A corpus-based study of the use of adverbial intensifiers ...

Ref. code: 25595821040408YGVRef. code: 25595821040408YGV

A CORPUS-BASED STUDY OF THE USE

OF ADVERBIAL INTENSIFIERS

WITH SEMANTIC PROSODY

BY

MISS SIRIPORN GAMPAENGGAEW

AN INDEPENDENT STUDY PAPER SUBMITTED IN PARTIAL

FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF ARTS IN CAREER ENGLISH FOR

INTERNATIONAL COMMUNICATION

LANGUAGE INSTITUTE, THAMMASAT UNIVERSITY

ACADEMIC YEAR 2016

COPYRIGHT OF THAMMASAT UNIVERSITY


A CORPUS-BASED STUDY OF THE USE

OF ADVERBIAL INTENSIFIERS

WITH SEMANTIC PROSODY

BY

MISS SIRIPORN GAMPAENGGAEW

AN INDEPENDENT STUDY PAPER SUBMITTED IN PARTIAL

FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE

OF MASTER OF ARTS IN CAREER ENGLISH FOR

INTERNATIONAL COMMUNICATION

LANGUAGE INSTITUTE, THAMMASAT UNIVERSITY

ACADEMIC YEAR 2016

COPYRIGHT OF THAMMASAT UNIVERSITY


(1)

Independent Study Paper Title A CORPUS-BASED STUDY OF THE USE OF

ADVERBIAL INTENSIFIERS WITH SEMANTIC

PROSODY

Author Miss Siriporn Gampaenggaew

Degree Master of Arts

Major Field/Faculty/University Career English for International Communication

Language Institute

Thammasat University

Independent Study Paper Advisor Assistant Professor Supakorn Phoocharoensil, Ph.D.

Academic Years 2016

ABSTRACT

In the use of English, word choice is probably one of the most important

keys to accomplishing effective communication. As the words carry not only

denotative meaning but also connotative and attitudinal meaning; therefore, it is

necessary, in particular, for L2 learners to be capable of applying words properly for

the correct and natural use of the language. This paper is thus aimed at investigating

the semantic prosody of the adverbial intensifiers really, certainly and clearly by

examining their adjective and verb collocations when in use and through observing

the authentic data of concordance lines from a language corpus which is the Corpus of

Contemporary American English (COCA). One thousand concordance lines of each

target word were analyzed by categorizing its prosodies of their adjective and verb

collocations. The results showed that these three adverbial intensifiers carry a stronger

positive semantic prosody than negative prosody as they are frequently applied to

intensify positive words, especially the words relating to states of mind, general

evaluation and description. Also, it was found that each AI has a particular set of

adverbs and verbs that it tends to co-occur with. Thereby, English learners should be

aware of using the AIs in accordance with their semantic prosodies for naturalness in

the use of English.

Keywords: adverbial intensifier, semantic prosody, collocation, corpus-based data


(2)

ACKNOWLEDGEMENTS

The completion of this study is a result of the support and encouragement

from several people. I would like to express my deep gratitude and appreciation to

them for their precious help and support.

Firstly, I would like to express my sincere gratitude to my admirable

advisor Asst. Prof. Supakorn Phoocharoensil for the continuous support and

motivation and for his comments and suggestions during the study. Besides my

advisor, I would like to thank all of the instructors at the Language Institute of

Thammasat University for providing me with extensive knowledge and useful

guidance, which helped me conduct this research study.

My sincere thanks also go to my CEIC classmates, in particular my close

ones, who always encouraged and helped me solve the problems which occurred

during the study; and the administrators of the Master of Arts Program in Career

English for International Communication at LITU for their prompt support and

coordination. Without their precious support, it would not be possible for me to

complete this study.

Finally, I must express my very profound gratitude to my family

including my mother, my father, my younger brother and all of my aunts, uncles and

cousins for their understanding, unfailing support and continuous encouragement

throughout my years of study. In addition, I would like to give a million thanks to Ms.

Pimsai Lertanugrom who is like my real sister. She is the one who has been inspiring

me and introduced me to study here. During the study, she always supported,

encouraged and soothed me.

This accomplishment would not have been possible without all of them.

Thank you very much.

Miss Siriporn Gampaenggaew


(3)

TABLE OF CONTENTS

Page

CHAPTER 1 INTRODUCTION 1

1.1 Background 1

1.2 Research Questions 4

1.3 Objectives of the Study 4

1.4 Definition of Terms 4

1.5 Scope of the Study 5

1.6 Significance of the Study 6

1.7 Limitations of the Study 7

1.8 Organization of the Study 7

CHAPTER 2 LITERATURE REVIEW 8

2.1 Corpus Linguistics 8

2.1.1 Definition and Characteristics 8

2.1.2 Types of Language Corpora 9

2.1.3 Corpus Processing Techniques 11

2.1.4 Use of Corpora 14

2.2 Adverbial Intensifiers 15

2.2.1 Definition and Types of Adverbial Intensifiers 15

2.2.2 Use of Intensifiers 18

2.3 Semantic Prosody 20

2.3.1 Background and Definition of Semantic Prosody 20

2.3.2 Main Features and Categories of Semantic Prosody 23

2.3.3 Semantic Prosody and Semantic Preference 24

2.4 Relevant Research 24


(4)

CHAPTER 3 RESEARCH METHODOLOGY 27

3.1 Target Adverb Intensifiers 27

3.2 Corpora 27

3.3 Procedures 27

3.3.1 Research Design 27

3.3.2 Selection of Target Adverbial Intensifiers 28

3.3.3 Data Collection 28

3.4 Data Analysis 31

CHAPTER 4 RESULTS 32

4.1 The Results of the Adjective and Verb Collocations 32

of the Adverbial Intensifier Really in COCA


of the Adverbial Intensifier Certainly in COCA


of the Adverbial Intensifier Clearly in COCA

CHAPTER 5 CONCLUSION, DISCUSSION AND RECOMMENDATION

44

5.1 Summary of the Findings and Discussion 44

5.2 Conclusion 47

5.3 Recommendations for Further Research 47

REFERENCES 49


(5)

LIST OF TABLES

Tables Page

1.1 Top 10 emphasizing adverbs in COCA 6

2.1 The most frequent words: ten-million-word corpus (CIC) 12

2.2 Frequency of expressions with time 13

4.1 Example of concordance lines of really collocating with top three 32

pleasant adjectives in COCA

4.2 Example of concordance lines of really in the sentence patterns S. + 34

Auxiliary V. + not + really + Main V., and S. + really + Auxiliary V. +

not + Main V. in COCA

4.3 Example of concordance lines of certainly collocating with frequent 35

positive adjectives in COCA

4.4 Example of concordance lines of certainly collocating with adjectives 36

ending with –able and –ible, comparative and superlative adjectives,

and proper adjectives in COCA

4.5 Example of concordance lines of certainly collocating with modal verbs 37

in COCA

4.6 Example of concordance lines of certainly collocating with main verbs 38

that express the meaning of ‘requirement’ in COCA

4.7 Example of concordance lines of clearly collocating with the top five 39

adjectives in COCA.

4.8 Example of concordance lines of clearly collocating with adjectives 39

ending with –able and –ible in COCA

4.9 Example of concordance lines of clearly collocating with unfavorable 40

adjectives in COCA

4.10 Example of concordance lines of clearly collocating with positive 41

verbs in COCA

4.11 Example of concordance lines of the top five verbs found with clearly 42

in COCA


(6)

4.12 Example of concordance lines of clearly collocating with verbs 42

that express the meaning of ‘to show’ and ‘to say’ in COCA


(7)

LIST OF FIGURES

Figures Page

2.1 Concordance lines for the word population generated by Antconc 11

2.2 Distribution of amplifiers across three periods 20

3.1 Extracted screen of the search for 1,000 concordance lines of 29

the adverb clearly when it collocates with adjectives

3.2 Extracted screen of the search for 1,000 concordance lines of 29

the adverb certainly when it collocates with verbs

3.3 Examples of the word clearly collocating with adjectives 30

in the concordance lines from COCA

3.4 Examples of the word certainly collocating with verbs 30

in the concordance lines from COCA

4.1 Percentage of the adjective collocations of the AI really 33

in three semantic prosody groups

4.2 Percentage of the verb collocations of the AI really 34


4.3 Percentage of the adjective collocations of the AI certainly 35


4.4 Percentage of the verb collocations of the AI certainly 37


4.5 Percentage of the adjective collocations of the AI clearly 38


4.6 Percentage of the verb collocations of the AI clearly 41



1

CHAPTER 1

INTRODUCTION

1.1 BACKGROUND

In English language study, it has been widely agreed that words are one of

the most important components which language learners must master in order to

comprehend not only their denotative meaning but also connotative meaning

including other necessary aspects such as their patterning and grammar in use. Two

main types of words that play important roles in English are content words and

function words. Linguistically, it is explained that a function word, sometimes called

a grammatical word or a closed class word, is a word that contains little lexical

meaning or has no meaningful meaning, but expresses a grammatical or structural

relationship with other words in a sentence (Hartmann & Stork, 1972; Quirk,

Greenbaum, Leech, & Svartvik, 1985; Au-Yeung, Howell & Pilgrim, 1998;

Hartsuiker, Bastiaanse, Postma & Wijnen, 2005). Function words include

prepositions, pronouns, determiners, conjunctions, auxiliary verbs, modals, particles

and quantifiers. In contrast to the function word, a content word or a lexical word or

an open-class word, is referred to as a word that has meaning itself and function to

carry the content of a sentence. According to Carter, McCarthy and O’Keeffe (2011),

English has four classes of content words, namely nouns, verbs, adjectives, and

adverbs.

At the word level, for the selection of words in use, English native

speakers heavily rely on their intuition. Davies (1991) and Stern (1983) claimed that

the individual has intuitive knowledge of the language, whereas L2 learners mostly

depend on dictionaries and strictly follow the grammar rules and structures, and

strategies prescribed because it is believed that these would lead to the correct use of

English.

However, using English is not only related to language in terms of

denotative meaning nor grammatical rules but also collocation, semantics and

pragmatics; supported by Sinclair (1991) expressing that denotative meaning and

grammar are not enough to allow the learners capable to reach English


2

accomplishment, the matter of collocation, semantics, and pragmatics are also the

keys. This means that correctness is important, but the naturalness in actual use of

language should also be counted and considered, in particular in communication, as it

is the requirement of coherence and communicative effectiveness (Sinclair, 1991).

The importance of naturalness in language use is also confirmed by McCarthy (1998)

stating that naturalness resides everywhere, i.e. word, pairs of words (also known as

collocational pairs), phrases, clauses, and the naturalness is a property of text and

there should be the study of the text’s nature. Likewise, it was concluded that

naturalness is not an inherent quality of language but a characteristic of the

relationship between the reader and the text (Ramsey, 1987). Thus, it could be

claimed that naturalness is essential for the effective use of language and

communication in practice. And in pedagogy, the learners should be encouraged to be

able to use the language naturally (Widdowson, 1980).

Focusing on Thai L2 learners, it has often been found that a great number

of learners produce English language based on fixed grammar rules with the

ignorance of naturalness leading to ineffective communication in both written and

spoken English, which results in inappropriate or unnatural language use; for

example, applying substitute words expressing improper connotative meaning,

placing the words in the wrong position unintentionally, and overusing or underusing

certain words which lead to misinterpretation and misunderstanding.

Regarding the natural use of English, one well-known term that has raised

the interests and awareness of most linguists and language academics in the

pedagogical field to pay attention with wide discussion and concern is semantic

prosody. In summary, semantic prosody is a term used to describe the phenomenon in

which a word whose meaning in general is neutral can be perceived with positive or

negative meaning, especially its connotative or attitudinal meaning, or associated

through frequent occurrences with particular collocations (Sinclair, 1987; Louw,

1993; Stubbs, 1996; Partington, 1998). It has been widely addressed that semantic

prosody associates the meaning with a particular purpose in a contextual

communication. So, this term is related to the natural use of language and has become

important for several decades.


3

The importance of semantic prosody is clearly stressed by Morley &

Partington (2009) who wrote that it helps maintain comprehensibility for the text

receiver and provides the receiver with an insight into the opinions and beliefs of the

text sender. They further emphasized that the awareness of semantic prosody can be

invaluable for the language learners in distinguishing among synonymous items for

the proper and natural word choice. Understanding the semantic prosody hidden in

language is important in order to maintain the connotational harmony and avoid

sending an ambiguous message (Morley & Partington, 2009). Hunston & Thompson

(2000, p.5) defined semantic prosody as “the speaker or writer’s attitude or stance

towards, viewpoint or feelings about the entities and propositions that he or she is

talking about”, in short, the “indication that something is good or bad”, in the aspect

of evaluative meaning. Apparently, there is a wide range of words that have semantic

prosody affecting their meaning in contexts. One of the well-known words is cause;

many studies show that cause actually has neutral meaning though it contains

negative prosody when it is applied in most contexts (Stubbs, 1995, 1996; Wei, 2002;

Xiao & McEnery, 2006; Limrosthip, 2013). Regarding the semantic prosody of

adverbs, the studies by Partington (2004) and Zhang (2013) confirmed that these

types of content words carry prominent semantic preference to some extent.

As naturalness is associated with semantic prosody, and this is confirmed

by many studies showing that a large number of words contains semantic prosody, the

study of semantic prosodic phenomenon occurring with adverbial intensifiers which

are also important to language production but often misused, overused or underused

by most Thai L2 learners and users, would enhance and improve the language use for

more natural and effective communication. Importantly, with more awareness on

semantic prosody, the learners’ comprehensibility would be highly developed.

In order to study the semantic prosody carried by words, a large amount

of data of authentic texts is needed to be able to explore the trend or phenomenon. For

this purpose, the data from corpora are extracted to be examined and analyzed. This is

to investigate how native speakers use intensifying adverbs in real situations and how

such adverbs commonly occur in texts.

The results of this study would enable Thai L2 learners to apply the most

appropriate words in a particular context naturally with the proper sense of


4

connotative and expressive meaning preventing misinterpretation or misunderstanding

to the receiver, and vice versa in receiving texts, in particular when communicating

with native speakers. More importantly, in pedagogical purpose, this study finding

would increase the awareness of learners to better use the adverbial intensifiers with

more comprehensibility and naturalness when necessary.

1.2 RESEARCH QUESTIONS

1.2.1 Which verbs and adjectives commonly collocate with the adverbial

intensifiers really, certainly and clearly?

1.2.2 What are the semantic prosodies of the adverbial intensifiers really,

certainly and clearly?

1.3 OBJECTIVES OF THE STUDY

The objectives of this study are the following:

1.3.1 To examine the frequent adjectives and verbs that collocate with the

adverbial intensifiers really, certainly and clearly.

1.3.2 To explore the semantic prosodies of the adverbial intensifiers

really, certainly and clearly.

1.4 DEFINITION OF TERMS

The definition of the terms of this study is as follows:

1.4.1. Adverbial Intensifier (AI) is an adverb that modifies an adjective, a

verb, or an adverb by adding stronger meaning and showing emphasis. It is a scaling

device that can intensify or weaken the meaning of words it modifies. The adverbial

intensifiers are classified into 3 main categories, namely emphasizer, amplifier and

downtoner.

1.4.2 Semantic prosody refers to the phenomenon in which certain

seemingly neutral words can be perceived with positive or negative meaning through

the habitually frequent occurrences with particular collocations.

1.4.3 Collocations refer to the words that are often used or co-occur with

another word, in the way that sounds correct to native English speaker. In corpus


5

linguistics, a collocation of a word co-occurs more often than would be expected by

chance (Kathleen, McKeown & Dragomir, 2000; Cunha & Manuela, 2009).

1.4.4 Corpus (plural, corpora) is a collection of written or spoken texts

that occur naturally in real use which is stored in computer-based readable format and

is available for qualitative and quantitative analysis (Sripichan, 2009). In this study,

the Corpus of Contemporary American English (COCA) is applied.

COCA is a freely-available corpus of English containing more than 520

million words of texts and it is equally divided into various genres, i.e. spoken,

fiction, magazine, newspaper, and academic.

1.4.5 Concordance line or key-word-in-context display is the line of texts

in which the searched items is presented in the middle. It is listed vertically to allow

reading outwards from the searched item.

1.4.6 Node word is the term used to describe the AIs that are selected to

be studied.

1.5 SCOPE OF THE STUDY

In this study, only the top three AIs, i.e. really, certainly and clearly were

investigated. These three words were selected as the target AIs because they are

intensifying adverbs and occur with the high frequency as ranked in the top 200

adverbs and top 5,000 words, in COCA (Figure 1.1). Another main reason is that they

are classified into the same category of AI emphasizer (Quirk et al. 1985).

Table 1.1

Top 10 Emphasizing Adverbs in COCA

No. Intensifying Adverb

(Emphasizer)

Top 200

Adverb

Top 5,000

Word Frequency

1 really 24 142 263414

2 actually 56 397 105155

3 simply 71 584 66712

4 certainly 74 664 59739

(Continue)


6

Top 10 Emphasizing Adverbs in COCA (Continued)

No. Intensifying Adverb

(Emphasizer)

Top 200

Adverb

Top 5,000

Word Frequency

5 clearly 87 849 45912

6 indeed 88 851 46184

7 obviously 112 1285 31299

8 definitely 164 2005 18214

9 surely 187 2434 13904

10 literally 189 2492 13425

Moreover, the node words were studied in a particular pattern where they

collocate with adjectives and verbs only. The other functions and perspectives, e.g.

grammatical patterns, discourse markers, were disregarded. In addition, the data

drawn for this study came from COCA, which represents American English.

However, the results should be sufficient to expose the sense of the use of AIs with

semantic prosody among all native speakers.

1.6 SIGNIFICANCE OF THE STUDY

In the global community in which English is used as a medium of

communication, using English in the positions of both sender and receiver with

comprehensibility is essential. Correctness in language use as well as naturalness is

necessary. In particular, among L2 learners, the awareness of natural use of language

should be raised in order that they become more aware of connotative and expressive

meaning or a word’s sense. Due to this reason, the study of semantic prosody of the

AIs really, certainly and clearly is beneficial. Furthermore, the main findings would

contribute to a more suitable pedagogical method for teachers in developing teaching

materials with regard to this concept. Additionally, the study also would be a

reference for further research concerning the use of other types of words with some

significant semantic preferences and semantic prosodies.


7

1.7 LIMITATIONS OF THE STUDY

In this study of semantic prosody of target AIs, each word is examined by

investigating its occurrences in 1,000 concordance lines: the first 500 lines of

adjective collocations and another first 500 lines of verb collocations, so this amount

of data may not be sufficient to be generalized to all uses of AIs in American English.

Moreover, since the concordance lines are drawn from COCA only, the study results

cannot be generalized to other varieties of English such as British English or

Australian English.

Furthermore, since this research study is a small scale study and was

conducted under time limitation, the researcher is unable to examine the AIs actually

and simply, being ranked in the top three in COCO, because of their variety of

meanings.

In addition, the collocations of the target AIs were studied only when they

collocate with adjectives and verbs, while the other perspectives, e.g. grammatical

patterns and discourses functions were disregarded.

1.8 ORGANIZATION OF THE STUDY

The study of the use of adverbial intensifiers with semantic prosody in

this paper is divided into five chapters. The first chapter provides background

information projecting the overall picture of the study, research questions, objectives

of the study, definitions of terms involved in this study including scope, significance

and limitations of the study. Four main parts of literature, i.e. corpus linguistics,

adverbial intensifiers and semantic prosody together with the relevant research are

reviewed in the second chapter, while the third chapter displays the methodology

explaining the target AIs, corpora, research design, selection of target AIs and method

for data collection and analysis. The fourth chapter exhibits the results of the adjective

and verb collocations as well as the semantic prosodies of the AIs really, certainly and

clearly, in figures and description. In the end, the fifth chapter contains summary of

the findings, discussion, conclusion and recommendations for further research.


8

CHAPTER 2

LITERATURE REVIEW

This chapter reviews the literature in four main areas: (1) corpus

linguistics, (2) adverbial intensifiers, (3) semantic prosody, and (4) relevant research.

2.1 CORPUS LINGUISTICS

2.1.1 Definition and Characteristics

In linguistics, a corpus (the plural form is corpora) that contains naturally

authentic texts is used as a methodology in language study and for a pedagogical

purpose (Kruger, 2002). According to O’Keeffe, McCarthy & Carter (2007, p. 1), “a

corpus is defined as a collection of texts which can be in the form of written or spoken

one, which is stored on a computer”. These electronically collected texts may vary in

size and length (Jones & Waller, 2015). Notably, a corpus is different from a general

text in terms of its purpose of use since it is created and collected to examine language

problems and to study particular perspectives of language (Hunston, 2002), It was

pointed out by Biber, Conrad & Reppen (1998) that a corpus is a principled collection

of texts which are available to be input for qualitative and quantitative analysis.

The abovementioned definitions represent three main features of the

corpus which are: a principled collection of written or spoken texts, as the corpus

must be designed in a principled way so that it represents data systematically; a

collection of electronic text stored on a computer which is usually large in size,

meaning that the large amount of texts can be stored and analyzed by using specific

software; available for qualitative and quantitative analysis, meaning that a corpus is

well planned in creation in order that it can be applied to study the language in various

aspects both qualitatively and quantitatively (O’Keeffe et al., 2007). Additionally,

another two distinctive features of a corpus are it is based on an empirical database

which contains authentic texts from various sources, so it represents the texts that

occur naturally in real use; and it is not very informative because it still needs

analysis, i.e. by human or software, for specific purposes of study (Sripicharn, 2009).


9

Regarding such definitions and characteristics, a corpus on the whole

could be summarized as “a principled collection of naturally-occurring written or

spoken texts in a computer-readable format which can be extracted and analyzed both

in qualitative and quantitative ways by using corpus analysis software” (Kennedy,

1998; McEnery & Wilson, 2001; Charles, 2002; Teubert & Cermakova, 2007).

2.1.2 Types of Language Corpora

Since the beginning of the time that corpus methodology has been

introduced (around 1960) and the first corpus named ‘Brown Corpus’ was created

(Lindquist, 2009), many types of language corpora have been established for language

study purposes.

Sripichan (2009) summarized that the categories of a corpus can be

divided into the following three major aspects. Firstly, a corpus can be categorized,

due to its communication mode, into written and spoken corpora. A written corpus

contains written texts, e.g. fiction, textbooks, news or students’ writing, and such texts

may be typed, scanned, or downloaded from websites and then entered into a

computer using corpus creation tools (O’Keeffe et al., 2007). Examples of solely

written corpora are Corpus of Global Web-based English (GloWbe), Brown Corpus,

International Corpus of Learner English (ICLE), and Longman Written American

Corpus. A spoken corpus consists of spoken language data, e.g. conversation in shops,

phone dialog, or radio interview, which are recorded, transcribed and changed into

electronic files to be stored in a computer and examples of spoken corpora are Bergen

Corpus of London Teenage Language (COLT), British Academic Spoken English

(BASE) Corpus, and Cambridge and Nottingham Corpus of Discourse in English

(CANCODE).

Secondly, corpora can be grouped in terms of data types or purpose in use

of data, into general and specialized corpora. A general corpus is often very large in

size; that is, it contains more than ten million words of a variety of language and data

so that the study and analysis from this kind of corpus may be somewhat generalized

(Bennett, 2010). Generalized corpora can be written-only corpus, spoken-only corpus

or the general corpus that contain both written and spoken language, for instance

American National Corpus (ANC), British National Corpus (BNC), Corpus of


10

Contemporary American English (COCA), and International Corpus of English (ICE)

(Lindquist, 2009). A specialized corpus is typically smaller than the general one

(Sripicharn, 2009). This corpus is composed of specific genres or registers of texts or

specialized data from some particular areas such as academic, business, arts or

science; these texts and data are collected for the purpose of representing the language

of its specialization and answering very specific questions (Bennett, 2010). Examples

of specialized corpora are Michigan Corpus of Academic Spoken English (MICASE),

which contains only spoken American academic language from the University of

Michigan; International Corpus of Learner English (ICLE), which contains both

spoken and written language of learners at many levels, and the CHILDES Corpus

(MacWhinney, 1992), which contains children’s language use (Lindquist, 2009;

Bennett, 2010). In recent years, specialized corpora, e.g. learner corpora: a corpus of

writing and/or spoken language used by students who are currently learning language;

and pedagogic corpus: a corpus of language used in classroom including textbooks,

course materials and conversational language in educational setting, have been widely

used in language learning and teaching (Bennett, 2010) from the major point of view

that they represent the authentic language in real use. In Thailand, Thai Learner

English Corpus (TLEC) has been created (Wirote, 2009). TLEC contains essays

collected from undergraduate students from Chulalongkorn University and

Thammasart University, and is categorized into two levels that are intermediate

learners and advanced learners. It also comprises Thai journalists’ writings from two

English newspapers that are the Nation and Bangkok Post. Totally, the corpus size is

approximately 1,240,000 words

(source: http://www.arts.chula.ac.th/~ling/TLE/)

Lastly, corpora can be classified in terms of number of languages

collected, into a monolingual corpus and a bilingual or multilingual corpus. A

monolingual corpus consists of texts written in one language, while a bilingual or

multilingual collects texts of two or more language, or contains sets of two or more

monolingual corpora which are designed “on the basis of similar design criteria”

(Baker, 1995, p. 232).

In addition to the three main perspectives listed above, Jones and Waller

(2015) separate corpus types into open-access corpus and make-up corpus. An open-


11

access corpus is the online corpus available on the internet so that it can be freely

accessed by users, for instance BNC, COCA or MICASSE, whereas a make-up

corpus is a collection of data which is compiled and created by using corpus analysis

software such as AntConc (a freeware corpus analysis toolkit developed by Prof.

Laurence Anthony), which can be freely downloaded from

http://www.laurenceanthony.net/.

2.1.3 Corpus Processing Techniques

According to its features, a corpus can be used to conduct a data analysis

systematically through the following basic processing techniques.

A core tool in corpus linguistics is concordancing, which means applying

corpus software to search for a particular word or phrase everywhere it occurs in a

corpus (O’Keeffe et al., 2007). The item searched or node item is presented in the

middle of concordance lines that are known as key-word-in-context displays (KWIC

concordances) which are listed and shown vertically so that they can be read outwards

to left and right, from the center or searched items (O’Keeffe et al., 2007; Sripicharn,

2009).

Figure 2.1: Concordance lines for the word population generated by

Antconc


12

Another basic function that most corpus software can perform is a

calculation of words which may be number of tokens, i.e. number of all running

words or number of types, i.e. number of words with the same type in which the

repeated words are not counted, in a corpus (Sripicharn, 2009), for example, in the

sentence ‘I know you know what I mean’, this sentence has seven tokens but five

types. The corpus software can also make wordlists or word frequency counts, which

can be in order of alphabet or frequency and be made on single lexical items or

clusters of words. In the following sample tables (Table 2.1 and Table 2.2), the most

frequent words in ten-million words from the Cambridge International Corpus (CIC)

and most frequent cluster of the word time are presented respectively:

Table 2.1

The most frequent words: ten-million-word corpus (CIC)

Word Frequency

1 the 439,723

2 And 256,879

3 To 230,431

4 a 210,178

5 Of 194,659

6 I 192,961

7 you 164,021

8 it 150,707

9 in 142,812

10 that 124,250

11 was 107,245

12 yeah 86,092

13 he 78,932

14 Is 75,687

15 on 71,797

16 for 69,392

17 but 64,561

18 she 61,406

19 they 58,021

20 have 55,892

(O’Keeffe et al., 2007, p. 34)


13

Table 2.2

Frequency of expressions with time

Expressions Frequency

all the time 1,019

the first time 834

at the time 733

a long time 657

by the time 583

at the same time 460

in time 323

the last time 238

at a time 216

a good time 127


For pedagogical purposes, this function is mainly used in identifying the

core vocabulary of English which can be an important tool for teaching English in

class. Also, Meunier (1998, p. 19) stated that “this facility is intended to support

stylistic comparison”, for instance the comparisons of various versions of translated

texts of the same story or articles on the same topic.

The other function that corpus software has is key word analysis, which is

performed by using a Keyness tool in order to compare key words between one

corpus and the other benchmark corpus (Sripicharn, 2009). This tool is helpful for

characterizing a register or a genre. For example, it is found that key words which are

found from an economic lecture relative to a general corpus of academic features are

tax, income, system, rate and supply (O’Keeffe et al., 2007). In addition to key word

analysis, the corpus technique can perform cluster analysis, that is to identify and list

the combinations of words or chunks as can be seen in Table 2.3 which shows the 20

most frequent 3-word combinations from 10-million words of CIC as follows:


14

Table 2.3

The 20 most frequent 3-word combination in ten-million-words from CIC

Word Frequency

1 I don’t know 588

2 a lot of 364

3 one of the 320

4 I don’t think 248

5 it was a 240

6 I mean I 220

7 the end of 198

8 there was a 193

9 out of the 190

10 do you think 177

11 a couple of 166

12 do you want 159

13 you have to 158

14 be able to 157

15 a bit of 155

16 you want to 153

17 and it was 148

18 it would be 142

19 do you know 138

20 you know what 137


2.1.4 Use of Corpora

Regarding the techniques that corpus software can perform, it makes a

corpus useful for various language analysis and study purposes. According to

O’Keeffe et al. (2007), corpora have been used in many areas, e.g. lexicography:

corpora have been utilized to make dictionaries such as COBUID (Collins

Birmingham University International Language Database), one of the first corpora-

based dictionaries created by John Sinclair and his team in 1980; grammar:

significant grammatical patterns and frequency can be investigated through corpus

techniques; stylistics: corpora can be applied for the study of literature and style of the

language people use; forensic linguistics: corpus techniques are sometimes used in

law and crime investigations analyzing language in documents or discourse in


15

courtroom setting; sociolinguistics: the difference of sociolinguistic variables such as

age, gender, education, or socio-economic, that affects the use of language can be

explained through corpus analysis; translation: in translation, corpora can be either

compared to translate the same or similar texts or utilized as tools in translation by

humans and machines; semantics and pragmatics: since in real use, meanings defined

by dictionaries do not represent all meanings in real use, in this case corpora are

helpful by comparing sentence to sentence or text to text to understand the meaning of

a word or phrase when it is mentioned in specific context; and discourse study:

recently, corpus concept has been employed in discourse analysis, especially political

discourse and business discourse (Jones & Waller, 2015).

As remarked by Jones & Waller (2015), corpus methodology can be used

to explore language and reveal the perspectives of language usage and patterns that

might not have been explained or were misused, in particular in terms of grammatical

analysis, as the corpora can disclose the variations of grammatical patterns in specific

contexts including distinctive grammar of spoken English and lexico-grammatical

profiles. However, the intention or thought of a writer or speaker during the language

processing is unable to be discovered by a corpus, so the data interpretation is

necessary (Jones & Waller, 2015).

In terms of the use of corpora in language study, Biber, Conrad & Reppen

(1994) identified two important advantages, indicating that, with corpora, the

language analyses are based on a large empirical data of naturally-occurring texts and

corpora allow researchers to extract specific issues or to answer some questions.

2.2 ADVERBIAL INTENSIFIERS

2.2.1 Definition and Types of Adverbial Intensifiers

English mainly has four classes of content words, namely nouns, verbs,

adjectives and adverbs. Among these, the most frequent type of content words is

nouns while adverbs are the least; however, its importance is not less (Carter,

McCarthy & O’Keeffe, 2011). Adverbs are used to add more information to a verb, an

adjective, another adverb, a clause and a whole phrase or sentence, in particular to

indicate the time, manner, place, degree and frequency of something in order to make


16

the meaning stronger as well as play roles as discourse markers and short responses

(Carter et al., 2011). It is further stated that any adverbial words or phrases are not

necessary to the structure of the clause or sentence but only establishes an extra

meaning to it (Carter et al., 2011).

One of the most common types of adverbs is called ‘adverbial

intensifiers’ or ‘intensifiers’ (Suzuki & Yamagishi, 1999). The term intensifier has

been firstly defined by Bolinger (1972, p. 17) as “any device that scales a quality,

whether up or down or somewhere between the two”. Similarly, intensifiers have been

acknowledged as adverbs or adverbial groups of words or phrases that are the subset

of adverbs of degree, as claimed by Quirk, Randolph, Greenbaum, Leech & Svartvik

(1985), who categorized intensifiers under the group of degree modifiers that express

either high or low point or intensity degree.

In general, AIs have two main functions which are to increase or to

decrease the modified items’ meaning, as described in the book English Grammar

Today. The intensifiers are identified as the adverbs or adverbial phrases that add a

stronger meaning and show emphasis (Carter et al., 2011, p. 255). Likewise, Vasko

(2010, section 9.3.2) defined intensifiers as “scaling devices which have either an

intensifying or a weakening effect on the meaning of the word they modify”.

Overall, it seems that not only should the change in a modified item’s

meaning be concerning but also its intensity scale or degree must be focused on.

Strongly confirmed by Lorenz (1998), adverbial intensifiers have a major function to

increase the intensity of the word not its meaning. Likewise, Pichler (2016)

maintained that intensifiers can have an amplifying or diminishing influence on the

modified head specific word or phrase.

According to Backlund (1973), in terms of notions of degrees, adverbial

intensifiers are classified into eight sub-classes, which are:

1. Adverbs expressing the complete or partial absence of the concepts

denoted by their head words;

2. Adverbs expressing the minimum degree or degree just above the

non-presence;

3. Adverbs expressing the low degree;

4. Adverbs expressing the low degree of a positive idea;


17

5. Adverbs expressing the moderate degree;

6. Adverbs expressing the increasing degree;

7. Adverbs expressing the high degree; and

8. Adverbs expressing the highest degree.

Furthermore, Quirk et al. (1976) clearly classified intensifiers by and

large into three semantic categories, namely emphasizer, amplifier and downtoner, as

explained below.

Emphasizers, as defined by Quirk et al. (1985), have a reinforcing effect

adding to the force of the modified words or phrases; as such, they do not require a

gradable predicate in intensifying. Examples of emphasizers are actually, certainly,

clearly, definitely, indeed, obviously, plainly really, surely, frankly, honestly, literally,

simply. Reviewed by Baumgarten, Bois & House (2012), it is summarized that

emphasizers’ function is to reinforce truth value.

Amplifiers have their function to intensify or amplify a certain quality

(Plo-Alastrue & Perez-Llantada, 2015), and are subdivided into maximixers and

boosters. Maximizers, for instance absolutely, totally, completely, entirely, fully, quite,

thoroughly, and utterly, relate to the extreme degree or occupy the highest position on

the intensify scale and most commonly modify non-gradable words (Huddleston &

Pullum, 2002). Meanwhile, boosters, for instance very, really, terribly, badly, deeply,

greatly, heartily, only indicate that it is very intense but its degree of intensity can be

more, and modify fully gradable words (Huddleston & Pullum, 2002).

For the last category, downtoners, Quirk et al. (1973, 1985) considered

that the main function of downtoners are to diminish the modified item’s intensity

degree, and divided them into four sub-classes, namely approximators, expressing an

approximation to the force of verbs, which reveals more than is actually relevant, e.g.

almost, virtually, nearly, as good as; compromisers, can be used for ether extreme

maximizing or minimizing, depending on the item they modify, e.g. sort of, quite,

rather, enough; diminishers, having a lowering effect and basically mean “to a small

extent” (Quirk et al., 1985, p. 599), e.g. only, somewhat, partly, slightly, a little; and

minimisers, scaling downwards considerably, e.g. hardly, scarcely, barely. Regarding

its functions, it was stated that speakers use downtoners in order to weaken the

assertion and to ease the unwelcome effect (Urrea, 2006). Consequently this makes


18

the terms downtoners and hedge able to be used interchangeably (Xiao & Tao, 2007).

However, Lakoff (1972) claimed that downtoners implicitly contribute to creating

fuzziness. As downtoners are very style-sensitive, it has been found that, “informal

downtoners (such as hardly, little and only) are rather frequent in informal spoken

discourse, whereas formal downtoners (such as nearly, merely and fairly) are

prevalent in academic writing” (Plo-Alastrue & Perez-Llantada, 2015). Furthermore,

in academic registers, downtoners are used to “specifying the amount of different”

(Biber et al., 1999, p. 567), and thus this type of intensifiers typically collocates with

an adjective different.

Similar to these groups classified by Quirk et al. (1985), Lorenz (1998)

categorized the intensifiers following the expression of their respective degrees,

namely maximizers, boosters, approximators, compromisers, dimininishers and

minimizers. It is explained that the maximizers and boosters work as intensifiers while

the other three groups function as downtoners (Lorenz, 1998).

2.2.2 Use of Intensifiers

Over many years, intensification in the use of natural language has been

studied in depth from many particular perspectives such as its common usage in the

real context, for example, a study found that “conversation uses a wider range of

common intensifiers than academic prose” (Biber, Johansson, Leech, Conrad &

Finegan et al., 1999). Not only its function, general application or pragmatic

viewpoints, but also the area of its semantics and usage in the social landscape have

been focused on broadly.

In terms of semantics, the recent findings (e.g. McCready & Kauffman,

2013; Bylinina, 2010; Irwin, 2013; Beltrama & Bochnak, 2013; McNabb, 2012) have

shown that the boosting impact of the intensification can occur through the various

semantic operations, and that the environments where intensification is found extend

well beyond the category of gradable expressions. For the social field factors, some

studies have revealed that the use of most intensifiers vary across the different social

categories and demographic categories, especially gender and age (Macaulay 2006,

Tagliamonte 2008, Tagliamonte & Roberts 2005). Affirmed by Beltrama (2014), the

intensification is often buried in “socially-conditioned variation” and the distribution


19

of intensifiers relates to the specific macro-social attributes of the speakers, in

particular their ages and to the specific communicative contexts. Concerning age, it

was found that intensifiers are commonly used more frequently among young

speakers, and on the contrary tend to decrease in use by older generations (Labov,

2001, Tagliamonte and D’Arcy, 2009, Kwon 2012). In a similar way, gender has also

been claimed to relate with intensifiers distribution as explored by Tagliamonte

(2008) presenting that some intensifiers such as so and pretty are predominantly used

by women in Toronto. Additionally, the education of the user affects the nature of

intensification; Ito & Tagliamonte (2003) observed the Northern English data from

York and noticed that the uneducated informants used fewer intensifiers than the

educated ones.

In addition to demographic and social categories, intensifiers have been

argued to be associated with specific genres. Biber (1988) examined the distribution

of intensifiers across different genres, i.e. news reports, academic writing, and fiction,

and found that intensifiers are most generally found in contexts where the speaker’s or

author’s intention is to express a high degree of personal involvement in the

communication. Furthermore, Xiao and Tao (2007) analyzed 33 English amplifiers,

supporting that intensifiers are used more in spoken registers than written ones.

Recently, Lim and Hong (2012) have investigated the use of intensifiers of Mandarin

Chinese, and result has been shown in the same direction, concluding that most

intensifiers are predominantly found in spoken genres.

In spite of the aforementioned variation in use of intensifiers, as same as

other types of main word classes which are nouns, verbs, adjectives and adverbs, the

adverbial intensifiers face the phenomenon of the change in use over the time (Xiao &

Tao, 2007). Xiao & Tao (2007) studied the use of amplifiers analyzing the data from

BNC covering three decades from the early 1960s to the early 1990s by dividing into

three different times; and it can be seen (as shown in Figure 1.2) that the use of

amplifiers declined over the intervening three decades.


20

Figure 2.2 Distribution of amplifiers across three periods

(Xiao & Tao, 2007, p. 257)

All in all, it could be summarized that an adverbial intensifier has the

main function to either amplify or diminish the intensity degree at some level of the

word, phrase or whole sentence it modifies. The language users may use it differently

to express their emotion and to serve a communicative purpose. Also, the use of

intensification could be different in terms of literature, dialectal variety, semantic

purpose, social attributes, social class, and educational background, for example.

2.3 SEMANTIC PROSODY

2.3.1 Background and Definition of Semantic Prosody

Cambridge Online Dictionary (COD) defines the meaning of the word

‘semantic’ as an adjective which means ‘connected with the meanings of words’ and

‘prosody’ as a noun which means ‘a. the pattern of rhythm and sound in poetry and b.

the rhythm and intonation (= the way a speaker’s voice rises and falls) of language’

and prescribes both words in the category of linguistic terms. Both words’ meanings

are similarly shown as ‘relating to the meanings of words’ and ‘the patterns of sound

and rhythm in poetry and spoken language, or the rules for arranging these patterns’

consecutively, in Longman Dictionary of Contemporary English (LDCE) under the

topics of linguistics and literature.

Among linguists, semantic and prosody have been combined and are well

recognized as one of the important terms ‘Semantic Prosody’. The term semantic


21

prosody has raised the interests and awareness of most language academics to pay

attention with broad discussion and exploration in depth for years, and is on the top of

language learning and teaching’s concern continually. The concept of semantic

prosody was originally noticed by J.R. Firth, the father of Firthian linguistics, in 1957

when he explored semantic interpretation in language utterance, and at that time it

was noted as ‘phonological prosody’. Later, in 1987, it was John Sinclair who firstly

coined the term ‘Semantic Prosody’ after he observed the phenomenon of semantic

prosody in the words’ habitual collocation. At last, semantic prosody was officially

introduced to the public by Bill Louw in 1993 by giving the first definition: “a

consistent aura of meaning with which a form is imbued by its collocates” (Louw,

1993, p. 157). His claim was that semantic prosodies are unable to easily be

recognized by human intuition or self-examination in using language and remain

hidden from the acknowledgement until John Sinclair initially studied lexis and

collocation with the machine driven lexicography from the large corpora.

Consequently, the evidence of semantic prosody can be discovered effectively using

corpora.

Sinclair (1996, p. 87) stated that “the initial choice of SP is the functional

choice which links meaning to purpose; all subsequent choices within the lexical item

relate back to the prosody”. It has been widely claimed that semantic prosody

associates the meaning with a particular purpose in a contextual communication. The

evidence was clearly shown in 1991 when Sinclair found that the lexical items

happen, break out, set in and bent on are all habitually linked to negative occurrences.

This concept is credited by Louw (2000), referring to semantic prosody as a form of

meaning developed through “the proximity of a consistent series of collocates”, which

are usually identified as positive or negative, and its function as the attitudinal

expressions of the message deliverer towards the pragmatic situation. Supported by

Stubbs (1996, p. 176), semantic prosody is referred to as “a particular collocational

phenomenon” in which two or more words are habitually co-occurring called

collocation. The classic word in the studies of semantic prosody which is cause was

initially investigated by Stubbs (1995, 1996). He observed the use of the word cause

in a corpus containing one million words, and found that almost 80% of occurrences

were negative (e.g. cause anxiety, cause errors, etc.), even a further study in another


22

corpus with 120 million words showed the same result of its negative prosody.

Furthermore, the term semantic prosody was explained in a new way by Alan

Partington in which he describes semantic prosody as an aspect of evaluative

meaning: “the spreading of connotational coloring beyond single word boundaries”

(Partington, 1998, p. 68). Partington (1998) as well supported the previously

mentioned linguists that corpus data reveals the statistical tendencies of the words or

phrases’ semantic prosodies since the language user’s intuition does not, as apparently

shown in his finding of the notion of the items timely, excessive, flabby which are said

to have a clearly favorable or unfavorable evaluation. In addition, Hunston and

Francis (2000) explained that if any word co-occurs typically with other words

holding a particular semantic set, such word may have a particular semantic prosody

correspondingly. Due to this view, semantic prosody is claimed as a phenomenon of a

node word frequently co-occurring with the lexical words or phrases under the same

semantic area (Hunston & Francis, 2000). Hunston (2002) further emphasized that the

sense of semantic prosody is able to be noticed only by the observation of “a large

number of instances of a word or phrase” because it is on the word or phrase’s use in

a typical way.

As the term semantic prosody has drawn the attention of most linguists,

its term’s concept, framework and definition have been endlessly studied and

extended. In 2005, William F. Whisitt offered a distinct concept of semantic prosody

which focuses on its pragmatic function. He also stated that “the essence of the

phenomenon of semantic prosody is, however, historical change: meaning being

transferred between terms which appear together frequently over time” (Whitsitt,

2005, p.294). Apparently, semantic prosody is closely associated with collocates or

collocation: the item’s affective or negative meaning does not show if it does not

occur along with the typical collocates in the context or it could be said that “the

semantic prosody of an item is the result of interplay between its typical collocates”

(Xiao and McEnery, 2006). This means that by only looking at the node items without

the consideration of the surrounding context and the frequent collocation is

impossible to extract the semantic prosody. This has been supported by Stubbs (2002,

p.225) observing that “there are always semantic relations between node and

collocates, and among the collocates themselves”, and Morley and Partington (2009,


23

p.150) stating that the semantic prosody was referred to as a product of collocation

with the reason that the prosody tends to change according to the collocational

patterns of items in the context.

2.3.2 Main Features and Categories of Semantic Prosody

Apart from its definition and meaning, the semantic prosody’s features

and categories have been interestingly discussed and described. According to Sinclair

(1987), the semantic prosody was recognized as connotation, pragmatic meaning and

attitudinal meaning and three salient features, namely functionality: linguistic choices

and communicative purpose were proposed, (Sinclair, 1996). Louw (1993) explained

that semantic prosodies involve either negative/ unpleasant events or neutral/ positive/

pleasant events and serve the primary function to express the attitude or evaluation of

speaker or writer (Louw, 2000). Similarly, Stubbs (1996) distinctly categorized the

semantic prosody into three groups which are positive, neutral and negative, while

Partington (1998) separated it into favorable and unfavorable evaluations. From the

study, the semantic prosody’s phenomenon could be determined and observed from

two main perspectives which are the functional perspective and the collocative

meaning perspective, and could be summarily classified as positive/ pleasant/

favorable, negative/ unpleasant/ unfavorable and neutral/ mixed prosody (Zhang,

2010). And it is suggested that the knowledge and awareness of semantic prosody

should be taught and raised among second language learners in order that they will

use English more naturally.

There have also been more and more studies on semantic prosody

insisting on the importance of semantic prosody in language such as the study of more

than 25 English lexical items which proved that the words apparently carry either

negative or positive prosody (Xiao & McEnery, 2006); the finding which confirms

that the classic word cause has a stronger negative prosody (Wei, 2002). Even in other

languages, the concept of semantic prosody has been brought to consideration as to

whether it exists or not, for example, Gan (2015)’s study which explored the

collocational behavior and semantic prosody of the words ‘mianzi’ and ‘lian’ in

modern Chinese and found that these two words are likely to have opposite semantic

prosody especially when they are used in the same sentence.


24

2.3.3 Semantic Prosody and Semantic Preference

On the other hand, there has been a dispute about the term semantic

prosody and the term semantic preference which are closely related to each other

whether they are the same or not. Sometimes, both terms are used to describe the

same phenomenon; however, it is argued that they should be considered differently at

other times. Stubbs (2002) referred to the semantic preference as the meaning

appearing from the shared semantic features of a given node word’s collocates.

Further explained, semantic prosody and semantic preference are not similar in terms

of operating scopes, semantic preference relates a node item to another item from a

particular semantic set and can be viewed as the collocates’ features; while the

semantic prosody affects broader stretches of text and can be viewed as the node

word’s feature (Partington, 2004, Xiao & McEnery, 2006). Partington (2004)

additionally noted that semantic prosody describes the phenomenon that a word or

phrase shows a preference to co-occur with surroundings that might be good or bad so

it can be explained as a sub-category of semantic preference.

2.4 RELEVANT RESEARCH

During the last two decades, the concept of semantic prosody has been

increasingly studied by many researchers and linguists (Hu, 2015). In order to analyze

the semantic prosodic behaviors in language, corpus methodology has been applied as

a main tool in observing and revealing the distinct semantic prosodies, especially in

near synonymous words (Xiao & McEnery, 2006). To date, there have been some

corpus-based studies exploring semantic prosodies in the use of words, as discussed

below.

Limrosthip (2013) investigated the semantic prosody of the words cause

and commit, collecting and exploring data from four selected language corpora, which

are British National Corpus (BNC), British News Corpus, Reuters Corpus and Corpus

of Contemporary American English (COCA). The results apparently revealed that

both cause and commit often occur in negative contexts, meaning that these words

carry negative semantic prosody. Additionally, the research showed that frequent

collocates of these words are also in negative sense, namely trouble, difficulty,

damage, cancer, injury and death. These findings supported the research of Wei


25

(2002), in that the word cause conveys a strong negative semantic prosody. In spite of

that, the further study by Wang & Wang (2005), comparing the English writing of

native speakers and Chinese learners, in the awareness of using lemma cause carrying

semantic prosodic meaning, showed that Chinese learners underuse the typical

negative semantic prosody and instead often overuse the atypical positive one, for

instance, development, progress and improvement.

Not only can words have semantic prosody but phrases also carry such

connotative meaning as proved by Begagic (2013) in the study on semantic prosody

of the collocation of a phrase make sense. The study was aimed at exploring the

behavior of make sense in COCA, focusing on its semantic prosody, and comparing

between the texts from newspapers and academic registers. It was found that make

sense often occurs in a negative environment and is in negative form, as well it is

more frequently found in a negative environment in newspaper genres than academic

one.

Recently, Hu (2015) compared the semantic preference and semantic

prosody with three synonymous adjective pairs, i.e. initial vs. preliminary, following

vs. subsequent, and sufficient vs. adequate. Such adjective pairs were selected from

3,000 core words in the Academic Vocabulary List (AVL), which is based on COCA

(Gardner and Davies), following clearly established criteria. The major findings

showed that an item may be involved with more than one group of semantic prosodies

and its particular semantic prosody could be noticed only when its collocates are

considerably scrutinized in the context and suggested that most of three synonymous

adjective pairs studied often carry neutral semantic prosody in academic registers.

However, it is suggested that more research in other types of texts such as

newspapers, magazines should be conducted to discover the differences as well.

Moreover, Xiao & McEnery (2006) conducted research to explore the

collocational behavior and semantic prosody of near synonyms, i.e. consequence

group, cause group, and price & cost group, from a cross-linguistic aspect, based on

data in English and Chinese. In their paper, it was emphasized that more than 25

lexical words in English have been shown to carry notably negative or positive

prosody. Their findings concluded that semantic prosody in Chinese can be noticed as

in English as well; consequently, it was clear that the importance of semantic prosody


26

should be explained to L2 learners to some extent. The findings were in line with

Zhang W. (2009), in that the ignorance of semantic prosody among ESL (English as a

Second Language) and EFL (English as a Foreign Language) learners are common,

and it was proposed that the concept of semantic prosody should be integrated into

ESL/ EFL vocabulary pedagogy in order that the learners’ communicative

competency will be improved.

Apart from the study of semantic prosody of verbs, adjectives and

synonyms as mentioned previously, a study in the case of adverbial intensifiers, in

terms of the change in semantic prosody, was conducted by Zhang, R. (2013). In the

study, Zhang examined four adverbial intensifiers that are terribly, awfully, horribly

and dreadfully by comparing their frequent collocational companies which may be

favorable or unfavorable words, drawn on historical corpus data: Bank of English

Corpus (BOE), and modern corpus data: Corpus of Late Modern English (CLME) as

the contrasted data, in order to study the semantic prosody change over different

period of times. The study showed that these four adverbial intensifiers, in the past,

carry the strong negative semantic prosody, especially negative emotional meaning;

however, over the centuries, their semantic prosodies have changed to be more neutral

or positive through their habitual collocation. This supports Louw (1993) who relates

semantic prosody to ‘contagion’ or ‘semantic transfer’. Another study of semantic

prosodic phenomenon in adverbial intensifiers was conducted by Partington (2004)

who examined seven amplifying intensifiers, namely absolutely, perfectly, entirely,

completely, thoroughly, totally and utterly. The finding revealed that the adverbial

intensifiers carry semantic prosody and semantic preference to some extent, and some

items have a stronger good or bad prosody than others.

This chapter reviews the related literatures in three main areas which are

corpus linguistics, adverbial intensifiers, and semantic prosody. Previous studies are

also presented and summarized, showing that words have particular sets of

collocations and carry semantic prosodies. Next, Chapter three reveals the

methodology including target adverbial intensifiers, corpora, procedures and data

analysis.


27

CHAPTER 3

METHODOLOGY

3.1 TARGET ADVERBIAL INTENSIFIERS

The study investigated the use of the three adverbial intensifiers (AIs)

really, certainly and clearly and their verb and adjective collocates in authentic texts.

For this study, 3,000 concordance lines (1,000 lines for each AI) from the Corpus of

Contemporary American English or COCA were extracted for the analysis. Those

data are the language samples produced by native speakers of American English, in a

variety of genres, i.e. spoken text, fiction, magazine, newspaper and academic text, so

it presents the authentic language use.

3.2 CORPORA

In this study, the Corpus of Contemporary American English (COCA),

as a reliable source of data, was used in the process of data collection in order to show

the concordance lines of authentic texts.

Additionally, the researcher consulted Longman Dictionary of

Contemporary English (LDOCE), online version 2016 as a main reference in

checking the collocates that co-occur with node words, their meanings and examples

of use. This dictionary contains more than one million corpus examples and the

words’ collocations. It is also claims to be a comprehensive online vocabulary and

grammar resource.

3.3 PROCEDURES

This section describes the research design and procedure of data

collection.

3.3.1 RESEARCH DESIGN

This study is an exploratory study which was conducted using a corpus-

based approach. The concordance lines of the target AIs were drawn from COCA

showing real texts used by native speakers for investigating their semantic prosody


28

and exploring their connotative meaning. For the examination of the semantic prosody

of node words, their adjective and verb collocations were collected and analyzed.

3.3.2 SELECTION OF TARGET ADVERBIAL INTENSIFIERS

In this study, only three AIs that are found in COCA were studied. Two

main criteria for selecting the target AIs were:

1. The selected words must be intensifying adverbs and be classified

into the category of adverbial intensifier emphasizer (Quirk et al., 1985).

2. The selected words must be the emphasizers occurring with high

frequencies, found in the top three ranked in COCA.

As shown in Figure 1.1, the top three emphasizers found in COCA are

really, actually, and simply. However, from the pilot study, it was discovered that the

words actually and simply, which are ranked in the top three, do not function as only

intensifying adverbs, instead, they have many varieties of use. Most of the frequencies

show their use as in the meaning of ‘as a matter of fact’ for actually, and ‘only’ for

simply, so the frequencies shown might not represent its real data as the AIs. Thus, the

researcher decided to study the next words ranked in the fourth and fifth places

instead, which are certainly and clearly.

3.3.3 DATA COLLECTION

The researcher collected the data from COCA online website at

http://corpus.byu.edu/coca/, provided free by Brigham Young University. The

researcher used its search feature ‘KWIC’ to extract the concordance lines of the node

words as designed.

1,000 concordance lines of the adverbs really, certainly and clearly were

searched for and extracted. To take each node word that functions as adverb and

collocates with adjectives, the part-of-speech tags were selected as r* and j*

respectively, as shown in Figure 3.1.


29

Figure 3.1 Extracted screen of the search for 1,000 concordance lines of

the adverb clearly when it collocates with adjectives

from http://corpus.byu.edu/coca/

Likewise, a similar method was used to collect the data of AIs’ verb

collocations. Only a part-of-speech tag was changed from adjectives, i.e. j*, to verbs,

i.e. v*, as shown in Figure 3.2.

Figure 3.2 Extracted screen of the search for 1,000 concordance lines of

the adverb certainly when it collocates with verbs

from http://corpus.byu.edu/coca/


30

After that, the program presented the concordance lines of the studied

words functioning as adverbs followed by adjectives and verbs, as shown in Figure

3.3 and 3.4, respectively.

Figure 3.3 Examples of the word clearly collocating with adjectives

in the concordance lines from COCA from http://corpus.byu.edu/coca/

Figure 3.4 Examples of the word certainly collocating with verbs

in the concordance lines from COCA from http://corpus.byu.edu/coca/

Then, the researcher scanned each line and removed the irrelevant data,

e.g. the concordance lines showing the target AIs functioning as discourse markers or

particles and mixed-up lines produced by automatic errors, to prevent

misinterpretation during investigation. Among these 1,000 lines of each node word,

the first 500 lines that met this criterion were selected for data analysis.


31

3.4 DATA ANALYSIS

After the procedure of data selection and collection, the adjective and

verb collocations that co-occur with the targeted AIs in the COCA were categorized

into 3 classifications regarding their semantic prosodies, i.e. positive, negative and

other or mixed prosody, by consulting the LDOCE to check the definition and

possible meanings of a word.

The raw data as the members of each semantic prosody group were then

counted and calculated into percentages for a clearer picture of the results. Next, all

the data were analyzed and interpreted in a descriptive way in order to respond to the

research questions. Finally, the conclusion was summarized based on the findings and

discussed.


32

CHAPTER 4

RESULTS

This chapter reveals the results of the adjective and verb collocations

of the adverbial intensifiers really, certainly and clearly, in COCA. Each AI’s

collocates were categorized into 3 semantic prosodies, i.e. positive, negative and

others. The adjective and verb collocations for which their sense of meanings present

as neither positive nor negative due to their meanings in use, or limited texts for

interpretation, were then classified into ‘others’.

4.1 THE RESULTS OF THE ADJECTIVE AND VERB COLLOCATIONS OF

THE ADVERBIAL INTENSIFIER REALLY in COCA

The results show that really occurs more often with positive adjectives,

such as great (9), cool (7), happy (7), glad (6), special (5), lucky (3), elegant (2),

brilliant (1), friendly (1), helpful (1), and smooth (1). Among those pleasant words,

the top three most frequent adjectives modified by really are good (68) followed by

important (18) and nice (15), as in Table 4.1.

Table 4.1

Example of concordance lines of really collocating with top three pleasant adjectives

in COCA.

1 So, I think that’s really good because it’s going to be a great balance…

2 -- you know, so far we feel really good about production.

3 my age ,made a good album, a really good album, last year, and I wouldn’t put it

4 They’re old, anyway, and people take really good care of them.

5 This is just a really important day where you celebrate the start of…

6 his television show, it was a really important phase in the development of country music…

7 One of us. Well, to me this is really important that the air gets clear, that we get rid of

8 And it’s just kind of a really nice luminescent effect again.

9 It’s going to be a really nice afternoon.

10 Now, it is really nice because I am at an age where I’ve done my


33

Nevertheless, really also collocates with negative adjectives, e.g. bad

(21), hard (10), difficult (5), serious (4), sorry (6), sad (4), low (2), and uncomfortable

(1), including other adjectives, e.g. early (1), local (1), similar (1), and substantive

(1), which cannot be classified into either positive or negative prosody due to

unspecified and limited context.

From the calculation into percentage, as shown in Figure 4.1, the positive

adjectives occur with really for 62.4%, while the use of really with negative and other

adjectives occur only 29.4% and 8.2% respectively.

Figure 4.1: Percentage of the adjective collocations of the AI really in

three semantic prosody groups

Through a close investigation of the individual significant collocates of

really, it was found that the adjectives typically occur with the node item relating to

emotions or state of mind such as happy (7), glad (6), sorry (6), tired (2),

embarrassed (1), satisfied (1), scared (1), and general evaluation such as good (68),

nice (15), hard (10), difficult (5), proud (5), confident (1). Significantly, size

adjectives such as big (14), small (1), huge (1) are also its common collocates.

In terms of verb collocations, the initial results do not show any

significant semantic preference. According to the limited context, most main verbs

that appear with really present neither negative nor positive sense of meaning, so they

were classified as others, as shown in Figure 4.2.

62.4%

29.4%

8.2%

PositiveNegativeOthers


34

Figure 4.2: Percentage of the verb collocations of the AI really in three

semantic prosody groups

From the data drawn, the top five most frequent verbs emphasized by

really are get (24), think (21), know (11), care (10) and look (10), respectively.

However, it can be notably seen that really typically occurs in the sentence patterns S.

+ Auxiliary V. + not + really + Main V.; and S. + really + Auxiliary V. + not + Main

V., as in Table 4.2.

Table 4.2

Example of concordance lines of really in the sentence patterns S. + Auxiliary V. +

not + really + Main V., and S. + really + Auxiliary V. + not + Main V. in COCA.

1 You know, they’re not really talking about it.

2 that you’ve now gathered that I hadn’t really heard before.

3 when she screamed “baby”. “I didn’t really think about it.”

4 Suddenly she didn’t really care about Joe Smith anymore.

5 they come back and start a fight. They are not really hating each other.

6 place, a smaller environment. We really weren’t doing anything different than we…

7 the phase of the campaign where voters really cannot recognize them, don’t know their…

8 and studying atoms in a way that really couldn’t be explained very simply.

9 Islamic fundamentalists, who really don’t owe their support to any particular

10 give us an instructive opinion, but really won’t put a period at the end of this

20.2%

20.2%59.6%

Positive

Negative

Others


35


THE ADVERBIAL INTENSIFIER CERTAINLY in COCA

The results of the adjective collocations of the AI certainly concerning

semantic prosody are similar to those of really. It shows that 66% of the adjectives

that co-occur with certainly have a pleasant sense of meaning, Figure 4.3.

Figure 4.3: Percentage of the adjective collocations of the AI certainly


The top three most frequent positive adjectives that co-occur with

certainly are true (87), possible (13), right (10) and interested/ interesting (10); and

the other sample positive words that collocate with certainly are good (9), glad (3),

correct (2), happy (2), worthwhile (2), nice (2), useful (2), attentive (1), pretty (1),

supportive (1), for example, as in Table 4.3.

Table 4.3

Example of concordance lines of certainly collocating with frequent positive

adjectives in COCA.

1 at levels of leadership, which is certainly true and a good start.

2 …, I think, that it’s certainly possible for Barack Obama to carry Virginia.

3 But you’re certainly right that an editor is an essential part of…

(Continue)

66%

20%

14%

Positive

Negative

Others


36

Example of concordance lines of certainly collocating with frequent positive

adjectives in COCA.(Continued)

4 I’m certainly interested in fashion. I like pretty clothes, just…

5 acknowledging that they are certainly interesting and worth serious discussion.

6 You know, the environment is certainly good for broadcasters in general.

7 The definition is certainly useful and has been corroborated by the…

Furthermore, with a closer investigation, it was found that certainly

frequently co-occurs with ability adjectives ending with –able and –ible, e.g. audible

(1), capable (5), plausible (4), understandable (3), questionable (2), changeable (1),

sustainable (1); comparative and superlative adjectives, e.g. better (4), higher (3),

bigger (2), easier (2), greater (2), harder (1), older (1); and proper adjectives, e.g.

African-American (1), Asian (1), Kurdish (1), Russian (1), Syrian (1), as in Table 4.4.

Table 4.4

Example of concordance lines of certainly collocating with adjectives ending with –

able and –ible, comparative and superlative adjectives, and proper adjectives in

COCA.

1 ...to COTTLE but there are

already

certainly questionable images that are out there like “USA”…

2 She looks serene and pastoral, certainly capable of mercy. I take this for a good sign.

3 It is certainly plausible that one’s personality and emotional…

4 …perhaps not irrelevant, is certainly changeable in the fashion of Watt’s narration:…

5 …to the value of the property and certainly greater than the rents for many years to come,..

6 You know, it’s certainly easier in my time than it was when Bonnie…

7 …to complain. On camera, it’s certainly hardier for Mr. Clinton to be bold than it was...

8 …as old possibly as the land itself, certainly older than the colony, which had exited…

9 There are certainly Asian gangs, Jamaican possess.

10 Certainly Russian leaders are now trying to prevent…

Regarding verb collocations, it can be seen that certainly generally co-

occurs with verbs showing other semantic prosody, Figure 4.4.


37

Figure 4.4: Percentage of the verb collocations of the AI certainly in


Remarkably, it was revealed that modal verbs, e.g. would (32), can (10),

could (9), should (3), must (2), frequently co-occur with certainly, as in Table 4.5.

Table 4.5

Example of concordance lines of certainly collocating with modal verbs in COCA.

1 “There certainly would be no harm in discussing that.

2 independent counsel, but he certainly can be swept up as part of something that…

3 …during the summer,, we certainly could have fought it longer.

4 The ship certainly should have awakened him if someone tried to…

5 …I think that AIDs patients certainly must be cared for, and doctors have an obligation

In addition, a large number of the main verbs co-occurring with certainly

express the meaning of ‘requirement’, e.g. want (19), need (5), ask (4), expect (3),

dream about (1), prefer (1), request (1), require (1), as in Table 4.6.

36.8%

20.5%

42.7%Positive

Negative

Others


38

Table 4.6

Example of concordance lines of certainly collocating with main verbs that express

the meaning of ‘requirement’ in COCA.

1 I certainly didn’t want to have one in front of an audience.

2 …security while the company certainly needs a degree of flexibility to change size,…

3 And I’ll certainly ask for their opinion as to where they…

4 We would certainly expect an accused stalker sample to show…

5 Research certainly requires funding, but just as important, the…


THE ADVERBIAL INTENSIFIER CLEARLY in COCA

In agreement with really and certainly, the results show that AI clearly as

well co-occurs with positive adjectives rather than negative or other items, as shown

in Figure 4.5, presenting the proportion in percentage.

Figure 4.5: Percentage of the adjective collocations of the word clearly


The sample positive adjectives found with this AI are better (6),

important (6), greater (4), good (3), logical (1), powerful (1). The data reveals the top

49.7%

34.0%

16.3%

Positive

Negative

Others


39

five adjectives that are visible (38), evident (20), articulated (17), aware (12), and

audible (7) and identifiable (7), as in Table 4.7.

Table 4.7

Example of concordance lines of clearly collocating with the top five adjectives in

COCA.

1 A dozen ships were clearly visible from where she stood, and more…

2 This attitude is clearly evident in classical Greece.

3 …requires the development of clearly articulated goals and objectives.

4 The court’s majority was clearly aware of the perils of entering what one…

5 …a situation in which unusual clearly audible background utterances frequently go…

6 ...temples comprise communities clearly identifiable as autonomous civil association,…

Similar to certainly, it was revealed that ability adjectives ending with –

able and –ible, e.g. visible (38), audible (7), identifiable (7), discernible (6),

distinguishable (4), observable (3), applicable (2), definable (2), questionable (2),

recognizable (2), commonly occur with clearly, as in Table 4.8.

Table 4.8

Example of concordance lines of clearly collocating with adjectives ending with –able

and –ible in COCA

1 …mass, together with several clearly visible abdominal masses.

2 …before that when he barks, it is clearly audible at her house, even with her music on.

3 Strong remnants of both are clearly identifiable even today.

4 Her spine was clearly discernible as she bent, a string of walnut-sized…

5 …groups of mounds that are clearly distinguishable from other groups.

6 …and economic power is most clearly observable in ethnically heterogeneous countries…

7 …and thus become much more clearly applicable and engaged with the world of the…

8 …ways rather than falling along clearly definable boundaries.

9 …, in questionable circumstances, clearly questionable circumstances.

10 …automatically attributed to any clearly recognizable figure in a picture.


40

However, there were a large number of unfavorable adjectives found.

Most of them are the words relating to unpleasant emotions or unfavorable states of

mind of people such as embarrassed (4), tired (3) annoyed (2), disappointed (2),

angry (1), confused (1), depressed (1), distressed (1), nervous (1), and worried (1);

and the adjectives with negative prefixes (dis- il-, im-, in-, un-), for example

uncomfortable (6), unhappy (4), illegal (4), inappropriate (2), unworkable (2),

unsuited (1), dissatisfied (1), impatient (1), impractical (1), incapable (1), inevitable

(1), unfamiliar (1), unfair (1), unstable (1), as in Table 4.9.

Table 4.9

Example of concordance lines of clearly collocating with unfavorable adjectives in

COCA

1 …students organizers were clearly embarrassed by the small turnout.

2 Melissa was clearly tired from work.

3 Eisner was clearly annoyed at his pushing, and at the increasing…

4 The mother was clearly disappointed at this outcome since she had wanted…

5 And he would be – and he is clearly angry about it.

6 Jimmy is clearly uncomfortable at the sight of these two.

7 But he’s clearly unhappy about the environment.

8 …since Watergate, that was clearly illegal and unconstitutional and yet the press...

9 …wrote, because a lot of it was clearly inappropriate and would be embarrassing if…

10 …is exceedingly unpopular, and clearly unworkable and remains unfair.

From the exploration of verb collocations, the results of clearly are

different from those of really and certainly, exhibiting that clearly more often co-

occurs with the verbs showing positive prosody, as the percentage result shown in

Figure 4.6.


41

Figure 4.6: Percentage of the verb collocations of the AI clearly in


Sample positive verbs include understand (5), enjoy (4), establish (4),

remember (4), love (3), produce (3), recognize (3), benefit (2), admire (1), care (1),

improve (1), like (1), promote (1), as in Table 4.10.

Table 4.10

Example of concordance lines of clearly collocating with positive verbs in COCA

1 She clearly understands far more than she can respond to.

2 McCain clearly enjoyed being a prime candidate, and he has

3 … nature of a work of art, clearly establishes an important bond between the life…

4 …Achille Lauro in 1985, still clearly remembered by the industry.

5 Playing directly to the camera, he clearly loves being in the spotlight.

6 While student evaluations of faculty clearly produce a large quantity of data…

7 College GPA is clearly recognized as being the most important factor...

8 Obama has clearly benefited from being black through his…

9 ..tomorrow in Philadelphia, and he clearly admires a career like Springsteen’s.

10 China clearly cares about the decision.

It was revealed that top five verbs found with clearly are define (19), see

(10), show (9), mark (7) and reflect (7), as in Table 4.11.

43.4%

26.8%

29.8%Positive

Negative

Others


42

Table 4.11

Example of concordance lines of the top five verbs found with clearly in COCA

1 …depend on a group’s ability to clearly define itself as an entity, an in-group, and to…

2 The postauricular nerve can be clearly seen distal to the chorda tympani nerves, …

3 The reenactment, OK, it clearly shows a Georgia man used the phone in the…

4 The exits are clearly marked for you to walk of, as I stand here without

5 This high technology orientation is clearly reflected in American international trade data.

Additionally, by further examining the other collocations of clearly, it has

been revealed that clearly is often used with verbs that express the meaning of ‘to

show’ and ‘to say’, e.g. show (9), demonstrate (6), indicate (5), say (4), spell out (4),

express (3), illustrate (2), state (2), enunciate (1), expose (1), describe (1), display (1),

present (1), suggest (1), as in Table 4.12.

Table 4.12

Example of concordance lines of clearly collocating with verbs that express the

meaning of ‘to show’ and ‘to say’ in COCA

1 They clearly show a man that knows what he was doing…

2 …final three baseline data points clearly demonstrate a downward trend.

3 …, the results of the post-survey clearly indicated a high degree of satisfaction and…

4 Sandusky is clearly saying he did not abuse anyone.

5 Because the law clearly spells out the grounds for expulsion and sets…

6 Noa has so much character, clearly expressing her likes and dislikes.

7 …loan guarantees for Israel clearly illustrated congressional reluctance to work against..

8 ...in case of Global Grind, they clearly stated hip hop was their jumping-off point.

9 …she poses in a man’s suit which clearly exposes her breasts.

10 …three lots that clearly describe Boulions fashionable ornamental…

From the data drawn from COCA, the results of the adjective and verb

collocations of AIs really, certainly and clearly were revealed and categorized into 3

semantic prosodies showing the trend of their senses of meaning. Moreover, some

distinctive characteristics of the AIs’ collocates were also disclosed. Next, in Chapter


43

5, the findings will be summarized and discussed. Then, the conclusion will be

presented, followed by recommendations for further research.


44

CHAPTER 5

CONCLUSION, DISCUSSION AND RECOMMENDATION

This chapter concludes this study. A summary of the research is

presented, and findings of the study are discussed and interpreted. The conclusion is

also drawn within the scope of this study. Finally, recommendations for further

research end the chapter.

5.1 SUMMARY OF THE FINDINGS AND DISCUSSION

This research studied the three adverbial intensifiers really, certainly and

clearly in terms of their semantic prosodies by examining the frequent adjectives and

verbs that collocate with them.

As shown in Chapter 4, the results reveal the frequent adjective and verb

collocations that co-occur with the AIs as follows:

Firstly, the study reveals the top three most frequent adjectives co-

occurring with the AI really are good, important and nice. It is also found that most of

the adjectives collocating with really are related to favorable states of mind and

pleasant evaluation, e.g. confident, easy, glad, good, happy, nice, proud, and satisfied,

as well as size adjectives such as big, huge and large. Considering its verb

collocations, it is shown that top five most frequent verbs emphasized by really are

get, think, know, care and look, respectively.

Secondly, the results show that certainly often co-occurs with positive

adjectives such as best, cool, famous, glad, good, happy, helpful, nice, and useful. Its

top frequent positive adjective collocations are true, possible, right and interested/

interesting. Moreover, it was discovered that certainly frequently co-occurs with

adjectives ending with –able and –ible, e.g. audible, capable, plausible, responsible,

and understandable; comparative adjectives, e.g. bigger, better, easier, greater and

higher; and proper adjectives, e.g. Asian, Kurdish, Russian and Syrian. In terms of its

verb collocations, the results show that certainly overwhelmingly collocates with

modal verbs, e.g. can, could, may, must, should, and would. Additionally, there are


45

plenty of verbs expressing the meaning of ‘requirement’, e.g. ask, expect, need,

request, require, and want found with certainly.

Thirdly, the study reveals that the top five adjectives collocating with the

AI clearly are visible, evident, articulated, aware and audible, respectively. Similar to

certainly, the adjectives ending with –able and –ible, e.g. applicable, audible,

definable, discernible, distinguishable, identifiable, recognizable, and visible,

commonly occur with clearly. However, many unfavorable adjectives relating to

unpleasant emotions or unfavorable states of mind such as angry, annoyed, confused,

depressed, disappointed, distressed, nervous, tired and worried, and the adjectives

containing negative prefixes, e.g. dissatisfied, illegal, impatient, impractical,

inappropriate, incapable, uncomfortable, unfamiliar, and unhappy, are also found.

From the investigation of verb collocations, the top five verbs found with clearly are

define, see, show, mark and reflect. The results show that clearly commonly co-occurs

with positive verbs such as admire, benefit, care, enjoy, establish, like, love, produce,

and promote. Remarkably, it reveals that clearly often co-occurs with verbs

expressing the meaning of ‘to show’ and ‘to say’, e.g. enunciate, express, define,

demonstrate, describe, display, present, say, show, state, and suggest.

Since this research aims at examining the frequent collocates of really,

certainly and clearly when they function as adverbial intensifiers, their adjective and

verbs collocations are mainly focused on. As summarized, the results show that AIs

have their habitual or common collocates, for example certainly overwhelmingly was

followed by the adjective true, while really tends to co-occur with the adjective good

very often, which supports the studies by Zhang (2013) showing that the adverbial

intensifiers terribly, awfully, horribly and dreadfully often intensify particular words,

especially unpleasant and unfavorable words. Likewise, the study of maximizers

absolutely, perfectly, entirely, completely, thoroughly, totally and utterly reveals a

similar result which displayed the common words’ collocations (Partington, 2004).

Additionally, according to the findings from Chapter 4, it could be

summarized that the studied AIs really, certainly and clearly are typically used to

intensify or emphasize good things since they tend to co-occur with positive

adjectives rather than negative or other ones, which can be interpreted as that they

possess positive senses of meaning. However, in terms of verb collocations


46

investigation, the results of really and certainly do not expose any particular semantic

preference except clearly as it is more commonly collocated with positive verbs such

as admire, care, enjoy, love, therefore seemingly being more positive in semantic

coloring. Therefore, the AIs really, certainly and clearly contain positive rather than

negative semantic prosody.

With regard to the literature reviews in this study, semantic prosody is the

term used to describe the phenomenon concerning a word whose meaning in general

is neutral but which can be perceived with positive or negative meaning or association

through frequent occurrences with some particular collocations (Louw, 1993;

Partington, 1998; Sinclair, 1987; Stubb, 1996). It has been widely agreed by many

linguists, e.g. Hunston & Francis (2000), Xiao and McEnery (2006), that a large

number of words carry this sense of meaning. The present study’s findings about

semantic prosody of selected AIs are confirmed by the study of the words set in,

happen, break out and bent by Sinclair (1991), revealing that they are all habitually

linked to negative occurrences. The present study also bears out Brown (1995) in that

some words seem to co-occur with words having negative meanings, e.g. cause

anxiety, cause errors.

Likewise, it can be seen from this study that the adverbial intensifiers also

carry semantic prosody exhibiting either a positive/ pleasant/ favorable or negative/

unpleasant/ unfavorable sense of meaning in context. The findings show that each

investigated AI has a tendency to co-occur with specific words so that a particular

semantic preference is perceived. The outcomes are in accordance with the study of

Partington (2004), who conducted research on the semantic prosody of the

maximizing adverbs absolutely, perfectly, entirely, completely, thoroughly, totally and

utterly, and confirmed that the adverbial intensifiers have particular semantic prosody

and semantic preference to some extent and some adverbs, for instance, really,

certainly and clearly, as shown in this study, have a stronger good prosody than others

in comparison.


47

5.2 CONCLUSION

This study investigated the semantic prosody of three adverbial

intensifiers by examining their adjective and verb collocations within the data drawn

from COCA.

The key findings reveal that the three AIs in this study carry a stronger

positive semantic prosody rather than negative prosody since they are used to modify

and emphasize the good words relating to states of mind, general evaluation, and

description. The interesting findings exhibit sets of particular words occurring with

specific adverbs, namely adjectives relating to emotion and evaluation, e.g. difficult,

embarrassed, glad, good, hard, happy, and sorry, and size adjectives, e.g. big, huge,

and small, frequently occur with really, comparative and superlative, e.g. better,

easier, harder, older, and proper adjectives, e.g. Asian, Russian, Syrian, commonly

occur with certainly; ability adjectives ending with –able or –ible, e.g. audible,

capable, definable, possible, and visible, frequently collocated with both certainly and

clearly, and adjective relating to unfavorable states of mind, e.g. confused, depressed,

nervous, and worried, adjectives added with negative prefixes, e.g. dissatisfied,

impractical, inappropriate, and unhappy, are frequently found with clearly.

Regarding verb collocations, the initial investigation of verbs collocating

with the studied AIs really and certainly does not show any semantic preference.

However, it was revealed that clearly has an obvious positive prosody since it

frequently co-occur with verbs containing a positive sense of meaning, e.g. admire,

benefit, enjoy, love, and produce. It is also found that clearly typically occurs with

verbs that express the meaning ‘to show’ and ‘to say’, e.g. express, demonstrate,

describe, present, and state. In addition, the results indicate that really, certainly and

clearly typically co-occur with ‘verbs to be’ and ‘modal verbs’ as they

overwhelmingly appear in the COCA.

5.3 RECOMMENDATIONS FOR FURTHER RESEARCH

In this study, the researcher only focuses on the top three emphasizing

adverbs functioning as AIs found in COCA that are really, certainly and clearly, and

also studied only adjective and verb collocations. Further studies may examine other

AIs and explore other collocational aspects to discover additional semantic prosodies.


48

Also, different corpora may be applied to study other varieties of English, such as

British English or New Zealand English.

Additionally, the corpus-based approach applied in this study can be

expanded to deeper exploration of the use of AIs in other perspectives such as

distributional placements, grammatical patterns, formal or informal use of them by

comparing the use from genre to genre. This approach can also be utilized to explore

Thai learners’ use of AIs, mainly with semantic prosody.


49

REFERENCES

Ali, S.A., & Muhammed, L.N. (2008). The translation of downtoners in the literary

texts into arabic. Adab Al-Rafidayn, 63, 33-60.

Au-Yeung, J., Howell, P., & Pilgrim, L. (1998). Phonological words and stuttering on

function words. Journal of Speech, Language, and Hearing Research, 41,

1019-1030.

Backlund, ULF. (1973). The collocation of adverbs of degree in English. Uppsala

University: Acta Universitatis Upsaliensis.

Baumgarten, N., Bois, I. D., & House, J. (2012). Subjectivity in language and in

discourse. Bingley: Emerald Group Publishing Limited.

Begagic, M. (2013). Semantic preference and semantic prosody of the collocation

make sense. Jezikoslovlje, 14(2), 403-416.

Beltrama, A. (2014). This is totally interesting, intensifiers between formal semantics

and social indexicality. Michicagoan, Linguistics Department.

Beltrama, A., & Bochnak, M.R. (2013). To Appear: Intensification without degrees

cross-linguistically. Natural Language and Linguistic Theory.

Bennett, R. G. (2010). Using corpora in the language learning classroom: Corpus

linguistics for teachers. Michigan: University of Michigan Press.

Biber, D. (1988). Linguistic features: Algorithms and functions in variation across

speech and writing. Cambridge: Cambridge University Press.

Biber, D., Conrad, S., & Reppen, R. (1994). Corpus-based approaches in applied

linguistics. Applied Linguistics, 15, 169-189.

Biber, D., Conrad, S., & Reppen, R. (1998). Corpus linguistics: Investigating

language structure and use. Cambridge: Cambridge University Press.

Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman

grammar of spoken and written English. London: Longman.

Bolinger, D. (1972). Degree words. The Hague: Mouton.

Carter, R., McCarthy, M., Mark, G., & O’Keeffe, A. (2011). English grammar today:

An a-z of spoken and written grammar. Cambridge: Cambridge University

Press.


50

Charles, F.M. (2002). English corpus linguistics: An introduction. Cambridge:

Cambridge University Press.

Cunha, C., & Manuela, M. (2009). Handbook of research on social dimensions of

semantic technologies and web services. Pennsylvania: IGI Global.

Davies, A. (1991). The native speaker in applied linguistics. Edinburgh: Edinburgh

University Press.

Firth, J.R. (1957). A synopsis of linguistic theory: Prosody. in F.R. Palmer (Ed.):

selected papers of J.R. Firth (pp. 1952-1959). London: Longman.

Gan, Y. (2015). Analysis on semantic prosody of ‘mianzi’ and ‘lian’: A corpus-based

study. In Lu, Q. Gao, H. (Ed.): Chinese lexical semantics. Lecture Notes in

Computer Science, 9332. Spinger, Cham.

Harsuiker, R.J., Bastiaanse, R., Postma, A., & Wijnen, F. (2005). Phonological

encoding and monitoring in normal and pathological speech. Hove:

Psychology Press.

Hartmann, RRK., & Stork, FC. (1972). Dictionary of language and linguistics.

London: Applied Science.

Hu, M. (2015). A semantic prosody analysis of three adjective synonymous pairs in

COCA. Journal of Language and Linguistic Studies, 11(2), 117-131.

Huddleston, R., & Pullum, G. K. (2002). The Cambridge grammar of the English

language. Cambridge: Cambridge University Press.

Hunston, S., & Francis, G. (2000). Pattern grammar: A corpus-driven approach to

the lexical grammar of English. Amsterdam: John Benjamins Publishing.

Hunston, S. (2002). Corpora in applied linguistics. Cambridge: Cambridge University

Press.

Irwin, P. (2014). So (Totally) Speaker-oriented: An analysis of “Drama SO”. In

Zanuttini, R. and Horn, L. R. (Ed.): Micro-syntactic variation in north

American English. Oxford: Oxford University Press

Jones, C., & Waller, D. (2015). Corpus linguistics for grammar: A guide for research.

London: Routledge.

Kennedy, G. (1998). An introduction to corpus linguistics. London: Longman.


51

Kruger, A. (2002). Corpus-based translation research: Its development and

implications for general, literary and bible translation. Acta Theologica,

22(1), 70-106.

Kwon, S. (2012). Beyond the adolescent peak of Toykey. In 48th Meeting of the

Chicago Linguistic Society.

Labov, W. (2001). Principles of linguistic change. New-Jersey: Wiley-Blackwell.

Lakoff, R. (1973). Language and woman’s place. Language in Society, 2(1), 45-80.

Lim, N.E., & Hong, H. (2012). Intensifiers as stance markers: A corpus study on

genre variations in Mandarin Chinese. Chinese Language and Discourse,

3(2), 129 – 166.

Limrosthip, K. (2013). A corpus-based study of semantic prosody of the words cause

and commit. Thammasat University. Master’s Project, M.A. (English),

Thailand.

Lindquist, H. (2009). Corpus linguistics and the description of English. Edinburgh:

University Press

Liu, D., & Espino, M. (2012). Actually, genuinely, really, and truly: A corpus-based

behavioral profile study of near-synonymous adverbs. International Journal

of Corpus Linguistics, 17(2), 198-228.

Liu, G. (2014). Investigating Chinese college learners’ use of frequency adverbs: A

corpus-based approach. Journal of Language Teaching and Research, 5(4),

837-843.

Longman dictionary of contemporary English. (2016). Retrieved from

http://www.ldoceonline.com/

Lorenz, G. (1998). Overstatement in advanced learners’ writing: Stylistic aspects of

adjective intensification. In Granger, S. (Ed.): Learner English on computer.

London: Longman.

Louw, B. (1993). ‘Irony in the text or insincerity in the writer? The diagnostic

potential of semantic prosodies’. In Baker, M. and Tognini-Boneli, E. (Eds.)

Text and technology: in honor of John Sinclair. Amsterdam: John Benjamins,

157-176.


52

Louw, B. (2000). ‘Contextual prosodic theory: bringing semantic prosodies to life’.

In C. Heffer, H. Sauntson, and G. Fox (Eds.): Words in context: A tribute to

John Sinclair on his retirement. Birmingham: University of Birmingham.

Louw, H. (2005). Really, too, very, much: adverbial intensifiers in black south

African English. North-West University, Potchefstroom Campus, South

Africa.

Macaulay, R. (2006). Pure grammaticalization: The development of a teenage

intensifier. Language Variation and Change, 18(3), 267–283.

MacWhinney, B. (1992). Transfer and competition in second language learning. In R.

Harris (Ed.): Cognitive processing in bilinguals. Amsterdam: Elsevier.

McCready, E., & Kauffman, M. (2013). Maximum intensify. Paper presented at the

Semantics Group, Keio University, 29 November.

McEnery, T., & Wilson, A. (2001). Corpus linguistics: An introduction. Edinburgh:

Edinburgh University Press.

McKeown, K.R., & Radev, D.R. (2000). Collocations. In Dale, R. Moisl, H. Somers,

H. (Ed.): Handbook of natural language procession. New York: Marcel

Dekker.

McCarthy, M. (1998). Spoken language and applied linguistics. Cambridge:

Cambridge University Press.

McNabb, Y. (2012). Cross-categorial modification of properties in Hebrew and

English. In Chereches, A. (Ed.): Proceedings of semantics and linguistic

Theory, 22, 365-382.

Morley, J.,& Partington, A. (2009). A few frequently asked questions about Semantic

– or evaluative prosody*. International Journal of Corpus Linguistics, 14(2),

139-158.

O’Keeffe, A., McCarthy, M., & Carther, R. (2007). From corpus to classroom:

Language use and language teaching. Cambridge: Cambridge University

Press.

Partington, A. (1998). Patterns and meanings: using corpora for English language

research and teaching. Amsterdam: John Benjamins.


53

Partington, A. (2004). Utterly content in each other’s company: Semantic prosody and

semantic preference. International Journal of Corpus Linguistics, 9(1), 131-

156.

Phoocharoensil, S. (2010). A corpus-based study of English synonyms. International

Journal of Arts and Sciences, 3(10), 227-245.

Pichler, H. (2016). Discourse-pragmatic variation and change in English: New

methods and insights. Cambridge: Cambridge University Press.

Plo-Alastrue, R., & Perez-Llantada, C. (2015). English as a scientific and research

language, debates and discourses: English in Europe. Boston: De Gruyter

Mouton.

Quirk, R., & Greenbaum, S. (1973). A university grammar of English. London:

Longman.

Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1976). A grammar of

contemporary English. London: Longman.

Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive

grammar of the English language. London: Longman.

Ramsey, S.R. (1987). The languages of China. Princeton: Princeton University Press.

Sinclair, J.M. (1987). Looking up. London: Collins.

Sinclair, J.M. (1991). Corpus, concordance, collocation. Oxford: Oxford University

Press.

Sripicharn, P. (2009). A Study of collocation and language pattern of high-frequency

nouns, verbs, adjectives and adverbs in a corpus of English compository

writing by Thai students. Thammasat University, Thailand.

Suzuki, H., & Yamagishi, M. (1999). Licensing some kinds of degree adverbs in

English. Report of the Special Project for the Typological Investigation of

Languages and Cultures of the East and West, 735-754.

Stern, H.H. (1983). Fundamental concepts of language teaching. Oxford: Oxford

University Press.

Stubbs, M. (1995). Collocations and semantic profiles: On the cause of the trouble

with quantitative methods. Function of Language, 2(1), 1-33.

Stubbs, M. (1996). Text and corpus analysis: computer assisted studies of language

and culture. Oxford: Blackwell.


54

Tagliamonte, S. R. (2005). So weird; so cool; so innovative: The use of intensifiers in

the television friends. American Speech, 80, 280–300.

Tagliamonte, S. (2008). So different and pretty cool!, recycling intensifiers in

Toronto, Canada. English Language and Linguistics, 12(02), 361–394.

Tagliamonte, S., & D’Arcy, A. (2009). Peaks beyond phonology: Adolescence,

incrementation, and language change. Language, 85, 58–108.

Teubert, W., & Cermakova, A. (2007). Corpus linguistics: A short introduction.

London: Continuum.

Uba, S.Y. (2015). A Corpus-based behavioral profile study of near-snynonyms:

important, essential, vital, necessary and crucial. International Journal of

English Language and Linguistics Research, 3(5), 9-17.

Vasko, A. L. (2010). Cambridgeshire dialect grammar: Studies in variation, contacts

and change in English 4. Helsinki: Research Unit for Variation, Contact and

Change in English.

http://www.helsinki.fi/varieng/series/volumes/04/chapter9.html

Wang, H., & Wang, T. (2005). A contrastive study on the semantic prosody of cause.

Modern Foreign Language, 28(3), 297-307.

Wei, N. (2002). Research methods in the studies of semantic prosody. Foreign

Language Teaching and Research, 34(4), 300-307.

Wei, N.X. (2006). A corpus-based contrastive study of semantic prosodies in learner

English. Foreign Language Research, 132, 50-54.

Whitsitt, S. (2005). A critique of the concept of semantic prosody. International

Journal of Corpus Linguistics, 10, 283-305.

Widdowson, H.G. (1980). Models and fictions. Applied Linguistics, 1(2), 165-170.

Xiao, R., & McEnery, A. (2006). Collocation, semantic prosody, a near synonymy: A

cross-linguistic perspective. Applied Linguistics, 27(1), 103-109.

Xiao, R., & Tao, H.Y. (2007). A corpus-based sociolinguistic study of amplifiers in

British English. Sociolinguistic Studies, 1(2), 241–273.

Zhang, C. (2010). An overview of corpus-based studies of semantic prosody. Asian

Social Science. 6(6), 190-194.

Zhang, R. (2013). A corpus-based study of semantic prosody change: The case of the

adverbial intensifier. Concentric: Studies in Linguistics, 39(2), 61-82.


55

Zhang, W. (2009). Semantic prosody and ESL/EFL vocabulary pedagogy. TESL

Canada Journal, 26(2), 1-12. Beijing: Springer International Publishing.


56

BIBIOGRAPHY

Name Miss Siriporn Gampaenggaew

Date of Birth April 14, 1988

Educational Attainment

2009: Bachelor of Technology

(First Class Honors)

Work Position Senior Project Assistant

Project of Japan International Cooperation

Agency

Work Experiences (January 2010 - Present): Senior Project Assistant

Project of Japan International Cooperation,

Ministry of Public Health of Thailand

A corpus-based study of the use of adverbial intensifiers ...

Documents