LINGUIST List 14.2658

Thu Oct 2 2003

Review: Discourse Analysis: Scheibman (2002)

Editor for this issue: Naomi Ogasawara <>

What follows is a review or discussion note contributed to our Book Discussion Forum. We expect discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in. If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for review." Then contact Simin Karimi at


  1. Julia Penelope, Point of View and Grammar

Message 1: Point of View and Grammar

Date: Wed, 01 Oct 2003 19:06:08 +0000
From: Julia Penelope <>
Subject: Point of View and Grammar

Scheibman, Joanne (2002) Point of View and Grammar: Structural
Patterns of Subjectivity in American English Conversation, John
Benjamins Publishing Company, Studies in Discourse and Grammar 11.

Announced at

Julia Penelope, unaffiliated scholar.


Beginning with an observation common among linguists who work with
language use -- that there seems to be little substantive information
passed among participants in conversations -- Scheibman asks why
speakers use a limited number of conventionalized structural patterns
in interactive discourse when linguists know that, theoretically,
language users have access to an infinite number of structural
combinations. She responds by positing that, in actual language use,
grammatical structures function more to indicate the speaker's point
of view, not to provide propositional information, as many theories of
language assume. In support of the thesis that the import of
conversational use is subjective, Point of View and Grammar presents
an analysis of the language of 33 adult speakers of American English
(21 women and 12 men) recorded in nine audiotaped informal
conversations. In all, 80 minutes of conversation were coded for
analysis, resulting in 2,425 utterances. Chapters 3 through 4 present
the results of the statistical analysis of those utterances. Appendix
A provides readers with the transcription symbols used to code
suprasegmental and metalinguistic elements of the utterances, and
Appendix B lists the 32 intermediate function verbs (e.g., ''be going
to,'' ''be supposed to,'' ''need,'' ''ought to,'' ''wanna,'' etc.) in
the database.

Chapter 1, ''Linguistic subjectivity and usage-based linguistics,''
introduces the theoretical background for the study and provides a
brief survey of scholarly research on how speaker point of view is
signaled by various grammatical elements (for example, the progressive
and perfect aspect, modals, tense and clause chaining) and specific
collocations such as the English verb ''remember,'' which is used most
frequently with ''I'' + ''don't.'' Her review of the literature
contextualizes the major thesis of Scheibman's study -- that the use
of specific grammatical structures, and even some lexical choices,
reflects the speaker's subjectivity -- and her challenge to the
privileged role of propositional transmission in theoretical
linguistics, assumed to be the primary function of language. Arguing
that there are entire subfields of linguistics that attend to the
expressive functions of language, she turns to the developing field of
usage-based linguistics and its emphasis on the interactive and
prosodic features of spontaneous conversation, features often treated
marginally, if at all, by linguistic theories that start by assuming
that the primary function of language is referential (presenting
propositions that transmit information about the world) rather than
focusing on how language functions expressively (providing information
about how speakers position themselves in the world and their
perceptions and evaluations of people, objects, and
events). Subsequent chapters present the evidence that the structures
that occur most frequently in conversation are those that express
speaker point of view.

Chapter 2, ''Classification and coding of conversational data,''
describes some of the problems with treating spontaneous conversation
as linguistic data (e.g., ''units of grammar are not necessarily units
in conversation,'' 18), the sources of the data that she analyzed, and
how the conversational utterances were coded for analysis. Noting that
one limitation of the study is the inaccessibility of gestural and
other kinesthetic information (because the data were collected on
audiotapes), Scheibman points out that such visual information,
although inherently significant to conversational interactions and the
expression of speaker subjectivity, does not bear directly on her
project, describing how speaker stance is communicated by ''the
frequency and cooccurrence of lexical and grammatical elements'' (20).
The remainder of the chapter describes in detail how the 2,425
utterances were coded and the kinds of decisions that had to be made
about how specific elements were to be understood. Coding the
utterances for analysis was not as straightforward as one might think,
and Scheibman's discussions of the role of interpretation in making
such decisions are detailed illustrations of why understanding
language use cannot be ignored in theoretical expositions. The coding
system was used to test hypotheses about how linguistic elements in
English signal speaker subjectivity and to make it possible to engage
in more open- ended exploration.

Chapter 3, ''Patterns of subjectivity in person and predicate,'' begins 
with a restatement of two theoretical assumptions introduced in Chapter 
1: (1) language, particularly conversational language, is subjective 
because speakers use it to express their point of view; (2) grammar -- 
conventionalized linguistic structures -- emerges from the repeated use 
of lexical and grammatical elements in natural discourse. Scheibman 
joins these assumptions to suggest two general hypotheses: (1) the 
linguistic elements that occur frequently in conversational use should 
be those that express speaker subjectivity (for example, adjectives 
such as ''great'' and ''good'' that express valuation); (2) there should be 
found a higher cooccurrence of items in combination that express 
speaker subjectivity than those that do not. She goes on to say that, 
in fact, the most commonly occurring combinations of subjects and 
predicates in the database are those that enable speakers to note their 
contributions to conversations, evaluate, and mark attitude and 
situation. The remainder of the chapter is organized by subject (1s, 
2s, 3, 1p, 3p); within those sections, she discusses the predicates 
that occur most frequently with each subject (for example, ''think,'' 
''know,'' and ''guess'' are the cognition verbs that occur most frequently 
with first person singular subjects, and ''know'' is the most frequently 
occurring verb in the present tense group). Twenty-seven tables present 
the statistical analysis of the types of verbs that cooccur with 
different subjects in the conversational database.

Chapter 4, ''The evaluative character of relational clauses,'' focuses
on the largest group of third person singular subjects, those that
occur with relational predicates, a majority of which are copular
clauses. Relational clauses occur the most frequently in the
database, accounting for 30 percent of the data tokens in Scheibman's
study and, she points out, describing the kinds of relations that
speakers usually express bears directly on her hypothesis that the
most frequent structures used in conversation are those that express
speaker subjectivity. Fifteen tables present the statistical analysis
of third person singular subjects by predicate type, tense, and
animacy of subject type, and so on. Scheibman concludes, on the basis
of this analysis, that a majority of third person singular utterances
do not describe the properties of people and events in the world;
rather, such utterances ''endow events, ideas, and entities with
characteristics'' based on speakers' evaluations'' (158). And, because
such utterances are not marked as first person, their use to convey
subjective information is covert, making it likely that they will be
accepted as objective descriptions unmediated by speaker point of

Chapter 5 presents Scheibman's conclusions drawn from her analysis in
Chapters 3 and 4: (1) the conventional classes and categories used to
describe linguistic elements, for example, subject and predicate, are
not as paradigmatically autonomous in interactive, spontaneous
discourse; (2) the functions of language -- the expression of
subjective valuations and the provision of objective propositions --
are not discrete and are best understood as a continuum; (3) in
contrast to authoritative speech, which assumes the consensus of
participants in o rder to present a stabilized reality, interactive
speech, which is filled with explicit markers of subjectivity, creates
a space that allows for the negotiation of meaning among participants.


Do not allow the numerous tables and statistics to keep you from
reading this book. It is one of the most exciting linguistic texts
I've read in years. The tables provide a visual presentation of the
statistical results, making it possible for Scheibman to devote large
portions of the text to exploring the implications of her findings. In
Point of View and Grammar, she brings together research done by some
of the best minds in linguistics over the past 40 years and presents a
synthesis, grounded in usage-based analysis, that promises a broader,
more fruitful approach to language and cognition and how language use
reflects our understanding of ourselves and others in the world.


Julia Penelope is now a freelance lexicographer and copy editor. While
actively engaged in linguistic research, she focused on language use,
in particular the variety of agentless passive constructions available
to English speakers and instrumental metaphors, and the interpretive
strategies such linguistic structures forced on hearers, and the ways
misogyny is expressed in English and in grammars of the language.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue