[- Oh, you rightly say the girl is going to somehow die, as it was announced <...>.]
b. Graboriai sako, žmonės gimsta tam, kad mirtų, <...>.
(fiction)
[Coffin makers say people are born to die, <...>.]
c. Aš tau sakau, stalas šoko!
(spoken)
[I'm telling you the table was dancing!]
d. Tu sakai, kėlei ranką prieš ją. Kaip? Kodėl?
(spoken)
[You say (you) raised your hand against her. How? Why?]
e. Taigi, jie sako, nereikia jo spausti prie sienos, <...>.
(journalism)
[So, they say (we) shouldn‘t push him to the wall, <...>.]
f. Ir jinai sakė, mes pabūsim pas ją, išgersim šampano, <...>.
(spoken)
[And she said we'd stay at hers, have some champagne,<...>.]
It must be mentioned that in English the corresponding verb say is reported as one of the most
frequent matrix verbs taking zero-that complements, thus the Lithuanian verb say is to a
certain extent similar in this respect.
Another surprising difference between the English and Lithuanian zero-that
complement clauses after the verbs of saying is that only few cases of zero-that complements
were detected in the spoken register of the LLC, which is completely opposite to the
tendencies reported in numerous corpus-based studies of English complement clauses, where
the spoken mode of communication, which is characterised as less formal than the written
mode, is named as one of the most influential factors on the omission of that (Storms, 1966:
262; Elsness, 1982: 39; Rissanen, 1991: 286; Kirk, 1997: 360, Biber, 1999: 144; HudsonEttle, 2002: 261, Kearns, 2007: 490). This also contradicts the general assumption that
Lithuanian asyndetic sentences are more common in the spoken mode of communication
(LKG III, 1976: 917: DLKG, 2006: 721). However, the humble amount of zero-that
complements in the spoken register of the LLC might be misleading and does not necessarily
imply that the omission of that is not characteristic of the spoken mode of communication in
Lithuanian. The low number of zero-that can be due to the fact that the spoken section of the
43
LLC is relatively new and still small in size in comparison to the rest of the corpus31. Thus,
many verbs, especially less frequent ones, do not occur in the spoken section of the LLC at
all. Therefore, at the moment, the cases of zero-that complements in Lithuanian can hardly be
properly studied in the spoken register of the LLC and compared to the cases of zero-that
complements in the BNC, where the spoken section is much wider and represents a
considerable variety of spoken texts in different registers, such as, for instance, conversations,
meetings, or broadcast news among others32.
On the whole, the examples of zero-that complements in Lithuanian that mostly
resemble the English counterparts confirm the tendency observed by Drotvinas (1961): the
matrix verbs are typically used in the present or past tense first or third person singular form
(1961: 191). They also conform to the feature of the asyndetic object clauses described in
LKG III (1976): the matrix verb is positioned at the end of the main clause (1976: 937). This
feature resembles the one defined by Elsness (1982) as closeness of the clause juncture (1982:
41) and is illustrated in the following example (24):
(24) a. Jis sako, mudu tikrai galėtume tai padaryti, <...>.
(ficiton)
[He sayswe two could really do that <...>. ]
b. Aš prisipažįstu, aš kaltas.
(spoken)
[I admit I'm guilty.]
c. Pats pripažįstu, tuomet aš buvau lyg apakęs.
(journalism)
[I admitI was quasi blind then.]
d. Aš tau byloju, kai kurios iš jų reformų yra tiesiog pasišlykštėtinos <...>. (ficiton)
[I'm telling you some of their reforms are just revolting <...>.]
e. Esu garbingas žmogus, todėl pažadu, jais nepasinauduosiu.
(journalism)
[I'm a respectable man, therefore (I) promise I won't use them.]
f. Tik įspėju, ta būtybė nepakenčiama!
(ficition)
[But I warn the creature is intolerable!]
The sentences in examples (24b), (24c), and (24e) also illustrate that subjects of the main and
the zero-that complement clauses in Lithuanian can be coreferential. However, the corpus
evidence from the LLC on the whole does not suggest that this is a typical feature of
Lithuanian zero-that complement clauses after the verbs of saying. The subject in the
31
32
In the LLC, which consists of approximately 140 million words, the spoken section comprises only 0.3% of
the corpus (557 822 words).
In the BNC, which consists of 100 million words, the spoken section comprises 10% of the corpus.
44
complement clause is often not coreferential with the subject in the main clause, as in (24a)
and (24d) and often is not even pronominal, as in examples (24d) and (24f) or (23a)-(23c). In
English, on the contrary, pronominal subjects in the complement clauses are reported as a
factor favouring that deletion. On the other hand, in Lithuanian the subject of the complement
clause, even if not pronominal, usually implies anaphoric reference, as in examples (24e) and
(24f), which corresponds to the tendency reported by Elsness (1982: 41). However, only a
quantitative corpus-based study could reveal how significant the type of the subjects in the
main and the complement clauses is for the choice of the zero-that complementizer in
Lithuanian.
Furthermore, the corpus evidence suggests that in Lithuanian the deletion of that in
complement clauses after the verbs of saying is influenced by structural peculiarities of the
clause. There is an observable tendency for the choiche of zero-that complementizer in case
the main clause is preceded by particles esą or neva, which both can be translated as
“ostensibly, allegidly, supposedly” or a parenthetical girdi “(you) hear”, which is very close
in meaning to the mentioned particles. The sentences in example (25) illustrate such cases:
(25) a. Dar kiti teigė, neva vaikus parduoda vergais turtingiems totoriams. (fiction)
[Still others claimed (?) the children were sold as slaves to rich Tatars.]
b. Jaunuoliai <...> ligoninėje melavo, esą juos sumušė chuliganai.
(journalism)
[In the hospital the young men <...> lied (?) they had been beaten by hooligans.]
c. Kai kurie galbūt pakomentuos, girdi, sekso reikaluose svarbu vien meistriškumas.
(fiction)
[Some might comment (?) only excellence matters in sex affairs .]
The particle esą is defined as a particle used to introduce the indirect speech (LKŽ, Wiemer,
2007: 177), the latest edition of DLKŽ (2012) classifies it as a modal lexical element (2012:
153); neva is defined as a particle used to express doubt (LKŽ; DLKG, 2005: 435; DLKŽ,
2012: 423), and girdi is classified as a parenthetical (DLKŽ, 2012: 177) or as a particle (LKŽ)
with the meaning synonymous to esą and used to report somebody's words (Wiemer, 2007:
177). On the whole, in the most recent studies all these three lexical items have been ascribed
to lexical means of expressing evidentiality33 (Ruskan, 2010: 7). Wiemer (2007) ascribes both
particles esą and neva, as well as the parenthetical girdi to the class of hearsay or reportive
33
According to Ruskan (2010), in contemporary linguistics evidentiality is considered a semantic (cognitive)
grammatical category used to indicate the cognitive or/and communicative basis of the speaker's utterance
(2010: 1).
45
evidential markers ( 2007: 177-180).
It has been generally assumed that the particles esą and neva, when operating over the
whole clause 34 , can function as complementizers (Būda, 1986: 55; Wiemer, 2007: 178,
Ruskan, 2010: 8); Būda (1986) also considers the parenthetical girdi as a possible
complementizer (Būda, 1986: 56). This would explain the omission of that in complement
clauses preceded by esą, neva, and girdi. On the other hand, corpus evidence shows that
complement clauses preceded by these lexical items also occur together with the
complementizer kad, as illustrated in example (26):
(26) a. O paskui <...> pareiškė, kad neva aš ir jos vyras esam kažkuo panašūs!
(fiction)
[And then <...> claimed that (?) me and her husband were similar in a way!]
b. Daugelis vairuotojų skundėsi, kad neva policininkai juos nubaudė neteisingai.
(journalism)
[A lot of drivers complained that (?) the police officers fined them unduly.]
c. Vyras <...> trumpai paaiškino, kad esą dabar dar ne laikas kalbėti apie užmokestį.
(fiction)
[The man <...>shortly explained that (?) it's not the time yet to speak about the pay.]
d. <...> pacifistai susirūpinusiais veidais aiškina, kad esą negalima erzinti Rusijos,
reikia su ja draugauti.
(journalism)
[<...> anxcious-faced pacifists comment that (?) we mustn't tease Russia, we have to
get on well with it.]
e. <...> Viktorijos tėvai į tokius žodžius atrėžė, kad, girdi, šis jau niekados
nebeparvyks.
(fiction)
[<...> Viktoria's parents retorted to these words that (?) he would never return again.]
f. Sibiriečiai ir šiais laikais juokauja, kad, girdi, jų krašto „įstatymas – taiga,
prokuroras – meška“.
(journallism)
[Siberian people even nowadays joke that (?) in their region “the law is taiga, while
the prosecutor is a bear”.]
This structural duality might be explained by an assumption that such lexical items,
34
These particles can also operate only over a part of the clause, for instance, a noun phrase :
Užuominomis jam pasakė, kad turėsiąs mokėti pinigus esą už blogą tarpininkavimą.
Using allusions he told him that he would have to pay money supposedly for bad mediation.
(Wiemer, 2007: 177)
46
permanentently positioned next to the complementizer kad/jog, over the time supersede it
(Būda, 1986: 57). However, in a language, the process of lexical items changing their
grammatical status and functions is very slow, uneven and gradual, therefore the co-existence
in a language of older and newer structures at the same time is an unseparable feature of the
syntax development (Ambrazas, 2006a: 25). Another fact signalling that these lexical items
still have not completely acquired the function of the complementizer is that they are
occasionally used in sentences with the matrix clause separated from the complement by the
colon, as illustrated in example (27), which happens only in asyndetic, i.e. conjunctionless
sentences:
(27) a. Pryn kartą su juo susiginčijo: esą barai jai patinka, <...>.
(fiction)
[Once Pryn argued with him (that) she likes bars, <...>.]
b. Jūs imsite prieštarauti: girdi, dar ne ledų metas, bet aš taip noriu!
(fiction)
[You will object (that) it's not the season for the ice-cream yet, but I want it so
much!]
c. <...> buhalteriai, galbūt ims mums priekaištauti: girdi, yra premijos, kategorijos,
<...>.
(journalism)
[<...> accountants may expostulate (that) there are rewards, categories, <...>.]
In regards to English complement that-clauses, it can be observed that there are no
directly corresponding structures. Therefore, it is difficult to compare them in the two
languages. On the whole, considering the tendency of esą, neva and girdi to adopt the
function of the complementizer, it would be logical to assume that these structures should
rather be regarded as that-complement clauses with the overt form of the complementizer and
the Lithuanian instances like those in example (25) should be translated with that in the place
of esą, neva and girdi. Such a translation, however, would not convey the evidential meaning
of these lexical items, thus the translator should incorporate some compensational strategies
in order to preserve it. What is more, in some cases of the translation from English into
Lithuanian observed in the English-Lithuanian Parallel Corpus of the LLC, translators
incorporate esą and neva when translating zero-that complement clauses into Lithuanian just
as evidential markers introducing indirect speech, as can be seen from the instances in
example (28) taken from the English-Lithuanian Parallel Corpus of the LLC:
(28) a. It hurts when I hear you say I am so serious and we don't have any fun.
47
(fiction)
Man skaudu girdėti iš tavęs, neva dėl mano rimtumo mes nepatiriame džiugesnių
akimirkų.
b. In the early 1900s the Vatican went so far as to say the Illuminati were a figment
of overactive imaginations.
(fiction)
Dvidešimto amžiaus pradžioje Vatikanas nuėjo tiek toli, kad ėmė teigti, esą
iliuminatų brolija apskritai buvo tik įaudrintos vaizduotėspadarinys.
What is peculiar, the English sentences in (28) do not contain any obvious evidential elements
refering to the speaker's doubt about the truth of the information reported. Thus, the
translators' choice to incorporate these lexical elements into the Lithuanian version is
probably due to the fact that they are typical strategies of refering to somebody's words in
Lithuanian. Still, the Parallel Corpus in the LLC contains too few examples of these structures
to draw any valid conclusions; the Lithuanian-English Parallel corpus of the LLC is even
smaller and does not contain the instances of neva, esą or girdi at all. In the future, when the
size of the parallel corpus and the variety of texts presented there grows, a corpus-based study
of Lithuanian structures containing these peculiar lexical elements and corresponding English
translations could shed more light upon how these structures are realised in English and how
influential they are in respect to the choice of zero that .
Another structural peculiarity of Lithuanian complement clauses after the verbs of
saying which seems to favour the deletion of that is the complexity of the complement clause
itself. The complementizer that is often absent when the complement clause is a complex
sentence of condition joined to the matrix clause by the conjunction jei/jeigu “if”, as
illustrated in example (29):
(29) a. Jis užsiminė, jei situacija Lietuvos rinkoje blogės, DINAMIKA trauks
koncertoti už Atlanto. (journalism)
[He mentioned (that) if the situation on the Lithuanian market deteriorates,
DINAMIKA will head to concert across the Atlantic.]
b. Seimo nariai-agrarininkai pagrasino, jei nebus žemės ūkiui skirta 8-10 proc. lėšų,
balsuos prieš.(journalism)
[The Agrarian party members of Seimas threatened (that) if agriculture does not get
8-10 percent of the funds, (they) would vote against.]
c. Vaivada pats pažadėjo - jeigu gerai pasirodysite, važiuosite į Briuselį. (fiction)
[Vaivada promised himself – if (you) perform well, (you)'ll go to Brussels.]
48
d. Dar pagrasino: jei jaunuolis neatveš dolerių, kreipsis į policiją. (journalism)
[(He) also threatened: if the youngster doesn't bring some dollars, (he) will go to the
police.
It must be mentioned that such structures typically occurr in the journalism section of the
LLC, and the most common punctuation mark in them is the colon, as in (27d). In the BNC
the corresponding complement clauses are typically used with the overt form of that, as
illustrated in (30):
(30) a. He threatened that if the Moldavian authorities did not notify him within 10
days of their implementation of measures to extricate Moldavia from its crisis of
"inter-ethnic confrontation" he would take "necessary measures" in accordance
with his newly enhanced presidential powers.
(non-academic)
b. Mr Kinnock promised that if Labour come to power, all that will change.
(spoken)
Another exclusive feature of the zero-that complement clauses after the verbs of
saying in Lithuanian suggested by the corpus evidence is that the colon as a punctuation mark
for the separation of the main clause from the complement clause is by far more typical than
the comma in written language, which accords with the numerous examples cited in
Lithuanian works on grammar. The examples of such structures were detected in LLC for 46
of the 75 studied matrix verbs:
(iš/pa)aiškinti “explain”, akcentuoti “emphasize”, byloti “proclaim”, pagrasinti
“threaten”, informuoti “inform”, pasakoti “tell”, įtikinėti “urge”, jokauti “joke”, liudyti
“certify”, murmėti “grouse”, pabrėžti “emphasize”, patikslinti “specify”, išpranašauti
“foretell”, pranešti “report”, pridėti (kalbant) “remark”, prieštarauti “remonstrate”, priminti
“remind”, pripažinti “acknowledge”, prisipažinti “profess”, prisiekti “swear”, prognozuoti
“forecast”, pareikšti “declare”, rekomenduoti “recommend”, sakyti “say”, paskelbti “declare”,
skųstis “complain”, susitarti “agree”, šaipytis “sneer”, šmaikštauti “quip”, prasitarti “blurt”,
teigti “state”, teisintis “plead”, patvirtinti “affirm”, užsiminti “hint”, pažadėti “promise”,
pažymėti “stress”, aimanuoti “complain”, kalbėti “speak”, pasakoti “tell”, pridurti
“add”,prikišti (daryti priekaištą) “expostulate”, priekaištauti “expostulate”, rėkti (skųstis)
“complain”, šnekėti “speak”, tikinti “assure”.
49
Some typical instances are illustrated in example (31), occasional instances with the
complement clauses preceded by esą, girdi and jei(gu) were illustrated in examples (27) and
(29d).
(31) a. Kauno darbininkų sąjungos pirmininkui G. Čižikui praneš: įstatai atitinka
Lietuvos Respublikos profesinių sąjungų reikalavimus, <...>.
(journalism)
[The chairman of the Kaunas labourist union G. Čižikas will be informed (that) the
statute meets the requirements of the Trade Union of the Lithuanian Republic,
<...>.]
b. Viena draugė mane įspėjo: 'Obsession' per saldūs.
(ficiton)
[A friend of mine warned (that) 'Obsession' was too sweet.]
c. Visą laiką akcentuodavo: reikia pakviesti tarptautinę komisiją.
(spoken)
[(He) would emphasize (that) it is necessary to invite an international committee.]
It must be noted that the Lithuanian asyndetic sentences with the complement clause
separated by the colon are the most frequent type of zero-that clauses in both the fiction and
journalism sections; however occasional instances can be found in the spoken section as well,
as illustrated in (30c). On the contrary, the colon is absolutely not typical as a punctuation
mark in zero-that complement clauses with the matrix verbs of saying in English. The BNC
evidence shows that the colon is occasionally used after the verbs of saying only in case the
direct speech indicated by quotation marks follows the reporting clause35, as illustrated in
example (32):
(32) a. The Darlington and Stockton Times reported: 'There are so many public
houses in the neighbourhood that the Palace will afford a very wholesome counterattraction to the temptations which are offered by these houses.'
(newspaper)
b. Scholar alleged: 'Tottenham, alas, never seemed to get his full attention.'
(newspaper)
Another significant observation about the Lithuanian verbs of saying taking zero-that
complements suggested by the corpus evidence is that these structures to a great extent merge
with another syntactic category: extended sentences with parentheticals. Parentheticals or, as
35
Interestingly, English grammars usually mention only the comma as the punctuation mark sperating the
reporting clause from the direct speech (LGSWE, 1999: 196; CGEL, 1995: 1113). However, the newspaper
register of the BNC contains instances with the colon used for this purpose.
50
they are labelled in LG (2006), interpolations36 are defined as “syntactic means of amplifying
a sentence by a broad <...> range of meanings” (LG, 2006: 685). Interpolated remarks are
added by the speaker to the proposition he/she makes in order to provide a remark,
clarification, to refer to his/her or someone else's opinion, as well as to express some modality
or emotional evaluation about the message (Sirtautas and Grenda, 1988: 137; Labutis, 2002:
343; DLKG, 2005: 644; LG, 2006: 688). The main feature of the interpolated remark is that it
is considered to be peripheral, providing additional meaning to the proposition, i.e. it can be
omitted without distorting the main information conveyed by the sentence (Labutis, 2002:
344; Sirtautas and Grenda, 1988: 136).
Interpolated remarks, which are diverse in their origin, form, and structure, include a
group of parenthetical phrases of clausal structure, with the predicate expressed by a finite
form of a verb, which can also be used alone as an interpolation (LG, 2006: 688). Such forms
of the verb sakyti “say”, which is one of the most frequent verbs taking zero-that
complementsin both English and Lithuanian, as sako “(he/she/they) say”, sakai “(you) say”,
sakykime “let's say”, sakysime “(we) will say” are reported functioning as interpolations
(Masilionis, 1985: 66; Sirtautas and Grenda, 1988: 133; Labutis, 2002: 349; DLKG, 2005:
643)37. Interpolations do not have a strictly fixed position in the sentence, and can typically
occur in the initial (or head), medial or final position (DLKG, 2005: 644), as illustrated in
examples from LLC (33a)-(33c) respectively:
(33) a. Sako, siunčia man pinigų, kokį tūkstantuką, <...>
(fiction)
[(She) says (she)'s sending me some money, about a grand, <...>]
b. Net pinigų nespėjo pasiimti. Dar ir dabar, sako, tebėra kalneliuose užkasti. (fiction)
[She even didn't have time to take the money. Even now, (she) says, it's still burried in
the hills.]
c. Tik merginom, sako.
(spoken)
[Only for girls, (they) say.]
As structurally interpolations of this type correspond to simple sentences (DLKG, 2005: 643)
and are separated from the rest of the sentence by the comma (Masilionis, 1985: 64; Sirtautas
36
37
The terms interpolation and parenthetical used in this work are synonymous, the same is valid for the terms
interpolated remark and comment clause.
Besides sakyti (say), the corpus evidence from LLC showed that certain forms of the following verbs of
saying can function as interpolations: (iš/pa)aiškinti “explain”, pagrasinti “threaten”, įspėti “warn”, pasakoti
“tell”, juokauti “joke”, lažintis “wager”, minėti “mention”, pabrėžti“emphasize” , priminti “remind”,
pripažinti “admit”, prisipažinti “confess”, prisiekti “swear”, pareikšti “decree”, pažadėti “promise”, psakoti
“tell, say”, šnekėti “tell, say”.
51
and Grenda, 1988: 136), when used in the head position, they can be confused with the matrix
clauses followed by zero-that complements: on the one hand, such verbs according to their
lexico-grammatical pattern require a complement (which from the syntactic perspective is
realised by the rest of the sentence in such cases), and, on the other hand, there are no strict
and explicit requirements for the structure of a parenthetical clause of such type. Only the
semantic criterion of the phrase being peripheral to the meaning of the sentence rather than
conveying its main idea and the clished nature of the phrase (the verb form, e.g. sakykime
“let's say”, or a set phrases, e.g. žmonės sako “people say”) help to identify such structures as
parentheticals. Thus, for instance, the verb form and the function of what syntactically can be
regarded as the matrix clause in examples (34a)-(34c) suggests that they should rather be
interpreted as interpolations, while the examples in (34d)-(34e), where the matrix clause is
manifested by a longer phrase than just the verb form, represent confusing cases:
(34) a. Sakai, šitas pameistrys turi šeimą, vaikų, <...>.
(fiction)
[(You) say, this apprentice has got family and kids, <...>.]
b. Sako, pašte jau yra laiškas.
(journalism)
[(They) say, at the post office there's already a letter.]
c. Pripažinsiu, kartais būdavo pagrindo nepasitikėti, <...>.
(journalism)
[(I)'ll admit, sometimes there was a ground for distrust, <...>.]
d. Aš... prisipažinsiu, aš niekada apie tai rimtai negalvojau...
(fiction)
[I … will admit I have never thought about it seriously...]
e. Aš tau sakau, stalas šoko!
(spoken)
[I'm telling youthe table was dancing!]
However, even the cases illustrated in (34a)-(34c) seem to lack an explicit criterion to be
clearly classified as interpolations, especially taking into account that the corpus evidence
provides examples where identical or very similar verb forms are followed either by zero-that
complements separated by the colon or the comma, or by that-complements with the overt
complementizer, as illustrated in example (35) respectively:
(35) a. <...> nors sąžiningai prisipažinsiu: visą likusią naktį nesudėjau bluosto.
(fiction)
[<...> though (I)'ll sincerely admit(:) I couldn't sleep a while through the rest of
the night.]
b. Prisipažinsiu, jaučiausi nei šiaip, nei taip.
52
(fiction)
[(I)'ll admit (,) I felt awkward.]
c. Prisipažinsiu, kad vien nuo šitos minties jaučiu fizinį diskomfortą.
(fiction)
[(I)'ll admit that the mere thought of it made me feel physical discomfort.]
Akelaitis (2002), who investigates parentheticals structurally resembling the main clause and
distingushes among such parentheticals groups of the verbs of saying as well as thinkig
(2002: 4), aims at establishing structural criteria for categorizing such finite verb forms as
interpolations. In his opinion, when functioning as interpolations, such phrases are not
structurally expanded: the verbs don't have subject and object arguments, which are
obligatory for these verbs due to their lexico-grammatical pattern, the subject in the first and
second person form is omitted, the subject in the third person form cannot have definite
reference (2002: 5-6). The scholar concludes that such verb forms used parenthetically lose
their syntactic connection to the rest of the sentence, which functions, therefore, as the main
sentence and cannot be considered the complement of such a verb. The lexical meaning of the
verb bleaches as well (op. cit.: 7). According to these criteria, then, the sentences in examples
(34d) and (34e) should be considered main clauses with zero-that complements, or as
Akelaitis labells them free constructions (laisvosios konstrukcijos) (2002: 6). However, this
classification does not account for the semantic difference between the sentences in (35).
A possible explanation for such confusing cases is suggested in some recent
researches. Ruskan (2010) observes that parentheticals of this type can also be ascribed to
means of expressing evidentiality, which means that these verbs are not used in their direct
meanings but in their evidential meanings (2010: 2, 6), which generally correlates with the
idea that the primary lexical meaning of the verbs used parenthetically bleaches. She suggests
that such verbs are undergoing the process of grammaticalization and gradually become
particles denoting the source of the information (op. cit.: 7). Such point of view matches the
intuitions about the English verbs of saying expressed in the works by Thompson and Mulac
(1991) and Kaltenböck (2004), which are reviewed in Chapter 3.1 of the current paper. These
scholars also consider that the matrix clause in such cases is subject to the process of
grammaticalization (Thompson and Mulac, 1991: 237), with the main clause loosing its status
and becoming peripheral in comparison to the complement clause, which conveys the new
information in the sentence, while the matrix clause just provides the reporting frame to it and
can be considered parenthetical (Kaltenböck, 2004: 65). The omission of the complementizer
that in such structures is natural as zero connection exhibits no overt mark of subordination
(Kaltenböck, 2004: 65). The case of Lithuanian parenthetical verbs shows that this process
has been already legitimized as grammaticalization and such sentences are classified as
53
simple extended instead of complex.
Nevertheless, the explanation above does not account for the fact that the same verb
forms are used in different structures, as it was illustrated with the Lithuanian verb
prisipažinsiu “(I)’ll admit” in the example (35) and for the difference between them, as
semantically they seem to be very close. A different interpretation presented in Kearns
(2007)38, who explains the peripheral meaning of the matrix clause not by grammaticalization
but by the use of the matrix verb in its epistemic meaning, provides more isights. Considering
the verbs of saying, especially say, Kearns also agrees that the matrix clause in the sentences
with that-complement clauses is often “of low informational prominence <...> in comparison
with the greater prominence of the embedded clause as the new content” (2007: 502), but she
argues that the complement clause as the main assertion can be also observed in the sentences
with the overt form of the complementizer that. According to her, both cases are best
explained by the meaning of the matrix verb: in such cases the verbs are used not in their
main meaning, but in the evidential meaning or modifier sense (op. cit.: 501). This
explanation would also account for the semantic closeness of Lithuanian structures
examplified in (35).
Considering the English zero-that complement clauses corresponding to the confusing
cases of Lithuanian verbal interpolations as opposed to zero-that complements, it should be
mentioned that the same verbs functioning as parentehticals in Lithuanian are involved in
problematic cases in English, and recently this issue has been widely discussed in numerous
English corpus-based studies. Matrix clauses which are peripheral in meaning to the main
information of the sentence conveyed by the complement clause, are most often referred to as
comment clauses (GCE, 1972: 778; CGEL, 1995: 1112; LGSWE, 1999: 197) and are reported
to function as pragmatic markers (Brinton, 2008: 2). Similarly to Lithuanian interpolation
remarks, the comment clauses are described as being loosely related to the rest of the sentence
and occuring in the initial, medial or final position (GCE, 1972: 778; LGSWE, 1999: 197).
Comment clauses comprise a group which structurally resembles the main clause (GCE,
1972: 778; CGEL, 1995: 1112; LGSWE, 1999: 197), like Lithuanian verbal interpolations, as
illustrated in the following examples from the BNC39:
38
39
(36) a. There was one reason, I admit.
(newspaper)
b. The council, you say, has done everything it can.
(spoken)
The study by Kearns (2007) is reviewed in Chapter 3.1 of this paper.
Besides say, the corpus evidence from BNC showed that the following verbs of saying can be used in
comment clauses: admit, claim, confess, emphisize, explaine, insist, mention, promise, report, swear, wager.
54
c. I've got to admit, I want to be able to do that too!
(magazine)
The major semantic functions of English comment clauses are reported as to “hedge”
or “express the speaker's tentativeness over the truth of the matrix clause”, to express the
speaker's certainty, emotional attitude towards the content communicated or to claim the
hearer's attention (CGEL, 1995: 1114; LGSWE, 1999: 197), which also corresponds to the
functions ascribed to Lithuanian verbal interpolations (Sirtautas and Grenda, 1988: 137;
Labutis, 2002: 343; DLKG, 2005: 644; LG, 2006: 688). Džežulskienė (2011), who compares
English and Lithuanian comment clauses on the basis of the Parallel Corpus of the LLC
evidence, mentions three main pragmatic functions of comment clauses, namely, phatic,
reporting, and mitigating, and additional illocutionary (ilokucinė), self-correction
(pasitaisymo), and tentativeness (dvejonės) functions (2011: 16). She claims that Lithuanian
parantheticals with the verb sakyti “say” used in the initial position in the sentence most often
perform the reporting function (as in examples (33) and (34b)), the phatic40 function (as in
example (34a)), and illocutionary41function (as in examples (24e)-(24f), and (34c)) (op. cit.:
16).
English comment clauses resemble Lithuanian interpolations in a few more ways: in
both languages they have a tendency to be used as cliched, “stereotyped” (CGEL, 1995: 1114)
or “formulaic” (LGSWE, 1999: 197) phrases, with the first, second or third-person pronoun
combined with the present-tense verb (Brinton, 2008: 2) and in both Lithuanian and English
they are separated by the comma (CGEL, 1995: 1113), which is different from the cases when
main clauses are followed by complements in English. This formal criterion, at the first sight,
helps to distinguish Englis comment clauses from the matrix clauses more easily than it can
be done in Lithuanian, where both parentheticals and subordinate clauses are separated by the
comma from the rest of the sentence. However, such a criterion can help only in the written
mode. Besides, the following examples from the BNC suggest that semantically the cases of
comment clauses vs. matrix clauses with zero-that complements can be rather confusing,
which is also the case in Lithuanian:
(37) a. I admit, the figure's gone a bit, but then it started going when I was still
twenty-five.
40
41
(fiction)
Phatic function is defined as the intention to maintain the contact between the speaker and the addressee
and claim the hearer's attention (Džežulskienė, 2011: 17).
Illocutionary function is defined as the intention of attracting the addressee's attention to the content of the
message conveyed by the speaker or the speaker's attitude towards this content (Džežulskienė, 2011: 17).
55
b. I admit I'm not most passionate person in the world <...>.
(spoken)
c. And you say, you know.
(spoken)
d. So you say they're fully competent.
(spoken)
It is generally recognised that in the initial position it is difficult to distinguish between
comment clauses and matrix clauses in written English (Storms, 1966: 269; CGEL, 1995:
1113; CGEL, Biber et al, 1998: 72; CAGEL, 2002: 896).
In fact, in scholarly discussions on the issue of the English verbs of thinking42 and
saying taking zero-that complements and the ambiguity between them and comment clauses
in the initial position in the sentence there are two main perspectives supported by different
researchers. The common element unifying both perspectives is related to the meaning of the
verb itself: if the verb is used in its primary lexical meaning, the clause functions as the matrix
clause with the zero-that complement. However, in comment clauses the verbs are used in
their modyfing and thus parenthetical sense; therefore semantically the matrix clause loses its
prominence and functions as a modifier to the complement clause, which, in turn, becomes
semantially central in respect to the information conveyed by the sentence (Thompson and
Mulac, 1991: 249; Kaltenböck, 2004: 65; Kearns, 2007: 503). Zero that is a more natural
choice in such a case, as it shows no formal sign of the complement clauses being
subordinated to the matrix clause (Thompson and Mulac, 1991: 249; Kaltenböck, 2004: 65;
Kearns, 2007: 503; Bogaert, 2011: 324). However, the modifying, parenthetical meaning of
the verb is crucial even in cases with the overt form of the complementizer that and in such
cases the matrix clause is still semantically peripheral (Kearns, 2007: 501; Bogaert, 2011:
322; Kaltenböck, 2009: 66). The difference between the two described perspectives lies in the
explanation of the mechanism underlying the semantic shift in the roles of the matrix and
complement clauses. Some scholars ascribe this shift to the process of grammaticalizatison
and claim that these verbs undergo semantic bleaching and matrix clauses containing them
change their category and become either epistemic modifiers or pragmatic and discourse
markers (Thompson and Mulac, 1991: 249;
Brinton, 2008: 2; Kaltenböck, 2008: 124;
Kaltenböck, 2009: 67; Kaltenböck, 2009a: 72), while others insist that the evidence for the
grammaticalisation of such structures is insufficient and these clauses should be still
syntactically classified as matrix clauses but with the parenthetic, modifying meaning of the
42
The problem of ambiguity between comment clauses and matrix clauses is even more relevant to the verbs of
thinking than saying; however, for the sake of convenience, the major observations on this issue are presented
in this part of the paper, while Chapter 2.2 provides only additional observations characteristic of the verbs
of thinking.
56
matrix verb, which is not a grammatical but rather pragmatical change (Kearns, 2007: 503;
Bogaert, 2011: 32). Still, this disagreement rather deals with the formal classification than
with the phenomenon itself, and Kaltenböck (2009) rightly observes that the choice “depends
on one's theoretical framework” (Kaltenböck, 2009: 67).
In Lithuanian, according to the recent grammatical trends43, cliched matrix verb forms
(e.g. sako “(he/she/they) say”, manyčiau “(I')'d think”, atrodo “(it) seems”) and cliched
phrases (e.g. žmonės sako “people say”) are ascribed to parentheticals (Labutis, 2002: 349),
which means they are considered to have undergone the process of grammaticalization, while
expanded structures including the first and second person pronominal subjects are considered
free constructions (Akelaitis, 2002: 6; Džežulskienė, 2011: 18) and rather conform to English
matrix clauses with epistemic verbs used in modifying meaning and followed by thatcomplements. The comparison of these structures in the two languages illustrates the formal
and semantic changes which are in progress in both languages. As it was already mentioned,
the processes of syntactic reanalysis and grammaticallization are very slow and gradual in the
language and this is the main reason for the co-existence of different forms and their semantic
ambiguity (Ambrazas, 2006a: 25).
On the whole, comparing English and Lithuanian zero-that complements after the
verbs of saying the following features can be observed. In both languages the main
information of the sentence is often conveyed by the part which formally can be classified as
the complement clause. Consider the following examples (38) and (39) comprised of the
instances from the LLC and BNC cited in the examples above and repeated here:
(38) a. Taigi, jie sako, nereikia jo spausti prie sienos, <...>.
(journalism)
[So, they say we shouldn’t push him to the wall, <...>.]
b. Ir jinai sakė, mes pabūsim pas ją, išgersim šampano, <...>.
(spoken)
[And she said we'd stay at hers, have some champagne, <...>.]
c. Aš tau byloju, kai kurios iš jų reformų yra tiesiog pasišlykštėtinos <...>. (ficiton)
[I'm telling you some of their reforms are just revolting <...>.]
d. Sakai, šitas pameistrys turi šeimą, vaikų, <...>.
(fiction)
[(You) say, this apprentice has got family and kids, <...>.]
43
In earlier works on grammar even cases with the verbal predicates used in present tense with the subjects I or
you omitted were cited as examples of conjunctionless complement clauses, for instance:
Žinau, tu atrasi namo kelią. [(I) know you'll find the way home.] / Manai, mūsų liaudyje nėra gyvybės?
[(Do you) think our nation doesn't have ennergy?] (Dambriūnas, 1963: 69)
57
(39) a. And they said it was our car.
(spoken)
b. The hotelier said it was Italy.
(fiction)
c. And I said I'm not talking about the council <…>.
(spoken)
d. You said you were thrown out by your landlord.
(fiction)
The matrix clause often refers to the source of information, as in (38a)-(38b) and (39a)-(39c)
or serves some modifying function, such as claiming the hearer's attention as in (38d) and
(39d) or expressing the speaker's certainty as in (38c).
In addition, both languages under consideration exhibit a considerable amount of
verbs of saying (in most cases these are the same verbs in both languages) followed by zerothat complements which can be confused with interpolations or comment clauses in initial
position, as it was examplified above. The context in the written mode and the intonation44 in
the spoken can help to interpret their syntactic function, while semantically the same
phenomenon of the verbs losing their primary lexical meaning and being used in evidential,
modifying meaning can be observed in both languages.
When used in primary meanings, English verbs of saying are more numerous with
zero-that complements in all registers and exhibit greater structural variety of both the main
and the complement clauses, as can be observed from the following instances:
(40) a. At least one member of the jury said he would not have given a sentence of
death.
(magazine)
b. She was waved on by a sharp-eyed young officer, who boasted he could smell a
smuggler from fifty yards away.
(fiction)
c. Minister Ronchey decrees they are to stay open ten hours a day.
(magazine)
d. Gascoigne went up to each player <...> and demanded he die for the club and the
cause.
(newspaper)
e. Suppose he needed you there to assert he'd walked into a trap?
(fiction)
f. They had one likeness of Oreste, sent at Christmas when Wilson had requested it
should be taken.
(fiction)
In Lithuanian, when used in primary lexical meaning the verbs of saying followed by zero44
In spoken English comment clauses are reported to be prosodically marked (Storms, 1966: 269;
CGEL, 1995: 1113) by a “higher or a more emotive tone“ (Storms, 1966: 269). Lithuanian parentheticals are
also reported to be marked by peculiar intonation in the sentence (DLKG, 2005: 641).
58
that complements are more typical in asyndetic structures with the colon, as it was illustrated
in examples (27), (29d) and (31).
In respect to structural peculiarities of the complement clause, it can be added that
English verbs of saying can be followed by a zero-that complement clause in the subjunctive
mood, as in (40d) and 40f), while Lithuanian verbs taking that-complement clauses of
motivation expressed in subjunctive mood, such as liepti “order” or pareikalauti “demand”,
occur only with the overt form of the complementizer kad. However, Lithuanian complement
clauses favour zero-that when preceded by particles esą and neva, a parenthetical girdi, or a
conjunction jei(gu), which is not typical of English that-complement clauses. In addition,
most of Lithuanian verbs of saying followed by zero-that complements occur in asyndetic
structures with clauses separated by the colon. These structures vary more in respect to the
tense forms of the matrix clause as well as to the subjects of both clauses than the structures
incorporatin the comma as a punctuation mark. Furthermore, Lithuanian asyndetic sentences
with the colon deal with the verbs used in their primary lexical meanings more often than
asyndetic sentences with the comma, as illustrated in (27), (29d) and (31). On the contrary,
separation by the colon is not typical of complement zero-that clauses with the verbs of
saying in English.
4.2 The deletion of the complementizer that in complement clauses after the verbs of
thinking in English and Lithuanian
The verbs of thinking under consideration “are concerned with thinking” (CGP, 1998:
99). The corpus evidence collected reveals that to a considerable extent the observations
which can be made about the zero-that complements after the verbs of thinking in both
languages resemble the observations made about the verbs of saying. Verbs of thinking are
obviously more frequent with the zero form of the complementizer in English than in
Lithuanian. From the 52 English verbs of thinking studied, for the following 27 verbs
examples of zero-that complements were detected in all registers investigated:
accept, acknowledge, agree, anticipate, assume, believe, bet, consider, decide, doubt,
dream, estimate, expect, fancy, forget, guess, hope, imagine, intend, know, presume, reckon,
regret, remember, suppose, suspect, think.
59
Most of these verbs occur in various tenses45 with both nominal and pronominal subjects, as
illustrated in the following examples:
(41) a. So I acceptthere are are a number of items here which I discussed with the
chairman <...>.
(spoken)
b. <...> she had to acknowledge she'd done wrong
(spoken)
c. Ruth knows and accepts she is important to us.
(fiction)
d. Mr Field has acknowledged he might lose,<...>.
(newspaper)
e. I dreamed I wished you were dead.
(fiction)
f. Carrie agreed it was an item that had completely slipped her memory. (fiction)
g. When Tim helps clients to decorate they have to accept they will not end up with a
'decorated' house in the London sense.
(magazine)
h. I believe you wanted to see me.
(fiction)
The closeness of the clause juncture reported by Elsness (1982) as influential for zero-that
complements is also clearly observable with the matrix verbs of thinking. Moreover, the
corpus evidence also suggests that coreferential subjects, as in (40b)-(40e), as well as the first
person singular pronoun I as the subject of the main clause are more frequent with the verbs
of thinking than with the verbs of saying. In addition, when the subject of the main clause is a
noun it is most often a proper noun denoting the name of a person.
Another similarity between the English zero-that complement clauses after the verbs
of saying and thinking is that with rare verbs the complementizer does not tend to be omitted.
The following verbs, for the most of which the frequency of occurences in the corpus is lower
than 2,000, occured with zero-that complements only occasionally:
dread, envisage, hallucinate, marvel, prefer, resolve, speculate, surmise, wonder.
A number of infrequent verbs, for which the frequency of occurence in the corpus is lower
than 1,000, have no instances of zero-that complements in the BNC:
concur, conjecture, disagree, disbelieve, envision, fantasize, fret, hypothesize, muse,
rejoice, theorize.
45
Progressive tenses do not occur with most of the verbs of thinking, which is typical of the verbs expressing
mental sates.
60
Some verbs, such as dread, foresee, hallucinate, prefer, or resolve, were detected with
zero-that complements in the fiction register only, as illustrated in example (42):
(42) a. Unless, of course, you're afraid to venture where you dread you'll prove less
than excellent?
b. She foresaw he would probably be difficult about it, <...>.
c. On the screen a taker of LSD, the Vision-of-Hell-drug, was borne off on a
stretcher, after having hallucinated he could fly.
d. Val would presumably preferyou didn't get arrested.
e. Theodora resolved she would do her best.
Considering the zero-that complement cases found in the LLC, it can be said the
structures marked by the comma, which resemble English zero-that complements most, are
even more infrequent with Lithuanian verbs of thinking than with the verbs of saying. The
most infrequent Lithuanian verbs listed below, with the occurences across the corpus lower
than 1,000, do not occur in the LLC with the complementizer omitted, which is similar to the
verbs of saying:
fantazuoti “phantasize”, nerimastauti “worry”, nervintis “fret”, nuogastauti “fear”,
pagrįsti “reason”, perprasti “grasp”, piktdžiūgauti “gloat”, susierzinti “fret”, prasimanyti
“phantasize”, nutuokti “grasp”.
Some more frequent verbs, namely atmesti “reject”, gailėtis “regret”, įvertinti “estimate”,
laukti “hope”, and nerimti “fret”, were not found with zero-that complements either. A
number of verbs were only occasionally detected with zero-that complements:
atsiminti “remember”, džiaugtis “rejoice”, samprotauti “reason”, sapnuoti “dream”,
spėlioti “conjecture” , spėti “hypothesize”,sutikti “accept”, svajoti “dream”, tikėti “believe”,
viltis “hope”, tikėtis “hope”.
Most of the verbs on the list above did not occur with zero that in the spoken register;
however, this might be as well due to the reason mentioned in Chapter 2.1, namely, the size of
the spoken section of the LLC, as many verbs of thinking occur in this register only several
times or even do not occur at all.
Only 2 out of 51 Lithuanian verbs were detected with zero-that complements in all
61
registers under consideration and exhibited some structural variety: manyti and žinoti, which
correspond to the English verbs think and know, and these are reported among the most
frequent English matrix verbs used with zero-that complements. Some typical instances are
illustreted in example (43):
(43) a. - Paparčio žiedą. - Tu mania nerastumėt?
(spoken)
[- Fern blossom. - You think (you) wouldn't find it?]
b. Aš manau, dabar į bažnyčią vaikšto nebedaug žmonių.
(fiction)
[I think nowadays not as many people go to church.]
c. Aš manau, kapinėms ten ideali vieta.
(journalism)
[I think that is the ideal place for the cemetary.]
d. Žinau, jūs aktorė.
(fiction)
[(I) know you are an actress.]
e. <...>, žinau, jie nemyli manęs.
(journalism)
[(I) know they don't love me.]
f. Taip, jie beveik nieko nežinojo, tiesiog žinojo, štai bus kažkokia lyriška daina.
(spoken)
[Yes, they knew nearly nothing, they just knew it was going to be some lyrical
song.]
However, even these verbs did not occur with zero-that complements frequently, and
structurally they are more limited than the instances of zero-that with the verb sakyti “say“:
the subject of the matrix clause is most often I and the verbs are usually in the present tense.
The following verbs were found with zero-that complements in occasional cases:
atsiminti “remember”, galvoti “think”, prisiminti “remember”, samprotauti “reason”,
sapnuoti “dream”, spėlioti “conjecture”, spėti “conjecture”, sutikti “accept”, svajoti “dream”,
tikėti “believe”, viltis “hope”.
Most often these examples belong to the fiction register. A few cases are illustrated in
example (44):
(44) a. Simonas atsiminė, jis turėjo brolį, Endriejų Grygą,<...>.
(fiction)
[Simonas remembered he had a brother, Endrey Grygas, <...>.]
b. <...> tai galvoju, reikia ir man ten kaip nors su juo susirišti.
62
(spoken)
[So (I) think I should also somehow get in touch with him.]
c. Aš jau žiemą galvojau, gal galą pasidarysiu.
(fiction)
[ In the winter I thought (I) might kill myself.]
Another similarity between the zero-that complements after the Lithuanian verbs of
thinking and saying is the presence of a particle, or a conjunction jei “if“ at the beginning of
the complement clause. However, the particles esą and neva are only occasional with the
verbs of thinking, which must be due to the fact that semantically these lexical elements are
related to speech rather than thought. Moreover, with the verbs of thinking they have different
distribution across registers: they are more frequent in the fiction than in the journalism
section. On the other hand, some verbs of thinking occur with zero-that complements when
the complement clauses are preceded by a particle tarsi “like“, a modal word gal “maybe“,
and an adverb lyg “as if“, which are also ascribed to reportive or hearsay evidential markers
(Wiemer, 2007: 178). Some typical cases of the described above structures are illustrated in
example (45):
(45) a. Vilimės, esą mus turi suprasti savaime.
(journalism)
[(We) hope (that?) we must be understood as a matter of fact.]
b. Sapnuoju, neva vilkelis laisvai šokčioja ant savo smaigalio.
(fiction)
[(I)'m dreaming (that?) a humming-top is jumping loosely on its spire.]
c. Daugelis žmonių klaidingai įsivaizduoja, jei jau esi politiko žmona, tai vakarais su
vyru aptarinėji politinę situaciją.
(journalism)
[Many people wrongly assume (that) if you are a polititian's wife, in the evenings
you necessarily discuss political issues with your husband.]
d. Vis galvojau, gal kas žino, kur mano sesuo.
(fiction)
[(I) still was thinking maybe someone knew where my sister was.]
e. Sapnavau, tarsi man kas būtų davęs vardą.
(fiction)
[I dreamed (that?) someone had given me a name.]
These examples show the similar tendency to the one observed with the verbs of saying: in
the Lithuanian language particles, adverbs and modal lexical items tend to acquire the
function of the complementizer (Būda, 1986: 54, 56), on the one hand, and, on the other hand,
in case there is another conjunction, i.e. jei(gu) “if“, at the beginning of the complement
clause, the complementizer tends to be omitted. In comparison, the English complement
clauses do not have such numerous alternatives to the complementizer that, as it was already
63
discussed in Chapter 2.1.However, zero-that clauses can occur with adverbs expressing
modality like maybe or perhaps, similar in meaning to Lithuanian gal, as illustrated in
example (46):
(46) a. I thought maybe you'd got it while you were abroad.
(spoken)
b. That summer Lewis thought perhaps he was on the verge of a nervous
breakdown.
(fiction)
One more common feature of zero-that complement clauses after the verbs of thinking
and saying in Lithuanian is that in written language for the majority of verbs such structures
incorporate the colon as the punctuation mark separating the clauses: 30 out of the 51 verbs
of thinking under consideration were found in such asyndetic constructions in the LLC,
namely:
apskaičiuoti “calculate”, argumentuoti “argue”, atsimiti “rememer”, prisiminti
“remember”, baimintis “dread”, bijoti “fear”, džiaugtis “rejoice”, galvoti “think”, įsivaizduoti
“envision”, įtarti “suspect”, jaudintis “worry”, jausti “feel”, manyti “feel”, nerimauti “worry”,
numanyti “anticipate”, nuspręsti “decide”, pamiršti “forget”, pasiryžti “resolve”, prisiekti
“swear”, samprotauti “reason”, spėlioti “conjecture”, spėti “conjecture”, stebėtis “marvel”,
sutikti “accept”, svajoti “dream”, tikėti “believe”, tikėtis “hope”, viltis “hope”, žinoti “know”.
Some instances are illustrated in example (47):
(47) a. Tuo pat laiku atsiminė: jeigu neišmoks dalyti, nebepamatys mamos.(fiction)
[At the same time (he) remembered (:) if (he) doesn't learn to divide, (he)'ll never
see (his) mom again.]
b. Aleksandro širdis džiūgavo: atėjo laikas kai tas didelis vaikas atiduodamas.
(fiction)
[Alexander's heart rejoiced (:) the time has come when that big kid is given
away.]
c. Kembridžo studentai džiūgauja: jie vėl pirmi.
(journalism)
[Cambridge students rejoice (:) they are on top again.]
d. Kartais aš galvodavau: mergina – genijus.
(fiction)
[Sometimes I would think (:) the girl is a genious.]
e. <...> sąmoningai nusprendė: jis atsisakys kalbėti.
[<...> (he) deliberately decided (:) he would refuse to talk.]
64
(journalism)
f. Rodžiau vaiką specialistams ir pats numanau: dešimties metų jam su jo širdies yda
nesulaukti.
(journalism)
[(I)'ve showed the child to the specialists and anticipate myself (:) he won't survive
until he turns ten with his heart defect.]
g. O žmonės visaip spėlioja: gal vaistinė bus, gal parduotuvė.
(journalism)
[People conjecture different ways (:) it might be a chemist's or it might be a shop.]
h. Maniau: barsis, kad iš paskos kaip uodega velkuosi.
(journalism)
[(I) thought (:) (he) would scold me for dragging behind him like a tail.]
The examples demonstrate that these sentences can be more versatile in respect to the tense
forms and the types of subjects; even rare verbs, such as džiūgauti “rejoice”or spėlioti
“conjecture” occur in these structures in Lithuanian, which is the same as with the verbs of
saying. Comparing the two languages in this respect, contrary to the verbs of saying, English
verbs of thinking occur with complement zero-that clauses separated by the colon, especially
in the fiction register, as illustrated in example (48):
(48) a. I thought: I am being psychoanalysed.
(fiction)
b. <...> she thought:no, they never learn.
(fiction)
c. I simply thought: I want to wash my hands.
(magazine)
d. <...> she knew: they were not murderers.
(fiction)
However, on the whole these instances are different from Lithuanian asyndetic structures with
the colon. These English cases represent reported thoughts, thus the subject of the matrix
clause is usually the first or third person singular pronoun. As the corpus examples illustrate,
reported thoughts are often not marked by quotes. They are very infrequent in the BNC46 and
usually occur with the most frequent verbs, such as think and know. Thus, similarly to the
semantic domain of saying, the colon is not typical of English zero-that complements after the
verbs of thinking, while in Lithuanian most asyndetic complement structures are of this type.
It can be noted that some Lithuanian structures with the colon are similar to English in that
the complement clause resembles reported thoughts, as, for instance, in examples (46d) and
(46h), but in general Lithuanian structures under consideration are more versatile.
Another major observation about Lithuanian verbs of thinking with zero-that
complements is that ambiguous cases of verbal matrix clauses which can be classified as
46
The colon is generally considered an infrequent punctuation mark in English (CGEL, 1995: 1620).
65
parentheticals are even more distinct with Lithuanian verbs of thinking than with the verbs of
saying. Some cases are illustrated in the following example (49):
(49) a. Atsimenu, tada dar kartu pasijuokėme iš to lapelio.
(fiction)
[(I) rememberwe laughed together then at that leaflet.]
b. Bijau, tai yra klausymas Baltiesiems rūmams.
(journalism)
[(I) fear this is a question for the White House.]
c. Galvojau, reik pabandyt, <...>.
(spoken)
[(I) thought (I) should try, <...>.]
d. Galvojau, ten kitokia technika.
(journalism)
[(I) thought the equipment there is different.]
e. Įtariu, busiu nuvylęs J. Andrijauską.
(journalism)
[(I) suspect (I) have disappointed J. Adrijauskas.]
f. Įtariu, dar bus sunku man.
(spoken)
[(I) suspect (I)'ll still find it difficult.]
g. Įtariu, mes kažko nežinom.
(fiction)
[(I) suspect we don't know something.]
h. Manau, jei sustosiu, gyvenimas neteks prasmės.
(fiction)
[(I) think (that) if I stop, my life will lose its sense.]
i. Žinau, ten irgi sodai yra ar kažkas tokio.
(spoken)
[(I) know there are also gardens or something like that.]
Such ambiguous structures, which typically consist of the present verb tense with the subject I
omitted, occur in both spoken and written Lithuanian and are extremely typical with the most
frequent verbs, such as manyti “think” or žinoti “know”47, which coincides with the high rate
of zero-that complements with the equivalent English verbs think and know. Similarly to the
verbs of saying, these forms can also be found in sentences with the overt form of the
complementizer kad/that and in asyndetic structures with a colon, as illustrated in example
(50):
(50) a. Prisimenu, buvo šilta saulėta diena.
(journalism)
[(I) remember it was a warm and sunny day.]
47 Other Lithuanian verbs that occur in the LLC in the function of parentheticals are: atsiminti “remember”,
prisiminti “remember”, bijoti “fear”, galvoti “think”, įtarti “suspect”, jausti “feel”, numanyti “anticipate”,
prisiekti “swear”, spėti “guess”, sutikti “accept”, tikėti “believe”, tikėtis “hope”, viltis “hope”.
66
b. Prisimenu: dar nesutemus ėme pustyti.
(journalism)
[(I) remember (:) it started snowing heavily before dark.]
c. Prisimenu, kad apie nieką negalvojau.
(fiction)
[(I) remember that I didn‘t think about anything.]
What is more, similar in meaning but structurally expanded clauses in the initial position, as
in example (51) below, which should be considered not parentheticals but free constructions
(Akelaitis, 2002: 6; Džežulskienė, 2011: 18), make the cases even more confusing:
(51) a. Aš prisimenu, aš turėjau tokias <...> kojinytes dryžiukais. (ficiton)
[I remember I had such <...> stripy socks.]
b. Aš tik prisimenu, kad atėjau atlikti savo darbo.
(journalism)
[(I) remember that I've come to do my job.]
c. Taip, aš prisimenu: Genė man rodė Zenono nuotrauką.
(fiction)
[Yes, I remember: Gene showed me the picture of Zenonas.]
As it was mentioned in Chapter 2.1 in relation to the similar problem with the verbs of saying,
such structures are difficult to classify on the syntactic level and seem confusing on the
semantic level.
The English corpus evidence displays a similar tendency: in the initial position a
matrix clause with a verb of thinking can be followed by a complement that-clause, by a zerothat clause, or even (though very rarely) by a clause separated by the colon, as illustrated in
example (52):
(52) a. I think that it is a very useful exercise.
(spoken)
b. I think it's great.
(fiction)
c. I think: perhaps he is angry.
(fiction)
Generally, the matrix clauses with the most frequent verbs of thinking, such as think or
know 48 are recognized as easily confused with the comment clauses in the initial position
(Biber et al, 1998: 72; Bogaert, 2011: 295) and recently have been the object of numerous
studies in corpus linguistics. Similarly to the verbs of saying, this ambiguity is due to the fact
that the semantic domain of thinking comprises a considerable number of epistemic verbs or
48
Other English verbs of thinking which were found in the BNC used in parenthetical structures are: agree, bet,
expect, guess, hope, imagine, presume, recon, suppose, suspect, believe.
67
mental predicates which can be used either in their direct, non-parenthetical, or modified,
parenthetical menanings. As it was mentioned in Chapter 2.1, in written English marking
comment clauses by the comma can serve as a formal criterion. In Lithuanian, where the
comma is obligatory after both a main clause and a parenthetical, it is often difficult to
distinguish which meaning of the verb is implied, as can be seen from the example (53):
(53) a. Žinai, miške visko atsitinka.
(fiction)
[(You) know, anything can happen in the forest.]
b. Žinai, aš už durų palaukiau.
(spoken)
[(You) know, I waited behind the door.]
Yet, the same problem of ambiguity is present in spoken English. Therefore, the initial
clauses with mental verbs have been recently often investigated on the basis of corpus
evidence from spoken language with the aim to determine their grammatical status. Moreover,
spoken language on the whole is a particularly valuable source of information, as language
changes in general originate in spoken language and grammar “can only arise from patterns in
the way language is used by speakers” (Thompson and Mulac, 1991: 250). As the shades of
meaning are reflected in prosody, studying grammatical patterns in spoken language helps to
distinguish and classify ambiguous cases. For instance, studies on the prosodical peculiarities
of the initial I think clause in spoken English conducted by Kaltenböck (2008, 2009, 2009a)
reveal four different prosodic patterns and show that these patterns do not depend on the
retention or omission of the complementizer that (Kaltenböck, 2009: 66), and that they are
rather due to different functions the phrase performs in discourse (Kaltenböck, 2009: 66).
Interestingly, the results of the research prove that even the initial I think is much more often
used in English in a modifying sense and serves as a “qualifier of the proposition in the
following clause” (Kaltenböck, 2009: 67).
An important observation is that the prosody reflects whether the verb is used in its
main lexical meaning or modifying sense. As parentheticals in Lithuanian are as well reported
to be marked by a peculiar intonation (DLKG, 2005: 641), corpus-based studies of spoken
Lithuanian in the future might reveal more about structural and semantic ambiguity of
Lithuanian parentheticals and free constructions with the verbs of saying and thinking in the
initial position as well as help to distinguish and classify them.
On the whole, in both Lithuanian and English, a considerable number of cases of the
verbs of saying and thinking deal with the verbs being used in their modifying sense. When
used in the main lexical, i.e. non-parenthetical, meaning, the English verbs take zero-that
68
complements much more frequently, while their Lithanian counterparts are less frequent and
the majority of them belong to asyndetic sentences marked by the colon.
On the other hand, closeness of the clause juncture is a distinct feature with both
Lithuanian and English zero-that complements with the verbs of thinking as well as with the
verbs of saying. Also, in Lithuanian zero that is likely after the verbs of thinking if the
complement clause is preceded by lexical elements which tend to acquire the function of the
complementizer, which is not the case in English. However, in case a modal adverb precedes
the complement clause, zero that is possible in both languages.
4.3 The deletion of the complementizer that in complement clauses after the verbs of
discoveringin English and Lithuanian
The verbs of discovering under consideration “are concerned with coming to know or
think something” (CGP, 1998: 100). The semantic domain of discovering is represented by
the smallest group of verbs in both English and Lithuanian. On the whole, the verbs of
discovering semantically resemble the verbs of thinking, with the main difference that the
verbs of thinking rather refer to a mental state, while the verbs of discovering usually deal
with changes in mental states or with mental actions. The corpus evidence collected reveals
that the use of zero-that complements with these verbs shares some features which were
described for the semantic domains of saying and thinking. On the other hand, there are some
peculiar features observable for the verbs of this particular domain.
From the 22 English verbs of discovering investigated, 9 were detected with zero-that
complements in all registers: calculate, conclude, determine, discover, notice, realise, recall,
recognize and see. The instances in example (54) can serve as an illustration:
(54) a. They quite sensibly concluded this wouldn't be a good idea.
(spoken)
b. They may conclude the symptoms are all in the mind.
(magazine)
c. I have concluded the same should apply to the match referee's report.
(newspaper)
d. He concluded he must indeed be on his way out.
(fiction)
e. Funabem, the federal child welfare agency, calculates there are seven million street
children across Brazil.
(newspaper)
These verbs are used with different types of subjects, such as nouns or noun phrases as well as
69
singular and plural pronouns and the the corpus evidence does not suggest that some type of
subject is more typical, as it was the case with the verbs of thinking, which very frequently
occurred with pronominal subjects I or you, except for the subject you, which is rather
infrequent with the verbs of discovering followed by zero-that complements.
Interestingly, in this semantic group even infrequent English verbs, namely ascertain,
deduce, infer, suss, and twig, the number of occurrences for which is less than 1,000 and
sometimes even less than 100 instances in the corpus, at least occasionally occur with zerothat complements, especially in the fiction register, as illustrated in example (55):
(55) a. <...> he'd been able toascertain she hadn't left him any lunch. (fiction)
b. I deduce the corpse we have just seen does not belong to James IV. (fiction)
c. She determined she would show him how generous she could be. (fiction)
d. Mr Orton infers there is an inconsistency in my contention. (newspaper)
e. <...> most of the audience has sussed there's no beer on sale.(magazine)
f. I suddenly twigged it was rain. (fiction)
The sentences in examples (54) and (55) also illustrate that the verbs of discovering are used
with zero-that complements in various tenses.
More frequent verbs observe, perceive, read, and sense occur with zero-that
complements only in written texts. The only three verbs of discovering that do not occur in
the BNC with zero-that complements are discern, intuit, and recollect, which are among the
most infrequent verbs; thus even in this small semantic group the tendency of rare verbs to be
used with that-complements with the overt form of the complementizer is observable.
In respect to structural peculiarities, the corpus examples show that closeness of the
clause juncture is as typical feature with the verbs of discovering as with the verbs of saying
and thinking: there are no intervening elements between the matrix verb and the complement
clause, as illustrsated in examples (54) and (55), and the pronominal subjects in the
complement clauses have definite, usually anaphoric, reference, as illustrated in (54a) and
(55a).
On the whole it must be noted that the verbs in the semantic domain of discovering are
generally less frequent than many verbs in the domains of speaking and thinking. In addition,
they are more formal and therefore less frequent in the spoken mode of communication.
The corpus evidence also suggests that the verbs of discovering are not as frequently
used in comment clauses as the verbs of saying and thinking. This goes together with the
observation about the infrequency of the verbs of discovering in the spoken mode, on the one
70
hand, and the variety of tense forms as well as the variety of subject types used in these
matrix clauses. As it was discussed in Chapters 2.1 and 2.2, comment clauses most frequently
incorporate only certain verb forms (typically present simple) and certain types of subjects,
such as I or you. These observations lead to the inference that the verbs of discovering tend to
be used in their primary lexical meanings, which are specific and have not bleached over
time. Therefore most of these verbs have not acquired any additional meanings and typically
are not used parenthetically49. However, some of them, such as calculate, conclude, discover,
infer, notice, and see occur in comment clauses and thus might be ambiguous when used in
the initial position in the sentence, as illustrated in example (56):
(56) a. They calculate they must have tried over seventy over seventy different shades
of paint before finding the correct.
(fiction)
b. He concluded he must indeed be on his way out.
(fiction)
The present tense of the matrix verb in (56a) as well as the modal verb in the complement
clause, which refers to a high degree of certainty, imply that the verb calculate is used in
modifying sense and the matrix clause should be considered peripheral, close in meaning to a
comment clause they think, which would merely refer to epistemic stance but would not be
related to the primary meaning “to determine (the amount or number of something)
mathematically” (NODE, 1998: 258). However, the clause they calculate is not marked by the
comma as a comment clause and thus is ambiguous for the reader. The sentence in (56b) can
also be ambiguous between two semantic interpretations of he thought (functioning as a
qualifier of the proposition) and primary meaning he “arrived at a judgement or opinion by
reasoning” (NODE, 1998: 381). Still, it should be mentioned that the corpus evidence
suggests that most often the verbs of discovering are used in their primary meanings.
Regarding the Lithuanian verbs of discovering, most features described for the English
counterparts coincide and even are more observable. These Lithuanian verbs occur with zerothat complements less frequently than the Lithuanian verbs of saying and thinking. Most of
them do not occur in spoken language, though it should be mentioned again that this might be
the result of the small size of the spoken section of the corpus.
Rare verbs išvesti (padaryti išvadą) “deduce”, įžvelgti “read”and nuvokti “realize” do
not occur in the LLC with zero-that complements.
49
This feature is also described in Ambrazas (2006a), who observes that the linguistic structures which
correspond directly to the semantic stucture, i.e. are iconic, and do not have any competing synonyms, are the
most stable in the language in respect to syntactic changes (2006a: 25).
71
Only two verbs, viz. matyti “see” ir pastebėti “notice”, were found with zero-that
complements in all three registers. However, these were only occasional instances, some of
which are illustrated in example (57):
(57) a. - Dar perjunk. - Tikrai trečias eina. Matai, parašyta.
(spoken)
[- Switch it over again. - It's really (channel) three. (You) see it's written.]
b. Mačiau, ten yra tavo katinėlis.
(fiction)
[I saw your cat was there].
c. Sakau mergaitei: - Matai, gėlių žiedai jau pavytę, duosiu du dolerius. (journalism)
[I said to the girl: (You) see the flower blossoms are withered, I'll give you two
dollars.]
d. <...> pastebėjau, tu man padedi įrašinėt ir kalbėt, <...>.
(spoken)
[<...> (I) noticed you help me to record and speak, <...>.]
e. Pastebėjau, jei vyksta koks nors tarptautinis chorų konkursas, gali būti tikras,
kad
tarp laimėtojų bus ir choras iš Lietuvos.
(journalism)
[(I) noticed (that) if an international choir contes takes place, (you) can be sure
that a Lithuanian choir will be among the winners.]
However, even some of these examples, for instance, (57c) and (57d) can be confused with
parentheticals pragmatically functioning as means for claiming the hearer's attention.
Example (57e), on the other hand, illustrates that the complementizer kad “that” can also be
omitted with the verbs of discovering in case the complement clause is preceded by a
conjunction jei(gu) “if”, which was also observed with the Lithuanian verbs of saying and
thinking. In this respect, it can be added that, differently from the two previous semantic
domains, the verbs in the semantic domain of discovering do not occur with particles esą,
neva “supposedly”, tarsi “as if”or a parenthetical girdi “(you) hear”, which can be explained
by the fact that semantically the verbs in the domain under consideration are not related to
reporting someone's words or thought, which the mentioned lexical items functioning as
conjunctions are reported to be most typical with.
Furthermore, most of the examples of the verbs of discovering with zero-that
complements belong to asyndetic structures with the colon, which also matches the tendency
observed for the verbs of thinking and speaking. These structures occur in both the fiction and
journalism sections of the LLC and were found with the following matrix verbs:
atrasti “discover”, įsidėmėti “notice”, išsiaiškinti “ascertain”, įžiūrėti “discern”, matyti “see”,
72
nustatyti “ascertain”, pamatyti “see”, pastebėti “notice”, nuspręsti “decide”, suprasti
“realize”, suvokti “realize”, sužinoti “learn”.
Some instances are illustrated in example (58):
(58) a. Visai netyčia atradau: Sokratas yra pasakęs, kad menas – gebėjimas gerai
padaryti daiktą.
(journalism)
[(I) accidentally discovered: Socrates said that arts is an ability to make something
well.]
b. Apolinaras tiktai dabar pamatė: aplinkui ruduo, <...>.
(ficiton)
[Only then Apolinaras saw (:) it was autumn, <...>.]
c. Išgirdusi patvirtinimą, nusprendė: jei gims mergaitė, pavadins ja Skaiste.
(journalism)
[Having heard the confirmation (she) decided (:) if she gives birth to a girl, she will
call her Skaiste.]
d. <...> aš išsyk supratau: tu nesi talentingas.
(fiction)
[I at once realized (:) you are not talented.]
Similarly to the previous observations, such instances are more versatile in respect to the
matrix verb tense and the variety of subejcts. However, the absence of intervening elements
between the matrix verb and the subject of the complement clause, as illustrated in example
(58), is typical of these cases as well. On the whole, the closeness of the clause juncture is a
structural peculiarity of zero-that complement clauses observed in both English and
Lithuanian examples from all three semantic domains investigated.
Another peculiarity of the zero-that clauses after the Lithuanian verbs of discovering
is that in most of the examples incorporating the comma as a punctuation mark separating the
clauses, the main clause is manifested by certain verb forms with the subject omitted, thus
these structures might be syntactically classified as parentheticals, with the verb used rather in
modifying sense than in its direct lexical meaning, as illustrated in example (59):
(59) a. Pamatysi, abiem bus daug paprasčiau!
(fiction)
[(You)'ll see it will be easier for both!]
b. Matai, man patinka kurti.
(journalism)
[(You) see, I like creating.]
c. Suprantu, reikia padėti savo šaliai.
(journalism)
73
[(I) understand one should help one's country.]
Thus, instead of being concerned with observation, (59a) rather implies maintaining the
contact with the addressee and encouragement and (59b) rather claims for the hearer's
attention. Example (58c) can be interpreted in two ways: as refering to epistemic stance, like I
know, or as directly meaning “perceive the significance of something” (NODE, 1998: 2016).
Interestingly, the verbs of discovering that are prone to be used parenthetically are not the
same in English and Lithuanian and the cases with them are not as obviously parenthetical as
with the verbs say “sakyti”, think “manyti”, galvoti “think”and know “know”.
Putting together the observations about cases ambiguous between comment clauses (or
parentheticals) and matrix clauses followed by zero-that complements in all three semantic
domains in both languages under consideration several additional observations can be made.
In most cases the verbs incorporated in such ambiguous structures are the same in both
languages, with the verbs say, think and know among the most frequent cases. At the same
time these verbs are reported by corpus-based studies among the most frequent verbs in
English and they are among the most frequent verbs in both BNC and LLC corpora of all the
verbs investigated in this research. On the other hand, more infrequent verbs, such as the
majority of those in the semantic domain of discovering do not occur as often in comment
clauses as the frequent verbs. Thus, it can be inferred that in general the high frequency of the
verbs, such as say, think and know, is due to the fact that very often they are used not in their
primary meaning, but in parenthetical, modifying sense and serve as epistemic markers or
discourse fillers. In such cases the complement clause conveys the main information of the
sentence and becomes more prominent and zero that connection, as it was discussed, is more
typical because it does not directly indicate subordination. This is in accord with the
Lithuanian language, where the bond between the clauses in asyndetic sentences is considered
looser than in syndetic sentences (Drovinas, 1961: 216) and relation between the clauses is on
the whole considered to be closer to coordination than to subordination (Holvoet, 2003: 112),
which allows to semantically interpret the clauses as least equally prominent. On the other
hand, the explanation suggested above would apply to the ambiguous cases of parentheticals
as well: Lithuanian exhibits a general tendency for omitting pronominal subjects, thus the
bleaching of the initial lexical meanings of the verbs, such as sakyti “say”, manyti “think”,
žinoti “know”, and the omission of pronominal subjects in the main clause as a concomitant
(as these subjects do not have a definite referent in such cases), the growing prominence of
the part of the sentence which initially functioned as a complement clause, and the omission
of the complementizer as a concomitant, might have been the reason for grammaticalization
74
of certain verb forms and changing their syntactical status to parentheticals, which resulted in
such structures being classified as extended, but not complex sentences. The co-existence of
alternative structures with the similar semantic interpretation, such as free constructions, or
the same verb forms followed by complements with the overt form of kad (as well as the
cases with modifying meaning of the matrix clause with the overt form of that in English) is
in accord with the fact that grammatical and syntactical linguistic changes are a very slow and
gradual process allowing co-existence of different syntactic forms sharing similar semantic
meaning. The same reasons would also account for the fact that rare verbs tend to be used
with the overt form of the complementizer that, as it is reported in numerous corpus-based
studies. Rare verbs are most often used in their primary meanings and comprise the main
assertion of the sentence with the complement realised by a clause; therefore, on the whole
there is no underlying reason for that to be omitted with them. Their occasional occurence
with zero-that complements can be rather explained by other factors considered influencial,
such as the spoken mode of communication or informality of style.
75
CONCLUSIONS
The thorough examination of the corpus evidence collected for the verbs of saying,
thinking and discovering followed by complement clauses with the complementizer that
omitted in the English and Lithuanian languages led to the following conclusions.
The zero-that complements are by far more frequent in English than in Lithuanian,
which proved to be the tendency for all three semantic domains considered. On the whole,
more English verbs are used with zero-that complements as well as with a wider structural
variety of the matrix and complement clauses in all three investigated registers.
Surprisingly few cases of zero-that complements occur with Lithuanian verbs in the
spoken register, which, on the one hand, is opposite to the English language, in which the
spoken mode of communication is considered a strong factor influencing the omission of that,
and, on the other hand, contradicts the theoretical claims about asyndetic sentences being
more characteristic of Lithuanian spoken language. However, this might be a result of the
small size and content insufficiency of the spoken section of the LLC.
Most Lithuanian cases of zero-that complements incorporate the colon as a
punctuation mark (the comma is less common), which is absolutely not typical for equivalent
English sentences and is due to the fact that the colon as a punctuation mark is generally
infrequent in English.
In Lithaunian zero-that is likely if the complement clause is preceded by an evidential
marker, such as esą, neva “supposedly”, tarsi, lyg “as if”, girdi “(you) hear”, or gal “maybe”,
which proves the theoretical claims that in contemporary Lithuanian these linguistic items
tend to function as conjunctions. The English zero-that complement clauses do not have
lexical elements alternative to the complementizer that.
A major similarity between the English and the Lithuanian zero-that complement
clauses is the structural peculiarity known as closeness of clause juncture, which is manifested
by the absense of intervening elements between the matrix verb and the subject of the
complement clause as well as by subjects in both the main and complement clauses being
pronominal and having definite and usually anaphoric reference. This finding is in accord
with previous corpus-based studies on the English zero-that complement clauses.
Another major similarity between zero-that complements in the two languages is that
the most frequent matrix verbs in both languages, such as say, think, and know, favour the
deletion of the complementizer, while the rare matrix verbs favour the retention of that. This
finding agrees with the previous corpus-based studies on English zero-that complement
clauses.
76
Last but not least, in both languages matrix clauses with the most frequent verbs of
saying and thinking can be often confused with parentheticals, which function as pragmatic
markers or discourse fillers and tend to be used with zero that. This is the result of the
bleaching of the lexical meaning of the matrix verb, on the one hand, and the growing
informational prominence of the rest of the sentence, on the other hand. The structural and
semantic ambiguity in such cases can be explained by the diachronic change resulting in
syntactic reanalysis and grammaticalization, which is more obvious in Lithuanian, as
sentences including parentheticals are classified as extended simple instead of complex. The
use of frequent verbs in parenthetical meaning can also account for the tendency of frequent
verbs to be used with zero complementizer and the propensity of the rare verbs to favour the
overt form of that. This finding accords with the results of the latest researches on English
verbs of saying and thinking with zero-that complements.
The current study was concerned with qualitative research, which enables one to detect
certain tendencies for the deletion of the complementizer that in English and Lithuanian but
does not allow to account for the significance of factors influencing the choice of the zero or
overt form of the complementizer. Therefore, taking into account that no corpus-based studies
on these structures have been conducted so far in Lithuanian, further corpus-based studies on
the issue would be essential. Firstly, a corpus-based study on Lithuanian verbs taking thatcomplements would result in a complete list of Lithuanian verbs of this lexico-grammatical
pattern. Secondly, quantitative corpus-based studies would provide more information upon
how influential the factors favouring the choice of zero-that arein Lithuanian. Thirdly, corpusbased studies, especially of spoken language, on Lithuanian structures ambiguous between
matrix clauses and parentheticals would help to establish more explicit criteria for
distinguishing these structures. Finally, parallel corpus-based studies of English-Lithuanian
and Lithuanian-English texts would give a more detailed account for similarities and
differences between that- and zero-that complement clauses in the two languages, which
would be of substantial value in translation theory as well as in language teaching.
77
SANTRAUKA
Šiame tyrime nagrinėjami veiksmažodžių valdomi anglų bei lietuvių kalbos
prijungiamieji aiškinamieji sakiniai. Tai tekstynais paremta lyginamoji anlizė, skirta šių
sintaksinių struktūrų abiejose kalbose panašumams ir skirtumams išsiaiškinti bei veiksniams,
darantiems įtaką jungtuko kad/jog praleidimui, nustatyti. Tyrimu siekiama atsakyti į klausimą,
kokie
kontekstinių
faktorių
panašumai
ir
skirtumai
būdingi
prijungiamiesiems
aiškinamiesiems sakiniams lietuvių ir anglų kalboje.
Tyrimo
tikslas
–
išanalizuoti
anglų
bei
lietuvių
kalbos
prijungiamuosius
aiškinamuosius sakinius bei remiantis tekstynų duomenimis aprašyti ir palyginti jungtuko
kad/jog praleidimo atvejus tokio tipo sakiniuose. Tikslui pasiekti iškelti šie uždaviniai:
arpašyti ir palyginti prijungiamųjų aiškinamųjų sakinių bruožus anglų ir lietuvių kalboje,
surinkti bejungtukių prijungiamųjų aiškinamųjų sakinių, valdomų kalbėjimo, mąstymo bei
suvokimo veiksmažodžių, tekstyno pavyzdžius ir aprašyti bei palyginti tekstynų duomenis,
skiriant ypatingą dėmesį šių sakinių vartojimo kontekstiniams faktoriams, tokiems kaip
bendravimo pobūdis (rašytinė ar sakytinė kalba) bei registras, aiškinamuosius sakinius
valdančių veiksmažodžių savybės bei tokių sakinių sandaros ypatumai, taip pat šių sintaksinių
struktūrų panašumams ir skirtumams anglų ir lietuvių kalboje. Tyrimas grindžiamas
lyginamosios ir kontekstinės analizės metodais bei kokybinės analizės metodu aprašant ir
lyginant iš tekstynų surinktus duomenis.
Surinkti tekstynų duomenys rodo, kad prijungiamieji aiškinamieji sakiniai su praleistu
jungtuku kad/jog žymiai būdingesni anglų nei lietuvių kalbai. Stebėtinai mažai tokių sakinių
pasitaikė lietuvių šnekamojoje kalboje, o tai prieštarauja teorijoje vyraujančiai nuomonei, kad
asindetiniai sakiniai labiau būdingi šnekamajai nei rašytinei kalbai; tačiau toks rezultatas gali
būti lietuvių kalbos tekstyno šnekamosios kalbos dalies mažos apimties bei riboto turinio
pasekmė. Taip pat paaiškėjo, kad daugelyje lietuvių kalbos asindetinių aiškinamųjų sakinių
šalutinis dėmuo skiriamas nuo pagrindinio dvitaškiu (kablelis vartojamas žymiai rečiau),
tačiau dvitaškis absoliučiai nebūdingas anglų kalbos asindetiniems aiškinamiesiems
sakiniams. Tekstyno duomenys rodo, kad jungtuko kad/jog nebuvimas lietuvių kalboje
tikėtinas, jei šalutinis dėmuo pradedamas evidencialumo raiškos priemonėms priskiriamais
žodžiais esą, neva, lyg, tarsi ir girdi, kurie šiuolaikinėje lietuvių kalboje neretai atlieka
jungtukų funkciją, bet anglų kalbai tokios struktūros nebūdingos. Abiejose kalbose
bejungtukiai aiškinamieji sakiniai panašūs sandaros atžvilgiu: nėra antros eilės sakinio dalių,
įsiterpiančių tarp valdančiojo veiksmažodžio ir šalutinio dėmens veiksnio. Abiejose kalbose
jungtukas kad/jog dažniau praleidžiamas su dažniausiai kalboje vartojamais veiksmažodžiais.
78
Be to, abiejose kalbose būna dviprasmiškų atvejų, kai bejungtukių aiškinamųjų sakinių
pagrindinį dėmenį sunku atskirti nuo įterpinio arba komentuojamojo sakinio. Toks struktūrinis
ir semantinis dviprasmiškumas gali būti aiškinamas diachroniniais sintaksiniais pasikeitimais.
79
REFERENCES
Corpora
5. BNC. British National Corpus [online]. University of Oxford, 2001 (second edition),
accessed 23 June, 2013, available from: http://corpus.byu.edu/bnc/
6. LLC. The Corpus of the Contemporary Lithuanian Language [online]. Vytautas
Magnus University, Centre of Computational Linguistics, 1998-2013, accessed 15
September, 2013, available from: http://tekstynas.vdu.lt/tekstynas/
Dictionaries
2. DLKŽ 2012. Keinys, S. (ed.) 2012. Dabartinės lietuvių kalbos žodynas, 7th edition.
Vilnius: Lietuvių kalbos institutas.
3. LKŽ Lietuvių kalbos žodynas [online], accessed 23 June, 2013, available from:
http://www.lkz.lt/startas.htm
4. NODE 1998. Pearsall, J. (ed.) 1998. The New Oxford Dictionary of English. Oxford:
Clarendon Press.
5. Piesarskas, B. 1998. The English-Lithuanian Dictionary. Vilnius: Alma Litera.
6. Sližienė, N. 1994. Lietuvių kalbos veiksmažodžių junglumo žodynas. Tomas I. Vinius:
Mokslo ir enciklopedijų leidykla.
7. Sližienė, N. 1998. Lietuvių kalbos veiksmažodžių junglumo žodynas. Tomas II (1).
Vinius: Mokslo ir enciklopedijų leidybos institutas.
8. Sližienė, N. 2004. Lietuvių kalbos veiksmažodžių junglumo žodynas. Tomas II (2).
Vinius: Lietuvių kalbos institutas.
Other references
4. Akelaitis, G. 2002. Veiksmažodiniai pagrindinio dėmens formos įterpiniai. Žmogus ir
žodis, I, 3-8, accessed 25 March, 2014, available from:
http://www.biblioteka.vpu.lt/zmogusirzodis/PDF/didaktinelingvistika/2002/akelait3-9.pdf
5. Ambrazas, V. (ed.) 2005.Dabartinės lietuvių kalbos gramatika. Vilnius: Mokslo ir
enciklopedijų leidybos institutas, Lietuvių kalbos institutas.
6. Ambrazas, V. (ed.) 2006. Lithuanian Grammar. Vilnius: Baltos lankos.
7. Ambrazas, V. 2006a. Lietuvių kalbos istorinė sintaksė. Vilnius: Lietuvių kalbos
institutas.
8. Anderson, W. & Cobett, J. 2009. Exploring English with Online Corpora: An
80
Introduction. Basingstoke: Palgrave Macmillan.
9. Balkevičius, J. 1963. Dabartinės lietuvių kalbos sintaksė. Vilnius: Valstybinė politinės
ir mokslinės literatūros leidykla.
10. Biber, D. & James, J. K. 2009. Quantitative methods in corpus linguistics, in Lüdeling,
A. & Kytö, M. (eds.) Corpus Linguistics. An International Handbook, Vol.2.BerlinNew York: Walter de Gruyter.
11. Biber, D. 1999. A Register Perspective on Grammar and Discourse: Variability in the
Form and Use of English Complement Clause. Discourse Studies, 1: 131. DOI:
10.1177/1461445699001002001,
accessed
15
March,
2014,
available
from:
http://dis.sagepub.com/content/1/2/131
12. Biber, D., Conrad, S., & Reppen, R. 1998. Corpus Linguistics. Investigating Language
Structure and Use. Cambridge: Cambridge University Press.
13. Biber, D., Johansson, S., Leech, G., Conrad, S., & Finogan, E. 1999. Longman
Grammar of Spoken and Written English. Harlow: Pearson Education Limited.
14. Bogaert, J. V. 2011. I Think and Other Complement-taking Mental Predicates: A Case
of and for Constructional Grammaticalization. Linguistics, 49: 2, 295–332 DOI
10.1515/LING.2011.009, accessed 22 June, 2013, available from:
http://web.a.ebscohost.com.ezproxy.vpu.lt/ehost/pdfviewer/pdfviewer?sid=ec45754e-60824385-b6fe-7a5d6b23fd52%40sessionmgr4004&vid=4&hid=4214
15. Brinton, L. J. 2008. The Comment Clause in English: Syntactic Origins and
Pragmatic
Development,
[book
online].
Cambridge:
Cambridge
University
Press,accessed 15 March 2014, available from:
http://assets.cambridge.org/97805218/86734/excerpt/9780521886734_excerpt.pdf
16. Būda, V. 1979. Papildinio prijungiamųjų sakinių sandaros elementai. Kalbotyra, XXX
(1), 1979.
17. Būda, V. 1986. Lietuvių kalbos sudėtiniai prijungiamieji sakiniai. Vilnius: LTSR
aukštojo ir specialiojo vidurinio mokslo ministerijos Leidybinė redakcinė taryba.
18. Conrad, S. 2010. What Can a Corpus Tell Us about Grammar?, in O'Keeffe, A. &
McCarthy, M. (eds.) Rootledge Handbok of Corpus Linguistics. London and New
York: Rootledge.
19. Dambriūnas, L. 1963. Lietuvių kalbos sintaksė. Chicago: Pedagoginis lituanistikos
institutas.
20. Detmar Meuers, W. & Müller, S. 2009. Corpora and Sintax, in Lüdeling, A. & Kytö ,
M. (eds.) Corpus Linguistics. An International Handbook, Vol.2. Berlin and New
York: Walter De Gruyter.
81
21. Dixon, R. M. W. & Aikhenvald, A. Y. (eds.) 2006. Complementation. A CrossLinguistic Typology. Oxford: Oxford University Press.
22. Drotvinas, L. 1961. Apie dabartinės lietuvių literatūrinės kalbos prijungiamuosius
sakinius be jungtukų, Dabartinė lietuvių kalba, 178-205. Vilnius: Valstybinė politinės
ir mokslinės literatūros leidykla.
23. Elsness, J. 1982. That vs Zero Connective in English Nominal Clause, ICAME
Newsletter of the International Computer Archive of modern English, 6, 1-45.
24. Elsness, J. 1984. That or Zero? A Look at the Choice of Object Clause Connectivein a
Corpus of American English. English Studies, 65: 519-33, accessed 10 April, 2013,
available
from:
http://web.ebscohost.com.ezproxy.vpu.lt/ehost/pdfviewer/pdfviewer?vid=3&sid=61ba
103d-7571-4e4f-923a-983a3474a03d%40sessionmgr15&hid=18
25. Finegan, E. & Biber, D. 1995. That and Zero Complementizers in Late Modern
English: Exploring Archer from 1650-1990, in Aarts, B. and Meyer, C. F. (eds.) The
Verb in Contemporary English. Theory and Description, 241-257. Cambridge,UK:
Cambridge University Press.
26. Frajzyngier, Z. & Jasperson, R. 1991. That-clauses and other complements, LiNgua,
83, 133-153. North Holland.
27. Francis, G., Hunston, S., & Manning, E. (eds.) 1998. Collins COBUILDGrammar
Patterns 1: Verbs. New York: Harper Collins Publishers Ltd.
28. Greenbaum, S., Nelson, G., & Weitzman, M. 1996. Complement clauses in English, in
Short, M. and Thomas, J.A. (eds.) Using corpora for language research. London:
Longman.
29. Gries, S. T. & Stefanowitsch, A. 2009. Corpora and grammar, in Lüdeling, A. and
Kytö , M. (eds.) Corpus Linguistics. An International Handbook, Vol.2. Berlin and
New York: Walter de Gruyter.
30. Holvoet, A. & Judžentis, A. 2003. Sudėtinio prijungiamojo sakinio aprašymo
pagrindai, in Lietuvių kalbos gramatikos darbai 1, Sintaksinių ryšių tyrimai. Vilnius:
Lietuvių kalbos institutas.
31. Holvoet, A. 2003. Sujungiamieji ir prijungiamieji sakiniai formos bei funkcijos
požiūriu, in Lietuvių kalbos gramatikos darbai 1, Sintaksinių ryšių tyrimai. Vilnius:
Lietuvių kalbos institutas.
32. Huddleston, R. & Pullum, G. K. 2002. The Cambridge Grammar of the English
Language. Cambridge: Cambridge University Press.
33. Hudson-Ettle, D. 2002. Nominal that Clauses in Three Regional Varieties of English :
82
A Study of the Relevance of Text Type, Medium, and Syntactic Function, Journal of
English Linguistics, 30: 258. DOI: 10.1177/0075424202030003003, accessed 12
February, 2014, available from: http://eng.sagepub.com/content/30/3/258
34. Judžentis, A. 2002. Mikalojaus Daukšos Katekizmo (1595) sudėtiniai aiškinamieji
sakiniai, Acta linguistica lituanica, XLVII, 19-29, accessed 20 March, 2014, available
from: www.baltistica.lt/index.php/baltistica/article/download/1066/992
35. Kaltenböck, G. 2004. That or no that? – That Is the Question: on Subordinator
Suppression in Extraposed Subject Clauses,VIEW[S], 13: 1, 49-68.
36. Kaltenböck, G. 2008. Prosody and function of English comment clauses,Folia
Linguistica, 42: 1, 83–134. Mouton de Gruyter – Societas Linguistica Europaea,
accessed 10 March 2014, available from:
http://web.b.ebscohost.com.ezproxy.vpu.lt/ehost/pdfviewer/pdfviewer?sid=53b8201a-1e284534-b749-fb05ea0b5a89%40sessionmgr112&vid=5&hid=113
37. Kaltenböck, G. 2009. English comment clauses: position, prosody, and scope. AAA.
Arbeiten aus Anglistik und Amerikanistik, 34:1, 51-77, accessed 10 March, 2014,
available from:
file:///D:/My%20Documents/Downloads/916-1502-1-SM.pdf
38. Kaltenböck, G. 2009a.Initial I think: Main or CommentClause? Discourse and
Interaction,
2:
1,
49-70,
accessed
15
March,
2014,
available
from:
http://anglistik.univie.ac.at/fileadmin/user_upload/dep_anglist/Mitarbeiter/Kaltenb%C
3%B6ck/Discourse_and_Interaction_Initial_I_think.pdf
39. Kearns, K.2007. Epistemic Verbs and Zero Complementizer, English Language and
Linguistics,11: 03, 475-505. DOI: 10.1017/S1360674307002353, accessed 12
February,
2014,
available
fromhttp://joutnals.cambridge.org/abstract_S1360674307002353
40. Kirk, J. M. 1997. Subordinate Clauses in English, Journal of English Linguistics, 25:
4, 349-64. DOI: 10.1177/007542429702500409, accessed 20 March, 2014, available
from: http://eng.sagepub.com/content/25/4/349
41. Labutis V. 2002. Lietuvių kalbos sintaksė. Vilnius: Vilniaus universiteto leidykla.
42. Marcinkevičienė, R. 2000. Tekstynų lingvistika, Darbai ir dienos, 24: 7. Kaunas:
Vytauto Dydžiojo universiteto leidykla, accessed 15 March, 2014, available
from:http://donelaitis.vdu.lt/publikacijos/marcinkeviciene.pdf
43. Marcinkevičienė, R. 2010. Lietuvių kalbos kolokacijos. Kaunas: Vytauto Didžiojo
Universitetas.
44. Masilionis, J. 1985. Lietuvių kalbos sintaksė. Chicago: JAV LB švietimo taryba.
83
45. McCarthy, M. & O'Keeffe, A. 2010. Historical Perspective: What Are Corpora and
How Have They Evolved?, in O'Keeffe, A. & McCarthy, M. (eds.) Rootledge
Handbok of Corpus Linguistics. London and Ndew York: Rootledge.
46. McDavid, V. 1964. The Alternation of ‘That’ and Zero in Noun Clauses, in American
Speech, 39, 102-13.
47. McEnery, T. & Hardie, A. 2012. Corpus Linguistics. Cambridge: Cambridge
University Press.
48. McEnery, T. & Wilson, A. 1996. Corpus Linguistics. Edinburgh: Edinburgh University
Press.
49. Milūnaitė, R. 2005. Kalbėjimo gyvumas – kalbos gyvybė. Vaikų ir jaunimo literatūros
vertimas. Conference Materials [2005-11-16], accessed 10 April, 2014, available from:
http://www.llvs.lt/?recensions=37
50. Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. 1972. A Grammar of
Contemporary English. London: Longman Group Ltd.
51. Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J.1995. A Comprehensive Grammar
of the English Language. London and New York: Longman.
52. Rissanen, M. 1991. On the History of that/zero as Object Clause Links in English, in
Aijmer, K. and Altenberg, B. (eds.) English corpus linguistics, 272-289. London:
Longman.
53. Ruskan, A. 2010. Evidencialumo raiškos priemonės lietuvių kalboje, Lietuvių kalba, 4,
accessed 15 March, 2014, available from: www.lietuviukalba.lt
54. Sirtautas, V. &Grenda, Č. 1988. Lietuvių kalbos sintaksė. Vilnius: Mokslas.
55. Storms, G. 1966. That-Clauses in Modern English, English Studies, 47, 249-270.
56. Thompson, S. A. & Mulac, A. 1991. The Discourse Conditions for the Use of the
Complementizer that in Conversational English, Journal of Pragmatics, 15, 237-251.
57. Tognini Bonelli, E. 2010. Theoretical Overview of the Evolution of Corpus
Linguistics, in O'Keeffe A. & McCarthy, M. (eds.) Rootledge Handbok of Corpus
Linguistics. London and Ndew York: Rootledge.
58. Wiemer, B. 2007. Lexical Markers of Evidentiality in Lithuanian, Rivista di
Linguistica 19: 1, 173-208, accessed 2 April, 2014, available from:
http://linguistica.sns.it/RdL/19.1/09.wiemer.pdf
84
APPENDIX I
The lists of the English verbs investigated
The verbs of saying
1. admit
2. advise
3. advocate
4. affirm
5. allege
6. announce
7. argue
8. assert
9. attest
10. aver
11. beg
12. boast
13. brag
14. caution
15. certify
16. claim
17. command
18. comment
19. complain
20. concede
21. confess
22. confide
23. contend
24. declare
25. decree
26. demand
27. deny
28. dictate
29. disclose
30. divulge
31. emphasize
32. enthuse
33. explain
34. forecast
35. foretell
36. grouse
37. grumble
38. hint
39. imply
40. indicate
41. insinuate
42. insist
43. instruct
44. joke
45. lament
46. maintain
47. mention
85
48. moan
49. opine
50. ordain
51. order
52. plead
53. pledge
54. posit
55. postulate
56. pray
57. preach
58. predict
59. proclaim
60. profess
61. promise
62. prophesy
63. propose
64. quip
65. recommend
66. recount
67. remark
68. remonstrate
69. report
70. request
71. reveal
72. say
73. signal
74. signify
75. sneer
76. specify
77. stipulate
78. stress
79. submit
80. suggest
81. swear
82. testify
83. threaten
84. underline
85. underscore
86. wager
87. warn
Verbs of saying that belong to more than one semantic domain (examined in other
semantic groups):
accept
acknowledge
affirm
concur
conjecture
THINK
THINK
THINK
THINK
THINK
86
estimate
guess
hypothesize
marvel
muse
regret
speculate
surmise
theorize
vow
THINK
THINK
THINK
THINK
THINK
THINK
THINK
THINK
THINK
THINK
conclude
note
observe
recall
recollect
DISCOVER
DISCOVER
DISCOVER
DISCOVER
DISCOVER
The verbs of thinking
1. accept
2. acknowledge
3. agree
4. anticipate
5. appreciate
6. assume
7. believe
8. bet
9. concur
10. conjecture
11. consider
12. decide
13. disagree
14. disbelieve
15. doubt
16. dread
17. dream
18. envisage
19. envision
20. estimate
21. expect
22. fancy
23. fantasize
24. foresee
25. forget
26. fret
27. guess
28. hallucinate
29. hold
30. hope
31. hypothesize
32. imagine
33. intend
87
34. know
35. marvel
36. muse
37. prefer
38. presume
39. reckon
40. regret
41. rejoice
42. remember
43. resolve
44. speculate
45. suppose
46. surmise
47. suspect
48. theorize
49. think
50. vow
51. wonder
52. worry
Verbs of thinking that belong to more than one semantic domain (examined in other
semantic groups):
calculate
realize
see
DISCOVER
DISCOVER
DISCOVER
pray
SAY
The verbs of discovering
1. ascertain
2. calculate
3. conclude
4. deduce
5. determine
6. discern
7. discover
8. infer
9. intuit
10. learn
11. notice
12. observe
13. perceive
14. read
15. realize
16. recall
17. recognize
18. recollect
19. see
20. sense
21. suss
88
22. twig
Verbs of discovering that belong to more than one semantic domain (examined in other
semantic groups):
decide
guess
remember
THINK
THINK
THINK
The lists of the Lithuanian verbs researched
The verbs of saying
1. (iš/pa)aiškint
2. akcentuoti
3. anonsuoti
4. atskleisti
5. bambėti
6. bėdoti
7. byloti
8. burbtelėti
9. dievagotis
10. dievažytis
11. ginčytis
12. girtis
13. pagrasinti
14. informuoti
15. įsakyti
16. įspėti
17. pasakoti
18. išpažinti
19. įtikinėti
20. jokauti
21. komentuoti
22. lažintis
23. liepti
24. liudyti
25. maldauti
26. melsti
27. minėti
28. murmėti
29. neigti
30. nurodyti
31. pabrėžti
32. pamokslauti
33. patikslinti
34. postuluoti
35. išpranašauti
36. pareikšti
37. pranešti
38. prašyti
“explain”
“emphasize”
“announce”
“reveal”
“grouse”
“moan”
“proclaim”
“grumble”
“swear”
“swear”
“argue”
“boast”
“threaten”
“inform”
“command”
“warn”
“tell, say”
“confess”
“urge”
“joke”
“comment”
“wager”
“order”
“certify”
“beg”
“beg”
“mention”
“grouse”
“deeny”
“specify”
“emphasize”
“preach”
“specify”
“postulate”
“foretell”
“declare”
“report”
“ask”
89
39. pridėti (kalbant)
40. prieštarauti
41. priminti
42. prisipažinti
43. prisiekti
44. prognozuoti
45. propaguoti
46. protestuoti
47. reikalauti
48. rekomenduoti
49. sakyti
50. sielotis
51. signalizuoti
52. pasiūlyti
53. paskelbti
54. skųstis
55. susitarti
56. šaipytis
57. šmaikštauti
58. prasitarti
59. teigti
60. teisintis
61. tyčiotis
62. patvirtinti
63. užsiminti
64. pažadėti
65. pažymėti
66. aimanuoti
67. kalbėti
68. meluoti
69. pasakoti
70. pridurti
71. prikišti (daryti priekaištą)
72. priekaištauti
73. rėkti (skųstis)
74. šnekėti
75. tikinti
“remark”
“remonstrate”
“remind”
“profess”
“swear”
“forecast”
“advocate”
“remonstrate”
“demand”
“recommend”
“say”
“lament”
“signal”
“suggest”
“declare”
“complain”
“agree”
“sneer”
“quip”
“blurt”
“state”
“plead”
“sneer”
“affirm”
“hint”
“promise”
“stress”
“complain”
“speak, say”
“lie”
“tell, say”
“add”
“expostulate”
“expostulate”
“complain”
“speak, say”
“assure”
Verbs of saying that belong to more than one semantic domain (examined in other
semantic groups):50
argumentuoti
THINK
atmesti
THINK
atsiminti
THINK
džiūgauti
THINK
gailėtis
THINK
įvertinti
THINK
manyti
THINK
piktdžiugiauti
THINK
prisiminti
THINK
50 See the list of the semantic domain indicated for the translation.
90
samprotauti
žavėtis
spėlioti
spėti
stebėtis
sutikti
THINK
THINK
THINK
THINK
THINK
THINK
The verbs of thinking
1. abejoti
2. argumentuoti
3. atmesti
4. atsiminti
5. atspėti
6. baimintis
7. bijoti
8. džiaugtis
9. džiūgauti
10. fantazuoti
11. gailėtis
12. galvoti
13. įsivaizduoti
14. įtarti
15. įvertinti
16. jaudintis
17. jausti
18. laukti
19. manyti
20. nerimastauti
21. nerimauti
22. nerimti
23. nervintis
24. numanyti
25. nuogastauti
26. nuspręsti
27. nutarti
28. pajusti
29. pagrįsti
30. pamiršti
31. pasiryžti
32. perprasti
33. piktdžiūgiauti
34. prisiminti
35. prisiekti
36. samprotauti
37. sapnuoti
38. spėlioti
39. spėti
40. stebėtis
41. susierzinti
42. sutikti
43. svajoti
“doubt”
“argue”
“reject”
“remember”
“guess”
“dread”
“fear”
“rejoice”
“rejoice”
“phantasize”
“regret”
“think”
“envision”
“suspect”
“estimate”
“worry”
“feel”
“hope”
“think”
“worry”
“worry”
“fret”
“fret”
“anticipate”
“fear”
“decide”
“decide”
“feel”
“reason”
“forget”
“resolve”
“grasp”
“gloat”
“remember”
“swear”
“reason”
“dream”
“conjecture”
“conjecture”
“marvel”
“fret”
“accept”
“dream”
91
“theorize”
“believe”
“hope”
“hope”
“marvel”
“know”
“phantasize”
“grasp”
44. teoretizuoti
45. tikėti
46. tikėtis
47. viltis
48. žavėtis
49. žinoti
50. prasimanyti
51. nutuokti
Verbs of thinking that belong to more than one semantic domain (examined in other
semantic groups):
apskaičiuoti
įsisąmoninti
nustatyti
pripažinti
spręsti, nuspręsti
suprasti, susiprasti
suvokti
sužinoti
DISCOVER
DISCOVER
DISCOVER
DISCOVER
DISCOVER
DISCOVER
DISCOVER
DISCOVER
maldauti
melsti
SAY
SAY
The verbs of discovering
1. atrasti
2. įsidėmėti
3. įsisamoninti
4. įsitikinti
5. išprotauti
6. išsiaiškinti
7. išvesti
8. įžiūrėti
9. įžvelgti
10. matosi
11. matyti
12. nustatyti
13. nuvokti
14. matyti
15. pastebėti
16. pripažinti
17. nuspręsti
18. suprasti
19. suvokti
20. sužinoti
21. užfiksuoti
“discover”
“notice”
“realize”
“ascertain”
“conclude”
“ascertain”
“deduce”
“discern”
“read”
“it is clear”
“it is clear”
“ascertain”
“realize”
“see”
“notice”
“admit”
“decide”
“realise”
“realize”
“learn”
“register”
Verbs of discovering that belong to more than one semantic domain (examined in other
semantic groups):
apskaičiuoti
THINK
92
atsiminti, prisiminti
numanyti
spėti
THINK
THINK
THINK
93