Chapter 1 Introduction
1.4 Research Questions and Hypotheses
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
in Chungli City than in Taoyuan City.
1.3.3 Socio-economic characteristics of the two cities
Taoyuan City is the administration center of Taoyuan County and a city of commerce, while Chungli City is a city of manufacturing industry. In comparison, the former is more like a cosmopolitan city, in which more Mandarin and less Hakka are used; and the latter is more like a town, in which Hakka is expected to be used more frequently.
In general, the two cities investigated in this study are different from each other in geographical area, population size, and socio-economic characteristics.
1.4 Research Questions and Hypotheses
To examine the target phonetic variation, the following research questions are proposed.
(1) Is this phonetic variation of /ɕ/ related to the phonetic environments, neighboring sounds, and syllable structure?
(2) Is this phonetic variation of /ɕ/ related to lexical frequency?
(3) Does this phonetic variation expand through lexical diffusion?
(4) Is this phonetic variation of /ɕ/related to situational formality?
(5) Is this phonetic variation of /ɕ/ related to geographical characteristics?
(6) Is this phonetic variation of /ɕ/ related to Hakka speakers’ social characteristics (specifically, gender, age, and education level)
Responding to the research questions given above, the following hypotheses are proposed.
(1) Phonetic environments (including syllable structures and neighboring sounds) would affect the surface realization of /ɕ/.
(2) Lexical frequency is related to the surface realization of /ɕ/, with characters of
5
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
high frequency yielding more nonstandard form [s].
(3) Situational formality is influential to the surface realization of /ɕ/ to [s], with the use of [s] increasing as the formality decreasing, and vice versa.
(4) Geographical characteristics are effective to the surface realization of /ɕ/ to [s], with people in Chungli City using more [s] than those in Taoyuan City.
(5) Social factors (including, gender, age, and education level) are effective in the surface realization of /ɕ/ to [s], with Hakka people of older age, of male, of lower education level using more [s]’s than their counter groups.
6
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Chapter Two Literature Review
In this chapter, theories and previous studies related to the target linguistic variation of this study are reviewed.
2.1 Theories of Linguistic Variation and Change 2.1.1 The formalism’s point of view
The Neogrammarian school proposes that language change is regular, and sound change is limited to a particular speech community at a particular time. Focusing on sound change, the Neogrammarians assume that it is observable that a given time in a given region, all the words which have a certain phonetic feature are subject to the same change. Verner’s law, illustrates a historical sound change in the Proto-Germanic language, a good illustration of such assumption; however, since it is hard to find extremely regular sound change, the school excludes sporadic changes, although Verner (1877) claims that there must be a rule within the irregulatities.
Structuralists propose that linguistic features can be described in terms of structures and systems. Originated from a Swiss linguist, Ferdinand de Saussure, in the early twentieth century, the following ideas came into every school of linguistics.
First, langue and parole are differentiated. According to Saussure (1916), parole, which includes spoken and written language, is concrete; however, langue, which is underlying the surface structure, is abstract. To Saussure, langue is more important than parole. Second, Saussure takes synchronic study as more critical than diachronic.
In synchronic study, one describes the state of the language, disregarding whatever
7
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
changes might be taking place; while in diachronic study, a language is studied from the point of view of its historical development. Third, Saussure distinguishes
syntagmatic and paradigmatic relations, with the former referring to the relationships between constituents in a construction, and the latter being defined as a class of associative relations. Fourth, the theory considers phonemes as the central units of a sound system. It was claimed that native speakers would be more aware of those contrastive phonemes than those that are not.
Structuralists did propose different views of sound change. They tried to explain the sound shifts from the structure of systems and the structural functions of language.
However, the theory still can not fully explain how and why language change happens.
Noam Chomsky, a representative of the generativists, in his Syntactic Structures (1957), proposes that all languages are rule-governed; therefore, human beings are able to learn these rules in finite number to produce infinite number of sentences.
According to Chomsky, it is such linguistic competence that would allow human to encode and to decode a message. In comparison with Structuralists, Generativists put more emphasis on linguistic data introspected, while Structuralist focused on actual linguistic data. Moreover, Generativists focus on underlying system of the rules; while Structuralists emphasize surface structure. Further, Generativists believe that all sound changes occur in grammar. Postal (1968:270), in describing the core idea of the
Generativist viewpoint about sound change, indicates that in early time, Generativists think that all changes must be simplificatory, translating more complex to simpler and more economical grammars. McMahon (1994) in his book Understanding Language
Change, points out some deficits of the Generativist’s theory. First, early Generativists
did not consider the effects of change on the system of sounds in a language. Instead,
8
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the Generativist give formal phonological rules, which reflect completed changes, but are not explanatory. Second, not all changes could be construed as simplificatory.
Third, there is no entire idea about what would constitute simplification, a concept that is relative rather than absolute. Fourth, the characterization of all changes as simplificatory presumes a view of change as constantly creating ever simpler
grammars. Unfortunately, there is no evidence to prove that languages, following this view point, must have been getting simpler gradually ever since language began.
2.1.2 Lexical diffusion theory
Linguistic variation is not instantaneous. Wang (1969, 1977) propose a theory of lexical diffusion, indicating that a sound change would spread gradually through lexical items carrying the target sound related to the sound change. In other words, a sound change may apply to some words initially, moves on to other words, and then still others, until that sound change is accomplished. Also, according to Bailey (1973), linguistic variation and change does not proceed across the affected lexical items at a uniform rate. To be specific, the initial and the ending stages of a change are slow, but in the intermediate stage, the change is rapid. Chambers and Trudgill (1998, as cited in Wardaugh 2002: 223), in their study in East Anglia and the East Midland of
England, find that the same sound change is complete in must and come, but hardly at all in uncle and hundred. Likewise, Hansen (2001, as cited in Wardaugh 2002: 225) supports lexical diffused sound change of French nasal vowels.
Phonetic variation may be evoked by linguistic factors, such as syllable structure and word frequency. It has been noticed that not all lexical items containing a target sound involved in a certain phonetic variation would experience the variation at the same time, in the same rate. Most scholars considered that the most frequent words are easier to comply to sound change. Schuchardt (1928) claims that the frequency in
9
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
using individual words that plays such a prominent role in analogical formation is also of great importance for their phonetic transformation. Therefore, it is said that “Rarely used words drag behind; very frequently used ones hurry ahead.’’ The view point was supported by Leslau (1969), and Fidelholz (1975). However, Hooper (1976) found that it is infrequent verbs, not frequent ones, that are the most possible ones to change to the regular –ed paradigm in English. Similarly, Phillips (1981), in his research of glide deletion in Southern American English words, finds that high frequency words would not necessarily induce higher percentage of sound change.
In this study, lexical diffusion theory is applied to examine the scope and the rate of the diffusion of the nonstandard variant [s] of Mandarin /ɕ/.
2.1.3 Sociolinguistic variation
One of the most important tasks in the field of sociolinguistics is the study of linguistic variation and change within social context. Although Labov (1994) admits that lexical diffusion is a fundamental mechanism of sound change, he still claims that internal variation and social motivations among speakers are the most important factors in linguistic variation and change. More detailed reviews are given below.
2.2 Linguistic Variation within Context 2.2.1 Speech community
In the realm of sociolinguistics, the criterion for a speech community is far beyond just sharing a linguistic code system. The definition of speech community is different researchers. Hudson (1996: 29) indicates that “our sociolinguistic world is not organized in terms of objective ‘speech communities,’ even though we like to think subjectively in terms of communities or social types such as “Londoner” and
“American.” In Hudson’s opinion, there is no clear definition of speech community.
Lyons (1970: 362) considered speech community as the congregation of the people
10
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
who use a given language (or dialect).
Giles, Scherer, and Taylor (1979:351) consider speech markers as important devices to divide social categories such as age, sex, ethnicity, social class, and situation. Although there are other ways to discriminate social categories, speech markers are still critical to social organization. According to Labov (1972: 120-1), a speech community is not defined by any marked agreement in the use of language elements, so much as by participation in a set of shared norms; these norms may be observed in overt types of evaluative behavior and by the uniformity of abstract patterns of variation which are invariant in respect to particular levels of usage.
Labov’s viewpoint to speech community focused on individuals’ feelings of being members in the same community. Gumperz (1971: 101), using the concept of linguistic community, instead of speech community, suggests that a speech
community can be either monolingual or multilingual, and speech community have to be independent in order to compare to other communities. Gumperz (1971: 114) provides another explanation of speech community, indicating that a community and the related language systems are not two independent entities, and that members in a speech community would have something in common, including grammatical rules, language use, and social norms. Following the definition given above, Taoyuan City and Chungli City are presumed to be two speech communities, with each having its own sociolinguistic norms for their choices between [ɕ] and [s].
2.2.2 Language contact
Very often people in the same geographical or social space speak more than one language (or dialect). Language contact is, hence, induced. According to Weinreich (1953), languages in contact would interfere each other: “Those instances of deviation from the norms of either language which occur in the speech of bilinguals as a result
11
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
of their familiarity with more than one language, i.e. as a result of language contact, will be referred to as interference phenomena." (Weinreich, 1953:56) For example, some Dutch would pronounce ‘finger’ /ˈfɪŋgə/ as [ˈfɪŋə] under the influence of the word ‘Vinger’ /ˈv̥ɪŋəʀ/ from Dutch.
Weinreich (1953) also suggests that social factors should be taken into consideration, so that the situation of language contact could clearly be observed.
Thomason and Kaufman (1988) include social factors in the theory, and they propose the importance of distinguishing borrowing and substratum interference, with the former referring to “ the incorporation of foreign elements into the speakers’ native language,” and the latter referring to native language structures’ affecting the second language (1988: 21).
Tse (1983) points out that most speakers in Taiwan are bilingual of Mandarin Chinese and Southern Min dialect. Hakka people in Taiwan are even forced to be multilinguals of Mandarin, Southern Min dialect, and Hakka dialect, with Mandarin being used in formal occasions, and Southern-Min and Hakka in less formal or casual occasions. In this study, it is curious that the emergence of the phonetic variant [s] of Mandarin /ɕ/ by Hakka people is a consequence of language contact with Hakka being the source language and Mandarin being the target language.
2.2.3 External factors of linguistic variation 2.2.3.1 linguistic variation by geography
Space is an essential element in analyzing linguistic variation and change. It is expected that different geographical regions would have dialectal differences.
Schmidt (1872), proposes “wave model” to explain linguistic variation, indicating that linguistic variation initiated in a region would spread to the surrounding areas, and the effect of the spreading would decrease with the increase of the distance.
Britain (2002) points out that geographical space is not the only component of
12
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the concept of space; instead, there were three types of space: (1) euclidean space, i.e.
the objective, geometric, socially divorced space of mathematics and physics), (2) social space, i.e. the space shaped by social organization and human agency, by the human manipulation of the landspace, by the contextualization of face-to-face interaction, by the creation of a built environment, and by the relationship of these to the way the state spatially organizes and controls at a political level), and (3)
perceived space (how civil society perceives it’s immediate and not so immediate environments-important given the way people’s environmental perceptions and attitudes construct and are constructed by everyday practice. Also, these various types of space do not exist independently. Therefore, when analyzing geographical variation, the social aspect of a linguistic variation should also be taken into consideration.
2.2.3.2 linguistic variation by formality
In stylistic variation, the influence of situational formality is examined. Labov (1972) proposes that speech styles can be categorized into careful speech and casual speech. In careful speech, speakers usually use more standard forms and less
nonstandard forms, while in casual speech, speaker usually use more nonstandard forms and less standard forms. In his study (1972), Labov designs different tools to measure situational formality by using conversation, reading passages, word list, and a list of minimal pairs, with formality decreasing accordingly. Labov (1966), in his study of the stylistic and the social variation between [t] and [θ] as variants of /θ/ in New York City, finds that the use of the nonstandard form [t] decreases with the increase of stylistic formality. In another study of Labov’s (1966), the influences of formality and class on the use of postvocalic /r/ are verified. Bell (1984) says,
“Variation on the style dimension within the speech of a single speaker derives from and echoes the variation which exists between speakers on the “social” dimension.”
Following the same line, Chan (1984), in her study of Mandarin /ʐ/ in Taiwan by
13
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Southern-Min speakers, finds that her subjects, in situation of higher formality (i.e. in reading passage, reading characters, and reading minimal pairs), use more standard form [ʐ] and less non-standard forms [l] and [z]; but they use more nonstandard forms and less standard form in casual conversation. Based on the findings of the studies given above, situational formality is included in this study.
2.2.3.3 linguistic variation by social space
As mentioned above in section 2.2.3.1, social space is an important factor to linguistic variation. A good example for the study of linguistic variation in social context is Labov’s (1963) investigation of vowel centralization of two diphthongs, /au/ and /a
ɪ
/, in words such as mouse and mice, in Martha’s Vineyard. Results of data analyses indicate that local people, tended to pronounce /a/ and /au/ as [əɪ] and [əu], in order to signal localism and loyalty to the island. In other words, this linguistic variation of vowel centralization is initiated and spread on the island for social causes.Bailey (1993) uses random telephone survey to examine the diffusion of the de-rounding of [
ɔ]
into [ɑ] in the center of Oklahoma City before 1945 and that after 1945. It is reported that the de-rounding process diffuses from big city to small city, from large town to small town. Also, Bailey concludes that different types of diffusion may aim at different social meanings that linguistic forms carry. However, opposite to Bailey’s finding, Trudgill (1974) found contra-diffusion in his study in East Anglia’s research, with phonetic variation shifting from rural north Norfolk to Suffolk, the city center. These findings mentioned above that the path of linguistic variation and change travels in all directions.People have been socio-economically categorized into different social groups. In 19 century, Karl Marx proposed a theory of social class. Warner (1941), in his study of the New England City of Yankee City, suggests that social factors other than income and occupation (including educational background, family background, and
14
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
so on) should also be taken into consideration, and he (1960) built Evaluated Participation and the Index of Status Characteristics to measure social class. Labov (1972), in his study of the use of postvocalic /r/ in three department stores in New York City, reports that subjects of higher class and those who assign themselves to higher class use more postvocalic/r/, the prestigious form, than subjects of lower classes.
Trudgill’s (1974) study of the Norwich England supports Labov’s point of view.
In Trudgill’s study The pronunciation of –ing as [ɪŋ] or [ɪn] is found to be related to the speakers’ social class. In his study, Trudgill makes use of occupation, income, education and other social factors to categorize people into 5 social classes, namely, middle middle class, lower middle class, upper working, middle working, and lower working class. The results of data analyses indicate that the lower classes tend to give higher percentages of using [ɪn] than the higher classes.
Labov (1966) proposes “change-from-above” and “change–from–below” to explain linguistic variation, with change-from-above being lead by the prestigious class (not necessarily the highest one), while change-from-below being evoked and lead by the working class. The use of postvocalic /r/ in New York City is an example of change-from-above; whereas the vowel centralization in Martha’s Vineyard is an instance of change-from-below. In this thesis, the phonetic variation of Mandarin /ɕ/
between [ɕ] and [s] may probably be affected by social class, which is represented by informants’ educational background.
Gender is one of the major factors that would determine people’s behaviors of various kinds. Cheshire (2002) defines the concept of sex as biologically determined, and the concept of gender as social-culturally identified. Labov (1990) lists three principles of gender differences: (1) Principle I: In stable sociolinguistic stratification, men use a higher frequency of nonstandard forms than women; (2) Principle Ia: In
15
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
change from above, women favor the incoming prestige forms more than men; and (3) Principle II: In change from below, women are most often the innovators.Levine and Crockett (1966), in their study of the postvocalic /r/ in southern states of United States, find that women, young people, the newer residents, and people of higher social status take the national r-norm as their speech model; while the linguistic behavior of males, older people, long term residents, and blue-collar respondents were found to comply to the southern prestige norm, namely, the r-less pronunciation of the coastal plain, Anshen (1969), in his cross gender comparison of the use of postvocalic /r/, finds a similar results that women tend to use more prestigeous form than men. Following the same line, Chan (1984) finds that female native speakers of Southern Min dialect use more prestigious forms of Mandarin /ʐ / than males do.
Nevertheless, there are counter examples. For instance, in Labov’s (1963) study of vowel centralization in Martha’s Vineyard, young men, who sought to identify themselves as native Vineyarders, reject the values and the speech style of the
mainlanders, and use the local prestigious forms [əu] and [əɪ] more often than women.
Also, Eckert (2012) points out that gender, like any other social factors, doesn’t work independently; instead, the significance of gender in linguistic variation has to do with how gender structures people’s lives at different places in society. In this thesis, gender is examined individually as well as in combination with other social factors, for its influence on the phonetic realization of Mandarin /ɕ/.
Linguistic variation in progress prescribes time. However, since it is hard to collect data diachronically, using data of apparent time becomes a way to solve the problem. Labov’s (1963) study of Martha’s Vineyard makes use of data of apparent time to observe the centralization of /aw/ and /ay/. With his subjects being divided into five different age groups, it is found that the odlest group shows the lowest rate of
Linguistic variation in progress prescribes time. However, since it is hard to collect data diachronically, using data of apparent time becomes a way to solve the problem. Labov’s (1963) study of Martha’s Vineyard makes use of data of apparent time to observe the centralization of /aw/ and /ay/. With his subjects being divided into five different age groups, it is found that the odlest group shows the lowest rate of