Professional Documents
Culture Documents
Faculty of Arts
Department of English
and American Studies
Jan Štěrba
Word-formation Processes in TV
Series
Bachelor’s Diploma Thesis
2015
I declare that I have worked on this thesis independently,
using only the primary and secondary sources listed in the bibliography.
……………………………………………..
Jan Štěrba
I would like to thank my supervisor doc, PhDr. Naděžda Kudrnáčová, Csc., for her time,
kindness and willingness to help.
Table of Contents
Introduction ....................................................................................................................... 6
1. Word-formation ............................................................................................................ 9
1.1 Definitions of Basic Terms ..................................................................................... 9
1.1.1 Lexeme, word-form, derivation and inflection................................................. 9
1.1.2 Root, stem and base ........................................................................................ 10
1.2 Definition of Word-formation ............................................................................... 11
1.3 Productivity and Lexicalization ............................................................................ 12
2. Word-formation Processes in English ........................................................................ 13
2.1 Compounding ........................................................................................................ 13
2.1.1 Definition ........................................................................................................ 14
2.1.2 Compounds and phrases ................................................................................. 14
2.1.3 Distinguishing compounds from phrases ....................................................... 15
2.1.3.1 Stress ........................................................................................................ 16
2.1.3.2 Spelling and lexicalization ....................................................................... 16
2.1.3.3 Inflectibility ............................................................................................. 17
2.1.3.4 Inseparability and modification ............................................................... 18
2.1.3.5 Semantic criteria ...................................................................................... 18
2.1.4 Types of compounds....................................................................................... 19
2.1.4.1 Primary and synthetic compounds ........................................................... 20
2.1.4.2 Syntactic and lexical compounds ............................................................. 20
2.1.4.3 Endocentric, exocentric, subordinate and coordinate compounds ........... 21
2.1.5. Neo-classical compounds. ............................................................................. 22
2.2 Affixation .............................................................................................................. 24
2.3 Conversion ............................................................................................................ 25
2.4 Clipping ................................................................................................................. 26
2.5 Blending ................................................................................................................ 26
2.6 Back-formation ..................................................................................................... 27
3. Analysis of the Corpus ................................................................................................ 28
3.1 Borderline Cases ................................................................................................... 30
3.2 General Overview ................................................................................................. 31
3.3 Compound Nouns.................................................................................................. 32
3.3.1 Noun + noun ................................................................................................... 33
3.3.2 Verb + noun .................................................................................................... 37
3.3.3 Adjective + noun ............................................................................................ 38
3.3.4 Particle + noun ................................................................................................ 40
3.3.5 Verb + particle ................................................................................................ 41
3.4 Compound Adjectives ........................................................................................... 42
3.4.1 Noun + adjective............................................................................................. 43
3.4.2 Adjective + adjective ...................................................................................... 44
3.4.3 Particle + adjective ......................................................................................... 45
3.4.4 Adverb + adjective ......................................................................................... 46
3.4.5 Noun + noun. .................................................................................................. 46
3.4.6 Adjective + noun ............................................................................................ 48
3.4.7 Particle + noun ................................................................................................ 48
3.4.8 Verb + verb ..................................................................................................... 49
3.4.9 Verb + particle ................................................................................................ 49
3.5 Compound verbs ................................................................................................... 50
3.5.1 Particle + verb................................................................................................. 51
3.5.2 Noun + verb .................................................................................................... 51
3.5.3 Particle + noun ................................................................................................ 51
3.6 Compound Adverbs .............................................................................................. 52
3.6.1. Particle + noun ............................................................................................... 52
3.6.2 Noun + adjective............................................................................................. 53
3.7 Phrasal Compounds ............................................................................................... 53
3.7.1 Phrasal nouns .................................................................................................. 54
3.7.2 Phrasal adjectives ........................................................................................... 54
3.8 Neo-classical Compounds ..................................................................................... 55
3.9 Comparing Friends and The Big Bang Theory ..................................................... 56
3.9.1 General comparison of the series ................................................................... 56
3.9.2 Comparison of characters ............................................................................... 58
3.10 Summary ............................................................................................................. 60
Conclusion ...................................................................................................................... 66
Bibliography ................................................................................................................... 69
Resumé in English .......................................................................................................... 74
Resumé in Czech ............................................................................................................ 75
Introduction
(Bauer, 1983, p. 206). Therefore, it proves particularly interesting to study the English
word-formation processes, which is also the topic of this thesis. This thesis will focus
regarded as the most productive English word-formation process (Kavka and Štekauer,
2006, p.52) and is thus the most difficult to delimit. The main aim of this thesis is to
The classification was done using contemporary British and American TV series
where studies regarding compound words are rather rare. The choice of series relevant
for this thesis was based on nature of the series and occurrence of compounding. Scope
of the thesis was also taken into account. This thesis will cover Friends and The Big
Bang Theory. These TV series are well-known and successful all over the world and
their focus also seems to be slightly different and so they allow comparison and
contrast. After going through transcripts of various episodes of these TV series, the first
ten episodes of the first season were analyzed for each series. Corpus was created using
This thesis mainly relies on Bauer’s book called English Word-formation (1983)
and his classification of compound words was adopted. Most of the terminology used in
the text also corresponds with Bauer’s approach to the theory of word-formation.
Differences from this terminology will be remarked on in the text. Another important
English Lexicology written by Jackson and Amvela (2007). However, this thesis also
6
The focus of this work is mainly practical, but sufficient theoretical background
is also needed. For this purpose, the first two chapters will have a largely theoretical
basis. The first chapter will be concerned with definitions of basic terms which will be
then used throughout the whole text. Also the definition of word-formation is provided.
compounds, because this division will be essential for the following analysis. Neo-
briefly than compounding. However, only those are included which are relevant for the
latter part of the thesis because of being on the borderline with compounding in some
cases. Therefore, this listing should not be treated as exhaustive. It will include
After the two theoretical chapters, the third chapter will form the central part of
this thesis. The whole chapter will be devoted to the analysis of the transcripts.
words which are on the borderline with phrases, proper nouns as well as scientific terms
will not be taken into account. Compounds will be classed into categories based on the
word class of their constituents. All the criteria will be described at the beginning of the
chapter. The most productive patterns will receive the greatest attention. The basic
7
division among the categories will be made according to semantic criteria.
Further division will be made and commented for each of the categories. Neo-
classical and phrasal compounds, which are special subtypes of compounds, will not be
classified in this way but will be also briefly mentioned. The analysis will be backed
compounding patterns between Friends and The Big Bang Theory. Language of main
characters will also be compared among each series with focus on compounding.
However, these comparisons will be rather brief as they will serve mainly for providing
a general picture. As a conclusion of the analysis, all the findings will be clearly
8
1. Word-formation
This chapter serves as a starting point to the whole thesis, since it presents the
basic notions of the English word-formation. It will not discuss the term ‘word’1,
word exists (Bauer, 1983, p. 8). At first, the terms ‘lexeme’ and ‘word-form’ will be
introduced. This will allow to define other constituents of word-formation which will be
recurring throughout the whole thesis. The term ‘word’ will be used vaguely between
itself will be provided which will be followed by the notions of productivity and
made. All examples will be written in italics, while non-grammatical forms will be
preceded by an asterisk and new terms will appear in bold the first time they are
mentioned.
This section will outline notions which will be used in the further text.
among scholars will be felt. This will be also the case here although larger discrepancies
“all the possible shapes that a word can have” (Bauer, 1983, p. 11). On the other hand,
1
For detailed discussion on the term word, see Matthews (1974); for overview of various potential
definitions of ‘word’, see Kavka and Štekauer (2006, p. 8-12).
9
occasion, word-forms2 are much more concrete and realize lexemes (Bauer, 1983, p.
11). This can be illustrated by the lexeme FLY3. This lexeme subsume all the words
like flies, flying, or flew and these words are word-forms of the lexeme FLY. These
forms are also termed inflectional (Huddleston and Pullum, 2002, p. 27) and therefore
process of creating new lexemes (Lyons, 1977, p.522). Free forms are those that can
occur alone, such as all the examples, while the term bound forms refers to entities that
cannot occur in isolation. Typical examples of bound forms are affixes5. They could be
defined as bound forms attached to lexemes that do not realize unanalyzable lexemes
(Bauer, 1983, p. 11). They are usually divided into prefixes such as dis- or re- and
suffixes (-dom, -ness). Prefixes are attached before the lexemes, while suffixes occur
behind them.
Contrary to affixes, roots are most usually free forms and are “at center of word-
derivational processes” (Stockwell and Minkova, 2001, p. 69). All lexemes include
roots, as they are the forms that remain after removal of all inflectional as well as
derivational affixes (Bauer, 1983, p. 20). Stem is concerned only with inflection and so
it is form of the word without all the derivational affixes, which are “analysable as
such” (Quirk et al, p.1519). Therefore, in unbeatable, beat is the root, while the stem is
unbeatable, as the word does not contain any derivational affixes. Finally, base is “any
form to which affixes of any kind may be added” (Bauer, 1983, p. 20). This means that
all stems and roots are bases, but this is not so the other way around. In the example of
unbeatable, beatable may be a base for a prefixation by un-, but neither root nor stem.
2
For further discussion regarding lexemes and word-forms, see e.g. Matthews (1974) or Beard (1995).
3
Lexemes are conventionally written in capital letters.
4
Derivation is also often defined as addition of affixes (Lipka, 1992, p. 72).
5
Several approaches to affixes were adopted by Lieber and Štekauer (2009).
10
Bases are normally free, but also bound bases can be found (Huddleston and Pullum,
2002, p. 1625). Words that can be broken down into smaller meaningful units are called
complex (Plag, 2003, p, 12). These units are named morphemes (Bauer, 1983, p.11).
Morphemes can be defined as “the indivisible semantic units” (Matthews, 1974, p. 20).
In contrast, words, which are not analyzable this way are simple, or simplex (Bauer,
1983, p. 30). For example, happiness or unbeatable are complex words, while beat or
After introducing the essential terms for the study of word-formation, definition
combination of lexical6 morphemes with each other” (p. 70) and also stated that it could
producing complex forms7 (Bauer, 1983, p. 31). It is important to remark, what should
be understood under the heading of word-formation. Lipka (1992) noted that word-
formation includes not only derivation (i.e. addition of affixes), but also “the
combination of free lexical morphemes” (p. 72). This combination will be in short
divided into derivation and compounding. However, some categories are not covered by
either of these headings (Bauer, 1983, p. 30). Inflection does not fit the definition at all
and therefore is usually separated from word-formation (Lipka, 1992, p. 70). But, as
admitted by Bauer (1983), there are sometimes no clear dividing lines between
6
Morphemes that are used to produce new lexemes, see Lipka (1992, p. 69-70).
7
Highly similar definition is also used by Plag (2002, p.17).
11
1.3 Productivity and Lexicalization
This introductory chapter will now be concluded with the notions of two
important terms, which are recurring throughout the whole thesis. Productivity will be
one of the main concerns in the analytical part of this thesis. A process is productive8,
when it is still able to create new words, or more precisely, new complex forms (Plag,
2002, p. 55). Plag dealt with productivity at the example of affix, but any process may
and rule-governed (Lipka, 1992, p. 92). When a process is not used for coining new
lexicalization. Lexemes are generally said to be lexicalized when they are or used to be
further morphologically analysable, but they cannot “be formed with their present
1629). Alternatively, it can be said that a complex lexeme gradually becomes a single
integration of lexeme with a specific meaning and form into already established lexicon
8
For detailed discussion regarding productivity and creativity, see Bauer (1983, p. 62-100).
12
2. Word-formation Processes in English
After dealing with word-formation in general, this chapter will provide overview
As the main aim of this thesis is to classify compounds in TV series, main attention will
similarity of compounds and syntactic word groups (phrases) will be pointed out.
Afterwards, basic types and terms regarding compounds according to their semantic
criteria will be provided before moving on to the sub classification of compounds. This
classification will also serve as the starting point for the practical part of this thesis. To
conclude this chapter, selected other word-formation processes will also be listed, as
they often overlap with compounding and sometimes they are also on the borderline
with compounding.
2.1 Compounding
(2000, p. 99) it is even often regarded as the most productive process of the English
word-formation. Plag (2002) also maintained this opinion. Furthermore, Plag (2002, p.
169) added that compounding is the most controversial process in English in terms of
linguistic analysis. This is due to the fact that “numerous issues remain unresolved and
convincing solutions are generally not easy to find” (Plag, 2002, p. 169). Because of
attempt to find one, which would be highly unfeasible, will not be made.
9
See Bauer (1983).
13
2.1.1 Definition
(1983, p. 11) who described it as “the process of putting two words together to form a
third”. But as this definition is rather vague at first sight and would also rule out many
established compounds, need for providing a more complex one can be felt. The most
suitable definition for purposes of this thesis may be the treating of compounds as free
lexical units consisting of two or more roots and “functioning both grammatically and
semantically as a single word” (Quirk et al., 1985, p. 1567). The “freeness” of its
scholars such as Lipka (1992, p. 83). Adams (1973, p. 30) added that order of
compound elements is fixed which leads to a conclusion that they cannot be separated
by inserting another free form between them. Lipka (1992, p. 83) also remarked that
English, it is most typical for compounds to consist of two bases, but constructions with
202) pointed out, they can be always broken into two analysable elements where each of
This chapter will once again focus on the problem of differentiating compounds
of delimiting compound from syntactic groups has not been found yet (Štekauer, 2000,
p. 100). This problem involves compounds such as school uniform or freezing point.
As Vachek (1976, p. 17) pointed out, this fact is due to the highly analytical character of
10
Exceptions to this rule will be discussed later under the heading “neo-classical compounds”.
11
They are known as “phrasal compounds” and will be also mentioned later.
14
Modern English. Therefore, boundaries between words and word-groups are
significantly vague. This is not the case in synthetic languages like the Slavonic ones,
insert (broadly speaking) infinite word-group, “that has been taken out of its semantical
environment” (Vachek, 1976, p. 17) before a given word. Consequently, its semantical
environment is changed as well as syntactical function of this word group, which can
even stand for the whole sentence. Vachek (1976, p. 17-18) demonstrated this on the
compounds and will be commented on in the analysis. This occurs indeed very often in
compound is treated as “one word” and phrases as “two or more words” (Matthews,
1974, p. 94), the reality is considerably more complicated, as pointed out in 2.1.2. When
which comes a result of this process may be either compound or a free phrase, as was
applied to those constructions (e.g. green house, black bird), but the question is whether
is it really possible to come up with a definite answer. Adams (1973) stated that there is
no answer to this problem which would cover all the cases. It can be however supposed
that distinctions can be made “on phonological, syntactic and semantic grounds”
(Jackson and Amvela, 2007, p. 93). Thereupon, these aspects will be discussed
15
successively from now on by mentioning selected criteria which may be used for
recognising compounds.
2.1.3.1 Stress
Stress may often distinguish a compound from a phrase. This is caused by the
fact that words in English tend to have one primary stress, so compounds can be often
recognized by “having a stress pattern and a lack of juncture” (Jackson and Amvela,
2007, p. 94). It is generally believed that when the primary stress is on the first
constituent, the construction is regarded as a compound. When the primary stress falls
on the latter element, than the expression tends to be recognised as a noun phrase. This
can be seen on an example12: a ‘dark room vs. a dark’ room where the first expression
is a compound, while the latter one a noun phrase. Quirk et al. (1985, p. 1569) state that
“almost all compounds have this accentual pattern”. However, not every compound
receive this stress. According to Štekauer (2000, p. 100) this criterion cannot be
regarded as a “hard-and-fast rule”, as there are multiple expressions with two main
stresses which are definitely accepted as compounds (e.g.: ‘trade ‘union). Moreover,
Štekauer (2000) says that number of compounds with this stress pattern is increasing
(1974) pointed out, this criterion is highly subject to variation, as there tends to be no
consistency of stressing the compounds, even among native speakers. It also often
depends on “sentence stress and intonation” (Matthews, 1974, p. 98) and therefore this
Compounds are spelled in three different ways (Lieber and Štekauer, 2009, p.
376) – they may be solid, i.e. written as one word (e.g. blackboard), two hyphenated
12
‘Marks the primary stress.
16
words (baby-sitter) or written as two separate words (living room). That already
indicates a problem, as not much effort is needed to find a compound which can be
written in either of the three above mentioned ways. Example of this can be the word
girlfriend where all the forms girlfriend, girl-friend and girl friend are acceptable.
Therefore, spelling is often regarded as not particularly relevant and unreliable criteria
compounds with lexicalized meanings can be found, such as blackbird, it was already
nonce formations13 and that is why lexicalization will probably never be treated as a
2.1.3.3 Inflectibility
functions of compounds” (Jackson and Amvela, 2007, p. 94), also appears frequently as
a test of compoundhood. An assumption is made in Jackson and Amvela (2007) that the
constituents of compounds are usually inflectionless and that they should be treated
(and inflected) as a single lexical unit. This works perfectly well in most cases, where
be found again, like girl’s club or children’s hour (Lieber and Štekauer, 2009, p. 376).
13
E.g. compound word crapweasel meaning insincere person, coined in The Friends TV series.
14
Consider also baby-sat.
17
2.1.3.4 Inseparability and modification
comparison to methods described above) among the scholars. Lieber and Štekauer
(2009, p. 377) even stated that “inseparability is perhaps the strongest test of
compoundhood”. Also Jackson and Amvela (2007, p. 93) said that all compounds are
normally inseparable in a way that no extra element may be put between its constituent
parts and such tests as separating the first element from the head were also successfully
applied by Adams (1973). This implies that while there is nothing wrong with a pretty
girl friend, *a girl pretty friend is not acceptable anymore and cannot be recognized as a
compound in English.
compoundhood of the original (Lieber and Štekauer 2009, p. 377), as shown above.
separately. As seen in example taken from Jackson and Amvela (2007, p. 93-94), air-
sick can occur in a phrase like seriously air-sick, where seriously modifies the whole
compound, but it cannot modify just the second element which means that phrase air-
Semantic criteria are the last ones that are to be discussed in this chapter, and
since the formal criteria mentioned above are not satisfying in many cases, semantics is
often emphasized (Matthews 1974, p. 95). This leads to Jespersen’s (1974) supposition
that “if the meaning of the whole cannot be logically deduced from the meanings of the
Jackson and Amvela (2007) stated that compounds tend to acquire very specialized
18
meanings and therefore their constituents may lose their original meanings. This works
very well for lexicalised compounds such as blackboard (as it may have also a different
color) or dustbin (it is not entirely restricted to dust), as exemplified in Jackson and
Amvela (2007, p. 94). However, this is not true for all compounds as there are also
expressions whose meaning can be easily predicted from the meaning of their elements,
idioms15, where an expression such as He made his mind up has a literal meaning in
only greatly limited amount of cases, while there are “good reasons” (Matthews, 1974,
After this overview of various tests of compoundhood, it is clear that none of the
above mentioned criteria would cover all the possible cases. However, it is to be noted
that these criteria “operate simultaneously” (Jackson and Amvela, 2007, p. 94) and
when applied all together, the probability of successful distinguishing compound from a
phrase rises.
In this section, only compounds consisting of two elements will be included. The
so-called “phrasal compounds” will not be mentioned here. Compounds can be divided
into main groups by taking lots of different approaches which appear to vary among
linguists. Various criteria are therefore applied. This chapter will provide an overview
of the most common ones and although they may have much in common and even
15
Expression whose meaning is not wholly predictable from its structural description (Lieber and
Štekauer 2009, p. 181).
19
2.1.4.1 Primary and synthetic compounds
named them synthetic (verbal) compounds (Štekauer, 2000, p. 102). Examples include
blackbird or birthday. Selkirk (1982, p. 24-25) also pointed out that while range of
possible relations between non-verbal compounds is broad, this is not the case with
As already indicated in the title of this section, this division is done according to
syntactic principles. This is proposed as the starting point by Stockwell and Minkova
(2001, p. 10). Like in 2.1.4.1., this sorting of compounds is once again a very general
one. By syntactic compounds, forms which are formed with respect to regular rules of
grammar are meant. Stockwell and Minkova (2001) added that those compounds are
formed similarly to sentences and are usually not listed in dictionaries (p. 10).
Therefore, birthday (the day of birth), or daylight (light which occurs during the day)
cannot be predicted from the regular rules of grammar, e.g. ice cream and water-proof.
This means that their meaning has to be looked up in dictionary (Stockwell and
20
compounds (Jackson and Amvela, 2007, p. 97). Štekauer (2000) added that constituents
in lexical compounds are “ordered differently from the corresponding syntactic phrases”
(1983), or Jackson and Amvela (2007). As this approach will be followed in the
practical part, this classification will be of the biggest importance in this thesis.
Terminology varies here, but this thesis will mostly stick to the Bauer’s (1983)
terminology including slight subdivision made by Jackson and Amvela (2007). Most of
the English compounds tend to have “modifier-head structure” (Plag, 2002, p. 173)
which means that the left-hand element modifies the right-hand one. As pointed out by
Ackema and Neeleman (2004, p. 81), this ensures that there is a semantic relation
They can be in short described “to have a head, and that head is on the right” (Selkirk,
1982, p. 19). Bauer (1983, p. 30) added that the left-hand element does not mark neither
gender, nor number in his classification. In his words (p. 30), “the compound is
type of chair and boyfriend is a boy who is also a friend. As the latter compound can be
often described vice-versa, difference between these examples can be seen. Compounds
such as armchair or call-girl will be thus named subordinate (Jackson and Amvela,
2007, p. 97) as they contain one head root and a modifier. On the other hand,
compounds having two head roots (e.g. girlfriend, maidservant) will be termed
coordinate8 compounds (Jackson and Amvela, 2007, p. 97). Bauer (1983, p. 30)
16
The same terminology is also used by e.g. Heine and Narrog (2010).
21
referred to this type of compounds as appositional compounds and said that both
present (Bauer, 1983). Therefore, mainly proper nouns are dvandva compounds in
English (Huddleston and Pullum, 2002). This type of compounds can be exemplified by
However, not all compounds fit into this category. Some of them appear to
“have no head at all” (Selkirk, 1982, 19). For example, skinhead is not a type of skin,
pickpocket is also not a kind of pocket at all. These compounds are known as
exocentric17, where the semantic head is not expressed (Bauer, 1983, p. 30). They are
also sometimes named bahuvrihi compounds (Bauer, 1983, p. 30). It is to be noted that
this terminology is far from constant and can differ considerably across the scholars.
Therefore, it was possible for Plag (2002, p. 188) to state that dvandva compounds are
Previous chapters were concerned only with compounds which correspond to the
definition formulated in 2.1.1. This means that all of their constituents are free forms.
However, a special type of compounds, where at least one of its elements is a bound
form, can be often found in English. This is due to the fact that English tends to borrow
from Greek or Latin extensively to coin new words (Adams, 1973, p. 129). These words
17
Some linguists argue whether exocentric compounds really exist in English, see e.g. Štekauer (1998).
22
contain so-called combining forms (Štekauer, 2000, 103), such as socio-, micro-, tele-,
phile-, phobe- and many others. Resulting words are generally known as
Latin origin are combined to form new combinations” (Plag, 2002, p. 199). They mostly
are highly common also in everyday speech. Therefore, proper clarification of how to
treat them under the heading of affixation18 but also added that arbitrariness is involved.
This comes from an assumption that neo-classical forms can be added to some bases
precisely like affixes (Bauer. 1983, p. 213). For example, music + al is a clear example
conclusion” (p. 213) that forms such as electrophile and galvanoscope, including only
prefixes and suffixes while having no roots, can exist. Furthermore, these combining
forms can perfectly combine with each other as shown above. This is not the case with
“regular” affixes (Bauer, 1983) and this implies that forms like *electroness or
*electroization do not exist. Also, they can be combined even with affixes “which is a
the combining forms can have their independently found counterparts serving as
independent words, like photo, mini or kilo. Based on these facts, neo-classical items
will be treated as combining forms and therefore as compounds in this thesis, mainly in
its analytical part. Bauer (1983) also observed that these compounds are very productive
18
Affixation will be discussed in 2. 2.
23
Of course, compounds can be classified further. As Bauer (1983) mentioned, this
can be done in many different ways, out of which none proves to be ultimately
successful. Furthermore, subtypes of compounds vary from four to one hundred among
not known yet. Compounds can be classified by e.g. the form classes of compound’s
clausal paraphrases” (Quirk et al., 1985, p.1570) and so on. Some authors, like Jackson
and Amvela (2007), or Adams (1973) use combinations of the above mentioned criteria.
However, this thesis will stick to classification proposed by Bauer (1983) which is
based on the form classes. The classification will be complemented by Jackson’s and
2.2 Affixation
Therefore, inflectional affixes as described by Jackson and Amvela (2007) will not be
taken into account, because they serve for creating new forms of the same word.
(Bauer, 1983) which was defined in 1.1. As Bauer19 (1983) continued, the lack of clear
dividing line between compounds and derivation is often felt, as described on suffixes
like –hood where it is unclear whether hood functions as lexeme, or as a bound form in
Štekauer (2000), infixes are almost non-existent in English although some examples,
such as abso-goddam-lutely may be found. Affixes can be also further divided into
19
Described in more detail on p. 35-36 in Bauer (1983).
24
class-changing and class-maintaining. Instances of the latter type is e.g. prefixation by
re- from consider to reconsider while example of the first type is suffixation by –ment
from pay to payment. Bauer (1983) observed that in English, “prefixation is typically
Matthews (1974) further stated that processes of suffixation are more common in
English, as they serve for most lexical derivations. As added by Jackson and Amvela
(2007), affix –ly which creates adverbs from a given stem, is the most productive
2.3 Conversion
word’s syntactic category without any commitment change of form” (Huddleston and
Pullum, 2002, p.1640). Therefore, it is possible to change the noun trouble into a verb
while not touching its form. Because of this, Adams (1973) treated conversion as a
formation process, because they feel it is rather using word in its less typical word class
(Štekauer, 2000). Conversion also creates lots of problems when classifying compounds
which will be the concern of practical part of this thesis. This is mainly due to the fact
that there are almost no rules of conversion (Bauer, 1983). Bauer (1983) also pointed
out that all the word classes and even affixes can serve as bases for conversion.
still extremely productive and its forms are even often given separate entries in
dictionaries.
25
2.4 Clipping
a lexeme is shortened without change of its meaning and word class (Štekauer, 2000, p.
111). The lexeme may be either simplex, or complex (Bauer, 1983). As remarked by
Stockwell and Minkova (2001), not only single words, but also phrases may be
shortened in this way. For example, ad or advert are clipped from advertisement, while
zoo from zoological garden. Because of this, clipping may be seen as an unpredictable
formation (Bauer, 1983) as there is no general rule which base of the word will be
shortened and how many syllables will be retained. Usually, the first part of the word is
retained as demonstrated on the example above, but also the middle can prevail like in
jams clipped from pyjamas. Clipping may also serve as a base for further word-
formation processes, typically for conversion (Štekauer, 2000). They can be also used in
2.5 Blending
process of merging two words into a single one (Štekauer, 2000). Furthermore, the
enough of both words is usually retained. Examples of blending include brunch, which
was created from breakfast and lunch, or smog, where smoke and fog were blended.
However, examples with another structure, such as Nixonomics where the first word
remains untouched, can be also found and therefore blending should be also regarded as
blends resemble coordinate compounds discussed in 2.1.4.3 and also Štekauer (2000)
26
(p. 111) or as a “formal reduction of compound” (p. 111). Bauer (1983) added that
while blending is rather not a clearly defined category, as it overlaps with compounding,
2.6 Back-formation
As the term suggests, back-formation occurs when the process of forming new
word from a base seem to be reversed (Adams, 1973). It usually refers to clipping of a
recognizable affix, such as how to burgle was formed from burglar, or cherry from
cerise (Stockwell and Minkova, 2001). As Bauer (1983) remarked, most of the words
resulting from back-formation are verbs. However, as pointed out by Stockwell and
Minkova (2001), the majority of words formed are not transparent, such as edit or dive
which both are not source for expected derivation, editor and diver respectively. On the
other hand, advisor was really formed by affixation from advise (Quirk et al., 1985). As
also noted by Quirk et al. (1985), back-formation is still highly productive nowadays.
27
3. Analysis of the Corpus
and its processes, this chapter will be concerned with practical analysis and will form
the central part of this thesis. It will focus on compounding and the main goal of the
analysis will be to classify compounds as found in a corpus. This corpus was created
using transcripts of episodes from two well-known TV series, namely Friends and The
Big Bang Theory (BBT). The corpus is enclosed to this thesis on a CD. The transcripts
of Friends were taken from Friendstranscripts.tk, while the source of BBT’s transcripts
is transcripts.foreverdreaming.org. The main reason for choosing these series was the
fact that their main focus is on creating entertaining episodes through series of dialogues
rather than concentrating on the plot itself. Furthermore, these series seem to be based a
bit differently which can prove as an interesting contrasting element. The above
mentioned facts lead to an expectation that number of compounds found in this corpus
should be high. And indeed, compounding was the most frequent word-formation
process present. Therefore, further subclassification can be made and it was pointed
As already mentioned before, this analysis will adopt Bauer’s (1983) approach
which relies on word classes of compounds’ elements. It will also largely consider
their form rather than function. As Bauer (1983) admitted, this classification is not
perfect in any means, but there is no classification that could not be considered non-
controversial (p. 202). This method of classification is perhaps the most straightforward
one and can be also clear for readers without knowledge of theoretical terms. Also,
(Bauer, 1983, p. 202). However, this does not prevent ambiguity, because it is often
28
highly unclear to what class a given constituent of compound should be assigned which
will be remarked several times. This is, according to Bauer (1983), mainly because of
the high amount of conversion20 in English (p. 202). That is the biggest disadvantage of
this method but, as noted several times before, problems are inevitable while studying
compounds, with the endocentric being split into coordinate and subordinate
with respect to Bauer’s terminology will be recognized too. What Bauer (1983) terms
endocentric coordinate compounds as written above. Remarks from other sources will
main character’s compound usage pattern21 was also recorded for each series. Because
of this, such transcripts were needed where there are names of the characters included
next to their utterances. Therefore, only first seasons of both series were analyzed, as
they contained these information for Friends as well as for BBT. Furthermore, not
whole series, but the first ten episodes will be discussed here, as the pattern remains the
same for the rest of the season. Language of individual characters as well as comparison
of both series will be later discussed in separate chapters. Before moving on to the
actual classification it will be useful to specify which compounds were taken into
20
See 2.3.
21
This is usually termed idiolect, see Saeed (2009).
29
3.1 Borderline Cases
This section deals mainly with formations discussed in 2.1.2 where the lack of
clear boundaries between compounds and phrases is the prevailing factor. Since any
universal solution of this problem has not been presented so far, treatment of formations
such as bowel movement, or school uniform is still arbitrary. Although most of the
compounds, this will not be followed here as a result of the discussion presented in
2.1.2. It is mainly due to the fact that these phrases are not considered as compounds in
synthetic languages, such as Czech. The only exceptions are names of rooms, such as
living room, in order to prevent the situation where some rooms (e.g. bathroom) are
included while others not. In addition, big bang was taken into account as well, as it can
be found in the title of BBT. However, this is not the case with these words written
hyphenated and such words will be included. On the other hand, phrasal compounds22
will be included, because they highly contribute to the flow and amusement of the
mentioned in the introduction to the analysis, will be included and analysed separately
terminology were also excluded due to the fact that they are not particularly relevant for
the study of the language itself. This is also the case with words which would not
probably seem to be compounds for many people, as they incorporated themselves into
the language core over the years. Examples of these words are pronouns such as
someone or everybody, adverbs like always or outside but also verbs like welcome and
become. This also affects nouns as afternoon, goodbye and so on. The same is the
22
Furthermore, they also tend to be written hyphenated, which corresponds to the criterion above.
30
situation with compounds containing names of people, locations, products, or events23.
All the compounds taken into account were highlighted in the corpus.
While the previous chapter may seem highly restrictive, it serves rather as a cut
to eliminate all the constructions not particularly relevant for a linguistic analysis. Many
compound words were still found. In total, 271 of them were recognized. This shows
that even after limiting the count of episodes to ten for each series, the number of
registered compounds was still considerable enough. Most of them were written as one
word, i.e. solid, for example cheesecake or blackboard. Examples of compounds written
hyphenated, which cover about one third of entries, include fat-free or pant-suit. After
disallowing most of the possible compounds written as two words in 3.1, only ten such
were found, where nine of them were names of rooms with the tenth one being big
bang, as described in 3.1. This subgroup would otherwise form the majority of
compounds. However, as noted in 2.1.3.2, many words can be written in more of the
ways mentioned above. Two compounds also contained linking elements, concretely
languages, such as German (Štekauer, Valera & Körtvélyessy, 2012), but apparently not
The table below presents the basic notion of compound patterns as found in the
transcripts. Adverbs were not highly represented because although many particles may
function adverbially, they were still classified as particles for consistency with Bauer’s
headings of columns classes of right-hand elements and ‘x’ marks that no compound of
23
They are generally termed proper nouns (Bauer, 1983).
31
that type occurred in the corpus. This labelling will be also kept in all the following
tables.
Adverb x now-pink x x x
Table 1
Of course, this table does not include classes of resulting compound words, as it
would become highly complicated and hardly understandable. Apart from these
patterns, phrasal and neo-classical compounds were also recorded but they are not
included there, as they will be discussed in separate chapters. This overview is mainly
to demonstrate the patterns found in the corpus in a general manner, while the
subsequent chapters will deal with the compound words already after their breakdown
to their corresponding formal classes. The classes found here are nouns, adjectives,
verbs and adverbs, with phrasal and neo-classical compounds to follow. Chapters with
(1983) noted, this is even more the case when dealing with compounds with more than
two constituents (p. 202). However, this will be the matter of concern in 3.7 and this
32
not unexpected that vast majority of recognized compounds fall into this group. In total,
roughly two thirds of all compounds found are nouns. However, this group has not the
highest variety of compounding patterns. Altogether, five patterns are found. For
Noun warehouse x
Adjective flatware x
Particle download x
Table 2
This chart explains, why the number of types of compound nouns is relatively
low compared to their total count. The main reason for this is that the second element is
a noun in extensive majority of cases, only a few noun compounds contain particles as
their second constituent, while other classes are not found at all in the second position.
All the compounds written separately fall into this class with vast majority of the rest
written as one word. The five patterns depicted above will be discussed in turn in the
following chapters. Each of them will feature a few randomly chosen examples of the
given patterns with information about its source series and episode provided in the
brackets, which will represent the compound types found and will be also mentioned
33
(5) …I happen to be a human beatbox. (BBT, 5)
This pattern is by far the most frequent one and it will therefore receive the greatest
attention. Ryder (1994) even devoted a whole book to noun + noun compounds
interpretations (Ryder, 1994, p. 4-5). Furthermore, this subgroup is also regarded as the
most productive pattern of the whole English word-formation (Huddleston and Pulum,
2002, p. 1647). This indicates that many compounds which does not seem to be very
similar can be found in this category. As Bauer (1983) proposed, many criteria would
be possible to classify compounds in this subgroup, but the division described in 2.1.4.3
As the examples (1) – (4) suggest, four subtypes of noun + noun compounds are
found in this category. Generally speaking, the vast majority of these compounds are
endocentric. They are represented in the examples (1) and (2). It was mentioned in
2.1.4.3 that endocentric compounds can be split into subordinate (1) and coordinate (2).
Endocentric subordinate compounds are the most frequent by far, further examples
include bathroom, eggplants, or jellyfish. Due to the restrictions applied earlier, only
two examples of gerund + noun compound, concretely living room and waiting room,
are found. This type is mentioned by Bauer (1983) to be particularly difficult to classify,
“since a gerund has both nominal and verbal characteristics” (p. 203). Furthermore, the
decided to keep this pattern within the noun + noun subgroup. The main reason was that
the semantic relationship between the gerund and the noun is most similar to the one
found in noun + noun compounds (Bauer, 1983) as opposed to verb + noun compounds,
34
which will be discussed in 3.3.2. Based on this, living room can be regarded as ‘a room
suitable for living’ and therefore seeming suitable for including it here, with waiting
room being treated similarly. Subordinate compounds contain words with literal as well
as more metaphorical meaning; the latter being represented by e.g. airport, snowman, or
painkillers. To illustrate the notion how diverse this group is, also compounds with part-
found. This indicates that it is virtually impossible to classify these compounds into
Coordinate compounds are much less productive in the corpus, while dvandva
compounds are not found at all. Apart from clear examples like boyfriend or girlfriend,
showbusiness. While stairwell and weekend also seem to fit into this category, the case
with compounds like staircase is a little different. Although it can be regarded as a type
of stairs, case has already lost its original meaning in this compound. Therefore, as it
does not refer to any particular kind of case or even to a case in sense that something
happens, it should not be treated as an endocentric compound at all. That implies that it
Exocentric compounds, which are represented by (3) and (4), are much less
frequent than the endocentric, although they happen to be more productive than the
endocentric coordinate compounds. As Lieber and Štekauer (2009, p. 378) pointed out,
exocentric compounds are often used as metonyms for people (3). Indeed, the class of
exocentric compounds turns out to be the most common in the corpus, represented by
crapweasel which was coined in the Friends TV series. Exocentric compounds not
referring to human beings (4) are of course found as well although they were in
35
minority. They can be exemplified by bedrock, cocktail and pitfalls. These tend to be
expressions which are used in an informal language, as it is the case with the majority of
felt here as well but is not strong enough to classify these compounds as exocentric.
The last type of noun + noun compounds, represented by (5), is also exocentric.
Two of this sort are present in the corpus: beatbox and workshop. But they seem to
differ from the types (3) and (4) discussed before. In previous examples, hammerhead
was neither kind of hammer nor type of head, mother-kisser was neither a mother nor a
kisser at all and so on. However, while workshop is also not a shop it is obviously
associated with work. Similarly, when beatboxing, one does not box but produces beat.
This could indicate that the semantic head was on the left. But as the left-hand element
generally serves to modify the right-hand one and the syntactic head is on the right,
because forms like *boxbeat or *shopwork are not grammatical, these compounds must
be considered exocentric as well in the end. This demonstrates that even among the
Examples of noun + noun compounds also include words which can be seen as a
compounds like mailman, congressman and snowman, the process of changing lexeme
MAN to suffix –man is ongoing (Bauer, 1983, p. 36). That is because it became
possible to say female mailman, so man starts to lose its original meaning. Furthermore,
mailman is pronounced with a reduced vowel as opposed to the lexeme MAN which
contains full vowel. However, as pointed out by Bauer (1983), “contrast is still made
between policeman and policewoman” (p. 36), which together with the fact that plural
of mailman is mailmen and not *mailmans suggests that MAN is still more lexeme than
36
suffix in English. However, not all compounds of this type are necessarily endocentric.
ambiguous case, because it refers to object made from snow which only resembles a
electronic device.
There are also compounds whose status depends on context and the degree of
literality in their usage. For example, showcase can be a genuine endocentric compound
when referring to a box which is containing exhibits, while it is highly borderline case
when representing an event. Also another example of derivation was found in this
retaining its compound status. One compound in this category, bridesmaid, also
cam was clipped from camera. This illustrates that word-formation processes can freely
(2) …he was trying to fight a bobcat for some licquorish. (BBT, 4)
As noted before, gerunds were classified as nouns and therefore this category will be
concerned with verb stem + noun compounds. (Bauer, 1983, p. 204). This subgroup of
noun compounds is not regarded as very productive in general (Lieber and Štekauer
(2009, p. 378) and it also appears rather minutely in the corpus; only eight examples are
recorded. Two patterns of verb + noun compounding are found. The first one (1) refers
to situation “where the noun is the direct object of the verb” (Bauer, 1983, p.205).
Another examples include crossbar and breakfront. These compounds are exocentric.
37
Compounds such as breakfast or breakfront fitting into the (1) pattern are also named
French (Kavka and Štekauer, 2006), or Romance24 (Bauer, 1998) type of compounds
because of their origin. This pattern is mostly rare and currently not very productive
The rest of the verb + noun25 category subsumes compounds, where the noun is
not the direct object (2). This pattern appears to be slightly more frequent in the corpus
with examples like chopsticks, notebook or popcorn. These compounds are endocentric
and all of them are subordinate as they have a clear head on the right. Altogether in this
category, no compounds denoting people were found and only bobcat refers to an
animate being. As remarked by Bauer (1983, p. 205), the (2) pattern is the most
problematic to limit. That is particularly because deciding whether the first element is
Therefore, it is indicated by Bauer (1983) that some of these compounds may also fit
into the noun + noun category. For example, notebook could be glossed as ‘a book into
which one notes’ or as ‘a book into which one takes notes’. In the second case, note
functions as a noun. However, for purposes of this thesis, compounds with this
This pattern is still productive, particularly when the compound constituents are written
separately (Bauer, 1983). However, this analysis is not concerned with such
24
For further discussion regarding this type, see Bauer (2008).
25
For detailed classification of verb + noun compounds, see Huddleston and Pullum (2002).
38
heading but they will be mentioned in a separate section. According to Huddleston and
Pullum (2002, p. 1649), the productivity of this pattern is actually quite low when
creating new words. Yet, it is one of the most numerous patterns in the corpus, yielding
only to the dominating noun + noun category. As further remarked by Huddleston and
Pullum (2002), many compounds from this category involve lexicalization and therefore
they highly differ from the syntactic phrases (p. 1650). Such compound is exemplified
in (1). Such compounds are exocentric and used for denoting animate beings. So,
(1998), “this existence of one of these with a particular meaning does not appear to
block its existence with another, unrelated, meaning” (p. 5). This also works on
examples from the corpus, as blue bell could definitely relate to ‘a bell which is blue’ as
well. However, this construction would lose its compounds status and would be
regarded as a phrase. That is not possible with, for example, exocentric noun + noun
different sense.
surprising. Most of the adjective + noun compounds were endocentric (2) and all of
them were subordinate. In total, 15 out of 17 compounds have this structure. This
subtype includes classic examples like big bang, flatware, gentlemen, softball or
whiteboard but also some unclear cases. Some of the compounds are spelled
treated as a borderline cases, as they would not be mentioned here if they had been
spelled as two words. Like formations discussed in the verb + noun category, goldfish is
also a compound, where the class of gold is not clear cut. It could be adjective in ‘a fish
39
which is gold’ or noun in ‘a fish that resembles gold’26. Once again, it was eventually
This pattern appears to be moderately productive with just over ten compounds present.
Prepositions and adverbs, which are collectively referred to as particles in this analysis,
both appeared as the left-hand elements. Adams (1973, p. 114) remarked that
complicated. However, solid spelled nouns with particles are treated as compounds by
3.3.1, particles may be seen to function as an attributive adjective. But Adams (1973)
background is usually not the main point of concern and so it stands in the back (either
literally, or metaphorically).
As far as the individual compounds are concerned, both exocentric (1) and
outlet, offspring and underdog. The last one is particularly interesting because although
its right-hand element indicates that is a kind of dog, it is typically used for
denoting people. This is also the only compound denoting animate being. Endocentric
compounds are more frequent and all of them are subordinate. They can be split into
two groups according to status of their right-hand element. The first one includes
26
This gloss is taken from Ryder (1994, p.5), where goldfish is treated as a noun + noun compound.
40
compounds where the noun did not result from the process of conversion (2). It contains
formations like background, overcoat or underside. The latter one is filled with
include download, outburst, overlay or underwear. These groups are almost balanced in
the corpus. Aside stands back-story which seems to fit into the pattern (2). However, it
instance of blending.
(1) This is the worst break-up in the history of the world. (Friends, 5)
This last subgrouping of compound nouns is similar to the previous category in some
aspects since it contains particles as well. They are equally frequent in the corpus as the
particle + noun type. However, these compounds are all exocentric, as they do not
contain any nouns. As Bauer (1983, p. 206) pointed out, compoundhood of some
formations in this group is arguable. That comes down to the fact that they can be seen
as nominalizations of phrasal verbs (1). Adams (1973) did not regard them as
compounds because of that. This type is common in the corpus and includes compounds
like blackout, break-up close-up, or knockout. The rest of compounds falling into this
category were not derived from phrasal verbs (2). Although Bauer (1983) remarked that
“they may be coined by analogy with phrasal verbs” (p. 206). They include flashback,
feedback, makeup or voiceover. The pattern (2) is slightly more productive in the
27
Adams (1973) used the term ‘zero suffix’ for this group.
28
See the Online Etymology Dictionary for more information about etymology of backstory.
41
All the nominal compounding patterns as well as their productivity are
illustrated in the following table. Semantic criteria used during the analysis are also
N+N 122 94 9 19
V+N 8 5 0 3
Adj+N 17 15 0 2
Par+N 11 7 0 4
V+Par 11 0 0 11
Table 3
As expected, compound adjectives are much less frequent than nouns. They
account for about one quarter of the total number of compounds found which still
makes them rather productive. Bauer (1983, p. 209) noted that number of patterns
which form adjectival compounds is high. That was also the case in the corpus, as
compound adjectives are the most variable category based on quantity of forming
patterns found in the transcripts. However, some of them were highly marginal. Both
Adams (1973) and Bauer (1983) pointed out that it is again difficult to decide about
42
ADJECTIVES Noun Adjective Verb Adverb Particle
Adverb x now-pink x x x
Table 4
As opposed to compound nouns, there is much higher class variety of the right-
hand element and also left-hand elements functioning as adverbs were found here.
Therefore, doubts regarding adjectival status of compound can arise particularly when
the second element is not an adjective. That was the case with the majority of adjectival
compounds. This is contrary to nouns, where almost all of the formations contained
noun as their right-hand constituent. All the patterns depicted above will be now
This is the most productive adjectival pattern with nineteen examples found.
Huddleston and Pullum (2002, p. 1657) observed that many noun + adjective
compounds involve high degree of lexicalization and it is therefore not always clear
whether they are endocentric, or exocentric. The majority of compounds in this category
are written hyphenated and include present or past participle as their right-hand element
43
(1a)-(1b) with the latter being more usual. Compounds containing participles are both
glazed and so on. Some of the compounds, such as carefree and fat-free, seem to be
more ambiguous cases, but based on discussion provided by Bauer (2008, p. 8-9) they
category are exocentric, for example filter-tipped or life-affirming which are both
denoting people.
well, like lifelong. Also some of the compounds formed by the pattern (3) which are
including self- as their left-hand element are exocentric, concretely self-destructive and
self-taught. But also an endocentric example fitting into this pattern, self-explanatory,
can be found. Compounds beginning with self- are often referring to people but also to
This pattern was less common than the previous one and contained mostly endocentric
compounds (1). Once again, almost all the words were written hyphenated. Bauer
category, but no coordinate compounds are found. All of the formations in this group,
such as high-powered, middle-aged or good-looking, may also seem to fit into adverb +
44
adverbially or adjectivally. However, because they are adjectives in form, they are
classified into this group. Participles are highly common for these compounds and
almost all of the compounds found include them, with exception of extra-tough. Past
participle is much more frequent and only one compound, good-looking, contains
of the compounds share very similar structure. However, they can refer to both people,
which is much more common in the corpus, but also to inanimate things such as in
open-faced sandwich. Only one exocentric (2) example, dumbfounded, is found in the
corpus.
This pattern is not particularly productive with only four examples found. However, it is
still present as opposed to adverb + adjective subgroup included in Bauer (1983). Bauer
(1983, p. 210) noted that this type of adjectival compounds is most frequent with
participles which proves true in the corpus, as all the compounds found contain
overrated, upset and upbeat. The two latter examples are participial forms of phrasal
verbs and only the first one may be considered endocentric. As Adams (1973, p. 118)
pointed out, it is usual for particle compounds that over- and under- can often prefix to
the same stem and so underrated would be possible as well. Compounds with particle as
the right-hand constituent (2) are even rarer and only the one example mentioned above
is found.
45
3.4.4 Adverb + adjective
This is a highly marginal case, which is even not listed in Bauer (1983). It is represented
only by two examples in the corpus and although it shows similarities with the adjective
case is not-mine, where not could be theoretically regarded also as particle but
From now on, all the adjectival compounds to follow will be exocentric because they
endocentric. This pattern is the second most productive in the corpus with fourteen
examples occurring there. Bauer (1983, p. 210) remarked that a dispute whether these
compounds really are adjectives can occur. This is mainly because of conversion which
turns verbs or adjectives into nouns in many cases. This situation may be also seen as a
noun compound modifying another noun compound and based on this, it can be
misleading to label these compounds as adjectives. However, this would result into a
“three-term noun compounds” (Bauer, 1983, p.210), which would be written as two
words and so they would not be discussed here with respect to 3.1 and that would mean
a contradictory situation.
in the corpus will be labelled as compound adjectives. Compounds which function both
46
as nouns and their modifiers will be classified within nominal compounds.
Another argument against not treating them as noun compounds is also proposed by
Bauer (1983) who said that these compounds often involve lexicalization as units,
independent of their elements. Furthermore, they can be used attributively with different
connotation than the same forms used as compounds nouns in non-attributive position
(p. 206). The last problem regarding their classification is connected with their form, as
vast majority of cases. Because of this, the noun they modify will be provided in
following examples.
Examples of these compounds include boy-girl party, fleabag motel, tin-foil hat
or workshop production. If they were nouns, these examples would provide a whole
range of exocentric, subordinate and coordinate compounds but this is not the case
when dealing with adjectives. One compound contained linking element in form of
indefinite article, that being workaday lives. The example (2) above illustrates
compound made up of two identical stems, as in bang-bang music. Kavka and Štekauer
claimed that it is not very productive in English29. In the corpus, altogether two such
examples were found. The last subgroup of noun + noun compounds include
expressions with numerals (3) which were quite frequent in the corpus. This subgroup
can be exemplified by 32-ounce steak, 80-foot dog or two-inch dowels. All the numerals
found function as nouns and therefore they all fall into this subgroup.
29
For a study regarding this pattern of compounding, see Hohenhaus (2005).
47
3.4.6 Adjective + noun
This subgroup contains six compounds and so it is rather marginally productive pattern
in the corpus. Bauer (1983, p. 211) suggested that these formations are not compounds
when used attributively, but noun phrases. As opposed to the adjective + adjective
doubts regarding their classes arise. All of the compounds are exocentric in the sense
that they do not include an adjectival head, but most of them are semantically
endocentric (1). These include easy-pour, high-resolution or high-def. The last one is
also an instance of clipping, as def was clipped from definition. All the compounds
fitting into (1) are written hyphenated. One example which is used for denoting people
This subgroup is also found rather marginally in the corpus with only five examples
recorded. These compounds are all exocentric and are used for denoting people, such as
upbeat, as well as inanimate beings, like upstairs hallway or under-panty things. Most
of them are written as one word and one (upbeat) includes participle. Bauer (1983, p.
211) pointed out that prepositional phrases are often turned into modifiers in this
48
3.4.8 Verb + verb
This pattern is extremely marginal and only the example above is found. It can be
argued whether it really fits into this category, as putt can be both noun and verb.
Maybe also because of this, this pattern is not listed widely. It was not mentioned by
Adams (1973) at all. However, putt-putt course could be glossed as ‘a course where one
putts’ which indicates that putt is indeed better classified as a verb. It is also an instance
of stem reduplication mentioned in Kavka and Štekauer (2006) and in 3.4.5. This
compound is exocentric.
The last subgrouping of adjectival compounds found is the more frequent of the two
containing verbs, but also highly marginal with three examples. Thus, Bauer’s (1983,
p.211) observation regarding high productivity of this pattern was not confirmed in the
corpus. All of the compounds are exocentric and two occur written hyphenated while
one is spelled solid. Leftover also contains participle. Two of them are converted phrasal
verbs, namely back-up hypothesis and stand-up comedian. The last one, leftover,
includes adverbial particle left and so at least some variability is seen here which
49
ADJECTIVES Total Subordinate Coordinate Exocentric
N+Adj 20 13 0 7
Adj+Adj 8 7 0 1
Par+Adj 4 2 0 2
Adv+Adj 2 2 0 0
N+N 15 0 0 15
Adj+N 6 5 0 1
Par+N 5 0 0 5
V+V 1 0 0 1
V+Par 3 0 0 3
Total 63 28 0 35
Table 5
While there were no disputes regarding the status of nominal and adjectival
compounds, such can arise when dealing with compound verbs. As mentioned by
Kavka and Štekauer (2006, p.69), many linguists argue whether these verbal formations
are really formed by compounding. Adams (1973) discussed three ways which serve to
create verbal compounds – linking two lexemes together, back-formation from nominal
and adjectival compounds and conversion. Plag (2002) remarked that most of
compound verbs with verb as a head are result of conversion or back-formation (p.
197). This leads to a conclusion that verbal compounds created by putting two lexemes
together, which had been the prevalent process so far, are in fact in minority. Thus,
compounds in general are rather rare in the corpus and vast majority of them are
50
containing particles. Only three patterns used for creation of compound verbs were
found and despite the facts above, the same classification as for nouns and adjectives
will be maintained.
This is by far the most productive pattern of compound verbs and still, only five
examples were found. As noted by Bauer (1983), most of these compounds are
conversion are recognized among this pattern. All of these compounds are endocentric,
subordinate and written as one word. The most used particle was under- (1) with three
examples of underestimate, undergone and underline. Over- (2) and in- (3) both
occurred once.
This subgroup of verbal compounds is exceedingly rare in the corpus as it occurred only
once. It is also the only verbal compound not containing particles. Bauer (1983, p. 208)
mentioned that most of the compounds among this subgroup result from the process of
back-formation. However, the verb role-play is an instance of conversion from the noun
(1) And Ross, with his over-pronouncing every single word? (Friends, 3)
This pattern could be also perhaps labelled as particle + verb pattern, as the status of
pronouncing as gerund is not ultimately clear. However, Bauer (1983) tended to classify
51
gerunds as nouns30. Furthermore, pronouncing is used possessively in the sentence
above which is more characteristic of nouns than verbs. Nevertheless, this is the only
compound since it does not contain any verb, but semantically it can be felt as being
These observations are confirmed by the following table which suggests that
Par+V 5 5 0 0
N+V 2 1 1 0
Par+N 1 0 0 1
Total 7 5 1 1
Table 6
Adverbial compounds receive very scant attention in literature and they are also
very rare in the corpus. Only two patterns are recognized and just one out of six
compound adverbs does not contain particles. They are not classified by either Bauer
(1983) or Adams (1973) as well as many other linguists. Therefore, they will be
discussed very briefly, but the classification based on word classes used above will be
continued here as well. All the examples will be provided with context this time in order
30
See the discussion in 3.3.1.
52
Jack from downstairs? (Friends, 4)
Most of the adverbial compounds fall into this subgroup and all of them are exocentric
endocentric. This pattern is quite rare in the corpus. These compounds seem to be
formed quite freely and therefore it is sometimes difficult to distinguish them from free
phrases, as discussed by Adams (1973, p. 122). All of them refer to a certain place, but
can also give specific information about people, as in “Jack from downstairs”.
Only one example is found within this subgroup. Its form is the same as in noun +
but is exocentric in form due to absence of adverbial element. This pattern is extremely
marginal.
Par+N 5 0 0 5
N+Adj 1 0 0 1
Total 6 0 0 6
Table 7
Compounds with phrasal structure were already mentioned in 2.1.1 and 2.1.2
and unlike other formations on the borderline with phrases, they were included in this
53
analysis. As stated by Kavka and Štekauer (2006), these compounds make use of
conjunctions and prepositions, “conjugated verb forms and articles, that is to say, words
and word-forms typical of syntax” (p.70). Therefore, they can bring more variety into
written as well as oral utterances, in this case into the studied TV series. This was also
the main reason for their inclusion. Bauer (1983, p. 207) noted that these formations
above, these compounds will not be classified according to lexical classes of all their
elements, but only based on semantic criteria. Altogether, seven phrasal compounds
were found. This number is probably surprisingly low, but it is to be noted that only
compounds written hyphenated were taken into account. Some of them function as
nouns while the rest as adjectives. Therefore, they will be split into two groups.
These phrasal compounds clearly function as nouns but they are very rare in the
transcripts. The example (1) is exocentric, while the two examples in (2) are endocentric
and subordinate. It may seem rather strange, as their head is the initial element, but they
are treated as endocentric in Bauer (1983). He also stated that endocentric compounds
having final element as their head are far less established (p. 207) which seems to be
54
…that whole "You're-not-up-to-this" thing again? (Friends, 5)
This subgroup is just slightly more frequent with one example more. Similarly to noun
+ noun adjectival compounds, their status of adjectives can be disputed but the
conclusion here is the same as in 3.4.5. As seen in these examples, a large variety of
word classes is used in these compounds. All of them serve to modify nouns and are
exocentric.
section will not provide detailed classification of those compounds for two reasons. The
first one is that there has not been any consistent classification proposed by any of the
linguists cited so far. This is also due to the fact (which is simultaneously the second
reason) that they are not a clearly defined word-formation process and therefore behave
very ambiguously (Kavka and Štekauer, 2006, p. 66-67). Therefore, they will be
mentioned here rather for illustration. The basic division can be easily proposed
according to type of combining form(s) they are made of. Interestingly, only neo-
classical compounds with initial combining forms (ICF) were found in the corpus. Neo-
classical formations were quite common in the corpus and occurred as nouns, adjectives
and verbs.
compounds found, like sonogram, also include so-called medial vowel –o-which was
(2002, p. 1662). Most of the neo-classical compounds in the corpus are nouns. The most
55
photo- (photographs, photocell) and tele- (telephone, television). A few other examples
are also found, like cardio-funk, hemisphere, minivan or multivitamin. They often refer
Neo-classical adjectives are much rarer and each of them include a different
question may arise whether there is any difference of compound patterns and usage
between these TV series. And even further, whether the main characters show any
was initially separate and was merged into a common one afterwards. All the episodes
last about twenty minutes and therefore the series are comparable. Taken into account
were only the major characters for each series who appeared in all of the ten episodes.
They are Chandler Bing, Joey Tribbiani, Monica Geller Phoebe Buffay, Rachel Green
and Ross Geller for the Friends series. As for the BBT, major characters are Howard
further text, they will be referred to by using their first names only. The comparison will
be done in two steps. Firstly, the series will be compared generally disregarding the
From the numerical point of view, compounds are distributed almost equally in
both Friends and BBT, with the difference of compound count being really marginal.
31
Her last name was not mentioned so far in the series.
56
Still, BBT manages to produce slightly more compounds. It may seem that in BBT,
compounds are distributed more ‘symmetrical’ and in Friends they tend to occur
frequently in two or three subsequent utterances while not appearing for a long time
afterwards. However, this is very difficult to measure precisely, so it will have to remain
more of a feeling than a fact. Furthermore, compound classes found are the same for
both series. In both of them, compound nouns were highly dominant, especially the
noun + noun subgroup. Also the frequency occurrence of the other classes is
proportionally about the same in both Friends and BBT. Difference of quantity within
each compound class never exceeds four when the series are compared. Nonetheless,
BBT would have contained much more neoclassical compounds if they had not been
excluded, this being thanks to its scientific background. These observations are
Adjectives (Adj) 32 36 67
Verbs (V) 2 6 8
Adverbs (Adv) 2 4 6
Phrasal compounds 4 3 7
Table 8
them since it contains two patterns more. Nevertheless, they are generally represented
by only one or two compounds and therefore they have to be regarded as highly
marginal cases. Based on the facts above, it can be thus said that distribution of
compound words is rather similar in the analyzed TV series. As for the individual
57
words, the most frequently occurring compounds are boyfriend, not-mine and bathroom
in Friends and bedroom, hallway and hamburger in BBT. They also share considerable
similarity, as most of them are names of the rooms or indoor locations which are usually
mentioned outside the characters’ utterances in the notes. Very common are compounds
unexpected compound from the listing is probably not-mine which comes as a result of
its repeating several times in one sentence. This suggests that also the individual
their classes across the series, the same is not practicable with the characters, as the
individual characters is to be done for each series separately. Compounds which were
not used by the main characters listed above were discarded for purposes of this section.
In Friends, Chandler produces the most compounds (33) while Rachel is the
least productive with 16. Furthermore, Chandler also contributes to two of the patterns
which are exclusive for Friends. Compounds he utters are not present in only two of the
compound subtypes, which makes him also the most variable character in terms of use
mentioned above, Rachel utters the lowest number of compounds, but also two out of
four phrasal compounds found. Monica and Rachel use lowest number of adjectives,
which is the only category with noticeable differences. However, characters in Friends
58
specific patterns were found. In order to present these ideas more clearly, a table
Chandler 15 1 1 3 2 8 2 0 1
Joey 13 0 0 2 1 2 0 0 0
Monica 12 0 2 0 1 2 0 0 0
Phoebe 15 2 2 0 2 9 0 1 0
Rachel 10 0 0 0 2 1 0 1 2
Ross 16 0 5 0 1 8 0 0 0
Table 9
However, this is not the case with BBT. There, one dominant character can be
found, namely Sheldon with 81 compounds which covers more than a half of the total
number of compounds. Other characters are left behind, mainly Raj (10 compounds
found) and Howard (21). Sheldon outscores the rest of the characters in usage of
utters exocentric compounds such as corn-husker and is generally the central point of
many entertaining situations. However, Leonard is the only one to use big bang which
also appears in the title of the series. He also produces unusual compound adjectives as
flash-frozen banana or date-like component. Generally, Sheldon utters most of the rarer
This can be caused by a relatively low amount of episodes studied (10 for each series),
but it is certainly enough for a general image of the situation. These statements are once
59
BBT Nouns Adjectives Verbs Adverbs Phrases
Howard 11 0 4 0 0 5 1 1 0
Leonard 27 5 3 2 1 9 4 0 1
Penny 26 0 1 1 1 1 1 0 0
Raj 6 1 0 0 1 1 1 0 0
Sheldon 40 5 5 3 1 15 1 4 2
Table 10
3.10 Summary
above. In order to do this properly, a series of tables will be provided as well as brief
comments suggesting the general tendencies observed in the corpus. Most of them will
discuss the two TV series, Friends and BBT, in general while the patterns of individual
characters will be illustrated only briefly. Afterwards, compound usage of the individual
characters will be the matter of concern. In the tables, the number of compounds fitting
into the corresponding classes or patterns will be included instead of listing examples
which had been done throughout the analysis. Abbreviations provided in the brackets
60
Friends BBT Total
Adjectives (Adj) 32 36 67
Verbs (V) 2 6 8
Adverbs (Adv) 2 4 6
Phrasal compounds 4 3 7
very frequent. As also seen above, vast majority of compounds from the corpus are
nouns, followed by adjectives. The other classes are already marginal. Altogether,
phrasal compounds will not be further classified. Neo-classical compounds are not
included here because they function as either noun, adjective or verb, so they were
assigned to the corresponding classes. Therefore, both of them will be omitted from the
patterns. These tables would once again suggest that distribution of compounds is
highly equal across the series, so they will not be distinguished from now on and all the
compounds will be discussed together. Comparison of the series was done in 3.9.1.
61
NOUNS Total Subordinate Coordinate Exocentric
N+N 122 94 9 19
V+N 8 5 0 3
Adj+N 17 15 0 2
Par+N 11 7 0 4
V+Par 11 0 0 11
Among the nouns, most of the compounds fall into the noun + noun category.
Only here, endocentric coordinate compounds are found. All the other classes are
the nouns from verb + particle are considered exocentric, as they do not contain any
noun.
N+Adj 20 13 0 7
Adj+Adj 8 7 0 1
Par+Adj 4 2 0 2
Adv+Adj 2 2 0 0
N+N 15 0 0 15
Adj+N 6 5 0 1
Par+N 5 0 0 5
V+V 1 0 0 1
V+Par 3 0 0 3
Total 63 28 0 35
62
As for the adjectives, it is apparent that more compounding patterns are
recognized in the corpus in comparison to nouns. However, most of them are marginal.
occurrence of compound adjectives. As in Bauer (1983), all the adjectives not including
adjectival head were recognized exocentric, although some of them may be regarded as
the adjectives.
Par+V 5 5 0 0
N+V 2 1 1 0
Par+N 1 0 0 1
Total 7 5 1 1
Par+N 5 0 0 5
N+Adj 1 0 0 1
Total 6 0 0 6
marginal case of compounding in the corpus. They were neither productive, nor
included many patterns. Most of the verbal compounds are endocentric, with one
coordinate as well as exocentric example recognized. All of the adverbs are exocentric
The last two tables contain compound usage of the characters. The first one is
concerned with Friends, while the latter one with BBT. They are included mainly as a
63
matter of interest and cover only the initial division into classes as depicted in table 4.
However, due to large number of noun compounds, individual patterns are also
Chandler 15 1 1 3 2 8 2 0 1
Joey 13 0 0 2 1 2 0 0 0
Monica 12 0 2 0 1 2 0 0 0
Phoebe 15 2 2 0 2 9 0 1 0
Rachel 10 0 0 0 2 1 0 1 2
Ross 16 0 5 0 1 8 0 0 0
Howard 11 0 4 0 0 5 1 1 0
Leonard 27 5 3 2 1 9 4 0 1
Penny 26 0 1 1 1 1 1 0 0
Raj 6 1 0 0 1 1 1 0 0
Sheldon 40 5 5 3 1 15 1 4 2
negligible, as all the not marginal patterns are used similarly by all of them. The only
compounds. As for BBT, which is depicted in table 10, compound usage differs much
more. In general, Sheldon uttered vast majority of compounds, while Howard and Raj
64
used only a small amount of compounds. As seen above, the scale is much wider in BBT
but the number of compounds is proportionally similar for the characters in all patterns
in Friends as well as BBT. Therefore, it can be said that no tendencies for characters to
65
Conclusion
also appeared to be the case in a corpus which was created for the purposes of this
thesis. The corpus was created using transcripts of two famous American TV series.
They are Friends and The Big Bang Theory. The first ten episodes of each series were
taken into account, which proved to be a satisfactory sample. Not only compounding in
general was taken into account but also compound usage of individual characters of
The main goal of this thesis was to classify the compound words found in the
corpus. For this purpose, Laurie Bauer’s classification as presented in his book English
them. The analysis also considers the status of compound’s head, which is inspired by
In order to back the analysis with facts, two theoretical chapters were included at
the beginning of the thesis. The first one discussed the basic notions of English word-
formation and also the definition of word-formation. The next chapter was focused on
although not much space was devoted to the definition of compounding. The main focus
of this chapter was the problem of unclear distinction between compounds and phrases
which causes various kinds of problems in English. Therefore, multiple criteria for
differentiating compound from phrases were presented. The second chapter also
66
introduced the basic types of compounds, which are then used in the analysis. On that
account, the terms endocentric, exocentric and copulative compounds were defined.
Endocentric compounds were further split into coordinate and subordinate according to
their head and these were the main compound properties discussed in the analysis.
mentioned as well. They were affixation, conversion, clipping, blending and back-
formation. All of them were represented in the analysis, as borderlines between the
The third chapter was concerned with analysis of the corpus. In the beginning,
the criteria for classification of compounds were described. This was followed by
clarification regarding borderline cases of compounding and all the compound types
that were not taken into account were emphasized. Afterwards, all the categories and
compounding patterns found were presented. The most productive and frequent patterns
received the biggest attention while the more marginal cases were commented only
briefly. The analysis was accompanied by tables for better orientation. Also two special
types of compounds, phrasal and neo-classical compounds, were recognized but not
corpus. The latter two were rather marginal cases and usually contained particles. The
vast majority of compounds were nouns and therefore the greatest space is devoted to
them. Adjectives were significantly rarer, but included a wider range of compounding
patterns. That was because almost every nominal compound included a noun as the
right-hand element. Most of the compounds were endocentric and subordinate, while
exocentric compounds occurred mainly among adjectives. Some of the compounds were
67
semantically endocentric, but formally exocentric and therefore considered exocentric in
the end.
The two TV series were compared as well as language of their main characters.
However, no significant differences were found. All the findings were then summarized
in tables which illustrated productivity and frequency of all compounding patterns and
68
Bibliography
Primary Sources
Friends:
http://home.versatel.nl/friendspic0102/transcripts/html/0101.html
The One Where Nana Dies Twice. [Television program] [Transcript]. Retrieved
http://home.versatel.nl/friendspic0102/transcripts/html/0108.html
http://home.versatel.nl/friendspic0102/transcripts/html/0109.html
http://home.versatel.nl/friendspic0102/transcripts/html/0104.html
http://home.versatel.nl/friendspic0102/transcripts/html/0107.html
The One With The Butt. [Television program] [Transcript]. Retrieved 9 October,
The One With The East German Laundry Detergent. [Television program]
http://home.versatel.nl/friendspic0102/transcripts/html/0105.html
69
The One With The Monkey. [Television program] [Transcript]. Retrieved 17
http://home.versatel.nl/friendspic0102/transcripts/html/0110.html
The One With The Sonogram At The End. [Television program] [Transcript].
http://home.versatel.nl/friendspic0102/transcripts/html/0102.html
http://home.versatel.nl/friendspic0102/transcripts/html/0103.html
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8537
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8544
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8542
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8538
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8543
70
The Hamburger Postulate. [Television program] [Transcript]. Retrieved 14
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8540
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8545
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8539
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8541
http://transcripts.foreverdreaming.org/viewtopic.php?f=159&t=8536
Secondary Sources
Longman.
Backstory. (n.d.). Online Etymology Dictionary. Retrieved March 13, 2015, from
71
http://www.victoria.ac.nz/lals/about/staff/publications/Bauer-
EnglishExocentricCompounds.pdf
Press.
Heine, B., & Narrog, H. (2010). The Oxford Handbook of Linguistic Analysis. Oxford:
Huddleston, R. D., & Pullum, G. K. (2002). The Cambridge Grammar of the English
Lieber, R., & Štekauer, P. (2009). The Oxford Handbook of Compounding. Oxford:
72
Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A Comprehensive Grammar
http://www2.uni-siegen.de/~engspra/plag-in-press.pdf
Stockwell, R. P., & Minkova, D. (2001). English Words: History and Structure.
Štekauer, P., Valera, S., & Körtvélyessy, L. (2012). Word-formation in the World's
Vachek, J. (1976). Selected Writings in English and General Linguistics. The Hague:
Mouton.
73
Resumé in English
series. These are Friends and The Big Bang Theory. They have become famous all over
the world and have a large fan base. Therefore, numerous transcripts of these series can
be found on the internet. Transcripts of the first ten episodes of each series are analysed
and serve as a basis for this thesis. The transcripts are used to create a corpus, which is
productivity and diversity. The main aim is to classify compound words found in the
corpus.
The first two chapters of this thesis are mostly theoretical. They provide
definitions not only of word-formation and compounding, but also other terms which
are essential for understanding the following text. The greatest attention is devoted to
compounding and different types of compound words. The theoretical chapters are
The third chapter is the main part of this thesis. The chapter is concerned with
analysis of the corpus and classification of compound words. The classification was
done according to multiple criteria. The main criteria are word classes of compound’s
elements and semantic relationships between them. All the criteria are explained at the
beginning of the analysis. Mainly compounds consisting of two elements are discussed.
However, compounds having three or more constituents are briefly mentioned as well as
compounds including elements of Greek or Latin origin. This chapter also examines
utterances of the main characters and compares their manner of using compound words.
Tables are often included to illustrate all the observations in a clear way.
74
Resumé in Czech
televizních seriálech. Těmi jsou Přátelé a Teorie velkého třesku. Tyto seriály se staly
internetu možno nalézt mnoho přepisů jednotlivých dílů. Analyzovány byly přepisy
prvních deseti epizod každého seriálu, přepisy zároveň slouží jako základ pro tuto práci,
a tak byly použity k vytvoření korpusu, který je k práci přiložen na CD. Za hlavní téma
této práce bylo vybráno skládání slov, a to především kvůli jeho produktivitě a
První dvě kapitoly této práce mají především teoretický ráz. Obsahují definice
nejen slovotvorby či skládání slov, ale také jiných pojmů, které jsou nezbytné pro
ostatních anglických slovotvorných procesů, které jsou často úzce spjaty se skládáním
slov.
Třetí kapitola tvoří hlavní část této práce. Zabývá se analýzou korpusu a
Hlavními kritérii jsou slovní druhy částí složených slov a jejich vzájemné sémantické
vztahy. Všechna tato kritéria jsou popsána na začátku analýzy. Zmiňována jsou
především složená slova skládající se ze dvou částí. Opomenuta však nejsou ani slova
skládající se z tří a více prvků, ani jako složená slova obsahující části řeckého či
latinského původu. Tato kapitola také zkoumá výroky hlavních postav obou seriálů a
porovnává, jakým způsobem tyto postavy užívají složených slov. Pro ilustraci jsou
75