- Shopping Bag ( 0 items )
Creativity and innovation appear in many different guises and on various levels of language, including sound, form, meaning and use. Instead of asking for Cheerios or cereal for breakfast, for example, my daughter blended them together and asked for Cheerial; when talking to some friends from a different region in the United States, my husband asked y'all want to join us? (rather than his own typical form you want to join us?); when describing a person who lived in our neighborhood, my son once coined the term back door neighbor to complement the term next door neighbor.
Linguistic creativity and innovation abound (even outside of my own immediate family!). For example, a speaker may know exactly about whom s/he is thinking when beginning a story about a specific person. But s/he may need to create a way to describe that person to an addressee that is more informative than the pronoun she, e.g. through a descriptive clause such as she- y'know that woman that I met when I went with Laura, last weekend, to that festival at Glen Echo? that actually tells a mini-story. And although we all have routine ways of asking for the salt (Can you pass the salt? or Salt, please), we may also vary our requests by saying This food is really bland or Are we out of salt? Likewise, the invitation Care to dance? - an utterance used as an access ritual (Goffman 1971a) that is part of the register of a particular social occasion - can be addressed to a woman (me) who accidentally bumped into a male stranger when she turned around too quickly in a checkout line at a busy shop in an international airport.
Public discourse also provides resources for creativity. Culturally familiar sentences (e.g. John F. Kennedy's Ask not what your country can do for you, but what you can do for your country) and lines from favorite books (Tolkien's All who wander are not lost) appear in high school term papers and college applications. And of course even single words travel to new public locations, as when the organization People for the Ethical Treatment of Animals protests the unethical treatment of animals by labeling a chicken dinner Holocaust on a plate on their web site and in print advertisements.
Despite the potential for wide ranging creativity in language, such creativity does not spread completely unfettered by restrictions. Numerous limitations arise simply because our sentences follow the implicit rules of our grammar. Although the sentence She wants to do it herself seems fine to speakers of standard American English, the sentence Herself wants to do it does not. And of course the innovative examples above actually follow regular linguistic patterns: Cheerials conforms to the syllable structure and stress pattern of Cheerios; it also reflects a semantic relationship of hyponymy ('Cheerios' are a type of cereal). Other restrictions are less formally grounded and may stem from our inability to clearly formulate the propositions that convey what we know (e.g. if we are trying to explain a complex equation), to state what we think (e.g. if we are trying to make a decision about something that we feel ambivalent about) or report what we feel (e.g. if we are still in emotional anguish forty years after a traumatic experience that we are recounting during an interview). Even if we may be perfectly able to access our knowledge, thoughts, and feelings, we may nevertheless find it difficult to verbalize them eloquently, in an appropriate manner or style, or in a way that fits the needs of our recipients or the demands of a situation. Although we are constantly speaking in innovative and creative ways, then, we are also limited as to what we are able to put into words, how we may do so, to whom, when, and where these new combinations and arrangements should appear.
The tension between innovation and restriction is partially reminiscent of, but also quite different from, two other oppositions inherent in our use of language to organize our thoughts, convey our intentions and manage our lives. Illustrated in Example are dichotomies between same and different, new and old:
|(a)||There is another person whose name is Deborah Schiffrin.|
|(b)||Although she has the same name as me, she spells it differently.|
The opposition between same and different is illustrated by the content of line (a): the same name is used for two different people. What makes this linguistically relevant is that names are rigid designators: they denote the same individual regardless of context.1 Names thus contrast with other ways of evoking people, such as titles and common nouns. There can be more than one person referred to as the Dean, and addressed as Dean even within a single institution. At Georgetown University, for example, there is a Dean of the College, the Graduate School of Arts and Sciences, Admissions, School of Foreign Service (and so on) at any one time. Likewise, occupants of these offices change over time. Or take common nouns. If I want to talk about 'a child,' I can either talk about a generic child or specify only one 'child' of the many children in the world.
The sentences in Example also illustrate the opposition between old and new. Once a referent has entered into the discourse, its information status changes: it is no longer new and we can use different words to evoke it. Thus, once 'another person whose name is Deborah Schiffrin' has been introduced in Example , line (a), I don't need to repeat it in all its detail in Example , line (b): instead, I can use she to evoke the old referent. Word order also reflects information status. In Example (a), 'someone shares my name' is new information: it appears at the end of the sentence after the semantically weak predicate there is. The alternative information order is awkward. Another person whose name is Deborah Schiffrin exists seems appropriate only if I am announcing something (e.g. Guess what!) or someone has questioned the issue (Are you sure?). Once line (a) has been presented, however, the information about a second 'Deborah Schiffrin' is no longer new and can become a sentence initial adverbial clause Although she has the same name as me, line (b). And then what is new information - the spelling of the names - can appear at the end of the sentence.
Not all of the oppositions - innovative/fixed, different/same, new/old - that characterize our use of language have been studied by linguists. One sort of difference that clearly matters for linguists is deciding whether a phonetic difference is associated with a different word meaning. If an unvoiced alveolar stop [t] is aspirated (ends with a puff of air) in Thai or Hindi, for example, it conveys a different word meaning than the same sequence of sounds with an unaspirated [t]. Not so in English: saying Sit! with an aspirated [t] would sound emphatic and perhaps angry, but still mean that I want my addressee (whether person or pet) to occupy a certain position in a chair or on the floor.
Of those oppositions that are pertinent to the systematic study of language, not all are equally interesting to the same linguists. Analyzing the role of repetitions, paraphrases and parallelisms in spoken discourse is interesting to linguists who study coherence (Becker 1984, Johnstone 1994, Tannen 1989) and intertextuality (Hamilton 1996), but perhaps less interesting to those who study reduplications, a form of morphological repetition common in pidgin and creole languages. Analyzing the organization of categories as prototypes (Rosch 1973, 1978, Taylor 1989) or radial categories (Lakoff 1987) might be interesting to cognitive linguists who study lexical meaning, but not to formal semanticists who study truth functional meaning. Other linguistic differences may be interesting to a variety of language researchers, but for quite different reasons. The analysis of speech errors, for example, interests psycholinguists because they can provide evidence for a particular model of language processing or production (e.g. Levelt 1983, 1999) or conversation analysts because of their role in the interactive construction of turns at talk (Fox and Jasperson 1995, Schegloff 1987).
One way that a subset of these oppositions - different ways of saying the 'same' thing - has been studied in Linguistics is through what sociolinguists have called variation analysis. After discussing this approach in Section , I turn to an overview of two aspects of language use on which this book will focus: reference, referrals to a person, place, or thing through a referring expression (Section ); narratives, sequences of temporally ordered clauses that cluster together to report 'what happened' (Section ). Each chapter (to be previewed in Section ) addresses some aspect of variation that arises when the referral or the narrative recurs in a 'second position' in discourse.
1.2 Variation analysis: 'same' vs. 'different'
One of the main functions of language is to provide information: language is used to convey information about entities (e.g. people, objects), over time and across space, as well as their attributes, states and actions (when applicable), and relationships. The terms used to convey this function vary: denotational, representational, propositional, or ideational. Yet language clearly has more than a referential function. In figure 1.1 I have adapted Jakobson's (1960) framework, which includes not only a referential function, but also five other functions defined by the relationship between utterances and facets of the speech situation.
|Image not available in HTML version|
Figure 1.1 Speech functions
The situational component is in upper case; the function is bracketed; I italicize terms that I use interchangeably with Jakobson's terms.
In addition to grounding the functions of language in the speech situation, Jakobson also makes another critical point: although an utterance may have a primary function, it is unlikely that it has only one function. Do you know the time?, for example, may have a phatic function (it opens contact), an emotive/expressive function (it conveys a need of the addressor), a conative/recipient-design function (it asks something of the addressee in a specific way), and a referential function (it makes reference to the world outside of language).
Despite the array of different functions that utterances serve, 'same' in Linguistics is usually understood as referential sameness: sounds, morphemes, words, sentences, propositions, and texts are the 'same' if they contribute to a representation of the same thing in a world. Suppose I am telling a story about how 'my dog' took (and then hid) most of the Halloween candy, and refer to her as Lizzy, she, a bad dog, and that clever girl. These expressions can certainly refer to different entities in other contexts. And even in my Halloween story, the different expressions convey different attributes of the 'dog' and different attitudes of the 'speaker.' But they are all ways of referring to 'my dog' and the referent that they index would be the same throughout the story.
The referential function of language has always had a central role in the study of sound (phonology) and form (morphology, syntax). For many scholars interested in meaning, however, restricting semantics to the study of relatively stable referential meanings, especially those that can be formally mapped as the conditions under which a proposition would be true, ignores other sources and types of meaning. One way that semantics is thus supplemented is through the subfield of pragmatics and its concern with contextually dependent non-referential meanings.
The semantics/pragmatics distinction (at least in so far as we are thinking of truth-functional semantics) provides a paradigmatic example of the same-different dichotomy. In the perspective known as Gricean pragmatics, for example, referential (truth-functional) meaning can remain constant, but a speaker's adherence to, or manipulation of, the maxims of Quantity, Quality, Relation or Manner (grouped together as the Cooperative Principle (Grice 1975)) can add additional communicative meanings. Thus the conjunction and may very well maintain its truth-functional meaning ('if P is true, and Q is true, then it is also true that P & Q is true') each and every time it is used. In a sentence like Example , line (a), however, there may be an added inference of temporal order (stemming from a maxim of Quantity, Manner or Relation), but not necessarily in Example , line (b):
|(a)||I got really busy on Sunday. I cleaned the closets and prepared things to give to charity.|
|(b)||I got a lot of exercise on Sunday. I ran a mile and swam twenty laps.|
To take another example, in Example either line (a) or line (b) may be used to convey to a spouse a plan to get young children ice cream. But it is only line (b) that would implicate (because of a violation of Quantity, Manner or Relation) the speaker's intention of surprising the young children by the upcoming treat (Levinson 1983: Chapter 3):
Let's go get some ice cream.
Let's go get some I-C-E-C-R-E-A-M.
Or consider the utterance It's cold in here, an example of an indirect speech act. Although the proposition conveyed in the sentence has the same referential meaning regardless of its context of utterance, a hearer would be remiss if - when hearing the utterance - he did not recognize its illocutionary force (Austin 1962, Searle 1969). Notice, however, that a variety of speech acts can be performed by It's cold in here: an assertion about the temperature, a directive to close the door, a request to get me a sweater, a complaint that the thermostat is too low. Because of its orientation toward action, and its contextual basis, the communicative meaning of an utterance cannot stem solely from referential meaning: it is now frequently considered in need of pragmatic analysis of how speakers (hearers) rely on context and inferential presumptions to convey (interpret) communicative intentions (Schiffrin 1994a: Chapter 6).2
Suggested thus far is that the referential function of language underlies phonology and syntax and helps to differentiate semantics from pragmatics. It has been less central to approaches to language that are more cognitively, socially, culturally or discourse based. In fact, it is in partial reaction to the emphasis on referential meaning and function that perspectives on language as varied as pragmatics, speech act theory, ethnography of communication, cognitive linguistics, functional grammar, interactional sociolinguistics, conversation analysis, critical discourse analysis and variation analysis have developed. Although the analyses in this book depend upon and draw from all of these perspectives, they will be brought together through a version of variation analysis (Schiffrin 1994a: Chapter 8) that alters and extends some of its traditional methodological and theoretical principles.
In an early statement describing the scope of variation analysis, Labov (1972a: 188) outlined the range of phenomena to be studied:
It is common for a language to have many alternate ways of saying 'the same' thing. Some words like car and automobile seem to have the same referents; others have two pronunciations, like working and workin'. There are syntactic options such as Who is he talking to? vs. To whom is he talking? or It's easy for him to talk vs. For him to talk is easy. Despite the lexical, phonological, and syntactic examples above, most early research on sociolinguistic variation (from the late 1960s through the 1970s) sought to explain the variation that had been largely ignored by phonologists: the non-referential differences in pronunciation or 'free' variation.3 Consider, for example, the alternation noted by Labov between working and workin', more precisely, between the velar and apico-velar nasal. This alternation is constrained both linguistically and socially. The more frequent use of in' in progressives and participles than in gerunds, for example, reflects different patterns of historical development (Houston 1989). In' is also socially constrained by speaker identity and style: it is used more frequently by working class than middle class speakers and by men than women, i.e. all speakers use in' more than ing when they are speaking casually instead of carefully (Labov 1972a; Trudgill 1974). Thus variation in allophones (different phonetic realizations of a single underlying phoneme) can be explained by social factors.
Since variationists try to discover patterns in the distribution of alternative ways of saying the same thing, an initial step is to establish which forms alternate with one another and in which environments they can do so, i.e. what aspects of context might matter. Those contextual features that might be related to the alternation among forms are viewed as constraints that have a systematic impact on the appearance of one variant rather than another. Depending on the variable, constraints may range all the way from the physiology of articulation to social identity. The process of identifying, coding and counting alternative realizations of a variable, identifying their environments and coding the constraints within those environments sets up its own requirements. One has to define and identify all the possible realizations of an underlying type (i.e. a closed set), classify the factors in the environment with which those variants may be associated, and then compare the frequencies with which different variants co-occur with different factors/constraints.
Despite interest in lexical and syntactic variation (as noted in Labov's quote above), variation analysis requires that we first establish which forms are the same "by some criterion such as 'having the same truth value'" (Labov 1972a: 188). This requirement has made it challenging (and, for some, theoretically problematic) to extend the notion of variable to levels of analysis beyond the phonological.
For example, the conditions in which David went to the store is true are not the same as the conditions in which David goes to the store would be true. Yet a referential difference between the preterit and the present tense can disappear in narrative, when the present tense is interpreted as the historical present tense and has a 'present time' meaning. Switching between the preterit and the historical present in oral narratives has been explained in relation to speech activity (Wolfson 1978), episode organization (Wolfson 1979) and evaluative (cf. expressive) function (Schiffrin 1981). Although these explanations all support Hymes' (1985: 150) prediction that "the recognition of social function brings recognition of new structure, transcending conventional compartments," the new structures that are recognized are quite different, in part because of the role of meaning. Wolfson (1976, 1978) argues against any exploitation of a 'present time' meaning. Rather it is speaker/hearer solidarity and textual structure that account for tense switches. Schiffrin (1981), however, argues that 'present time' meaning underlies the evaluative function of the historical present, i.e. function has moved from a referential to expressive and interpersonal plane. Thus, both analyses agree that the historical present tense has social and textual functions; they disagree on whether tense retains and exploits semantic meaning, and if so, in what pragmatic realms it does so.
Suggested thus far is that morphological, lexical and syntactic forms convey meanings that may (or may not) be neutralized in different discourse environments (Sankoff 1988). Forms that are semantically distinctive in one environment may not seem so in another. However, it is difficult to say whether residues of those meanings remain with each occasion of use and whether they are exploited for textual and/or pragmatic purposes (Lavendera 1978).4
The extension of variation analysis to discourse has taken two different directions, each of which has added further complications. The study of discourse variation attempts "to find patterns of language use that characterize the spoken language of a definable group in a specific setting" (Macaulay 2002: 284). As such, it encompasses a very broad view of discourse, variation and contextual factors (e.g. situation, participants). Studies grouped under this description could thus include analyses of all social and cultural differences in language use above and beyond the sentence: differences in conversational style by region (Tannen 1984) and gender (Tannen 1990), in contextualization cues by culture (Gumperz 1982), in narrative by gender (Johnstone 1980), race (Michaels and Collins 1984), or culture (Tannen 1980), in politeness strategies (Pan 2000) and dinner table conversations (Blum-Kulka 1997) by culture and situation. The list could clearly go on to include all studies of acting and speaking that are socially/culturally motivated and constructed.
Variation of a sort, to be sure - but not the sort of variation easily conceptualized as 'same' at an underlying level of language structure with surface variants contingent upon linguistic and social factors. Thus it would be difficult to identify and categorize all the possible realizations of an underlying style, narrative, politeness strategy, or conversation, classify the contexts in which those variants might occur, and then compare the frequencies with which different variants co-occur with different factors. Nor would this necessarily be the best approach. Many aspects of discourse are locally negotiated and co-constructed: identifying them and understanding why they appear, and how they do so, requires close attention to minute details of emergent properties and sequential contingencies of multi-functional units in discourse that are notoriously difficult to identify, let alone count. Likewise, the idea that ongoing discourse "constrains" its constituent parts bypasses the ways that sentence and text can co-constitute one another and that situational meanings emerge from what is said and done. Because of these differences, analyses of social/cultural differences in ways of speaking often reject the logic underlying Labovian variation analysis and can better be characterized as part of pragmatics, interactional sociolinguistics, discourse analysis, linguistic anthropology and the ethnography of communication.
|4||Reactive and proactive prototypes||110|
|7||Retelling a story||241|
|8||Who did what (again)?||277|
|9||Redoing and replaying||314|
|App. 1||Transcription conventions for data excerpts||341|
|App. 2||Four versions of Susan Beer's capture story||342|
|App. 3||Jack Cohen's narrative about Joey Bishop's childhood prank||349|