Linguistic Theory as Discourse
‘Surveys’ of ‘linguistic theory’ have become so numerous that a new one
calls for some justification. It seems to me that even though linguistics is about
language, the major works in linguistic theory have seldom been analysed and
synthesized as language, specifically: as a mode of discourse seeking to
circumscribe language by means of language. Perhaps this lack is due in part to
the limitations imposed by theorists who did not address discourse as a
linguistic phenomenon, or only marginally so. Perhaps too, it was tacitly
assumed that theories do not critically depend on the language in which they
happen to be expounded. Today, however, discourse has become a major area of
concern; and the dependence of concepts and arguments on the discourse that
constitutes them is widely acknowledged.
Therefore, to examine linguistic theories as discourse constructions is by no
means to discount their conceptual importance, but to insist on attending very
carefully to the emergence of those conceptions within the original discourse
before proceeding on to the more usual stages of abstraction and paraphrase.
This insistence can be particularly instrumental in tracing the development of
terminology, and the continuity, evolution, or change in the major lines of
argument not merely between theorists, but within the work of an individual
On the whole, the history of the ‘science of language’ has not been
unmanageably diffuse. Major theoretical works and frameworks have not been
overly numerous. And on the whole, the discipline has been fairly parsimonious
in its theorizing, indeed resolutely so in the face of the complexity of
language. Yet we can certainly not claim that the problems addressed by our
predecessors have by now vanished or been completely resolved. Instead, we
frequently sense a need to return to those problems and re-examine the
principles set forth decades ago to approach them.
In that situation, surveys of linguistic theory should be cautious about
imposing an artificial, retrospective sense of order and direction on the
discipline by distilling out a few main ‘ideas’, ‘schools’,
‘trends’, or ‘paradigms’. That method can abbreviate or conceal the
complexity and diversity of scientific interaction and discourse. A
counterbalance could be attained by surveying linguistics as a ‘model
science’ perpetually in the process of situating itself in respect to
Such a survey is a problematic and arduous project, but I hold it to be urgent
for several reasons. First, many of the issues in linguistics that preoccupy
linguistic theorists today were recognized and deliberated by our predecessors.
We cannot get a full sense of our domain by reducing the works of the founders
to a handful of precepts and slogans, without due regard for the overall
argument and context, including important qualifications and reservations. That
strategy tends to covert complicated, energizing research programmes too eagerly
into inhibiting new orthodoxies. And in hindsight, we may get the utterly
mistaken impression that linguistics did not properly appreciate the depth and
difficulty of the issues.
Second, linguistic theory is essentially a domain of work in progress, a
discipline always in search of itself. Leading theorists often voiced their
dissatisfaction with the state of linguistics as they saw it (13.3). But if we
construe their discontent as a pretext for writing off the past, we incur the
risk of repeating the same shortcomings they perceived and strove to alleviate.
Third, certain signs indicate that linguistic theory has for some years been
moving into a phase of stagnation and diminishing returns. Despite decades of
effort, the relations between theory and practice, between model and domain, or
between method and evidence, have not been definitively established, and seem to
be shifted once again by every new school or trend. In consequence, the history
of the discipline may appear discontinuous and non-cumulative, with research
projects typically clustered around sporadic bursts of theorizing. The status of
theoretical entities, even such central ones as ‘word’ and ‘sentence’,
remains in dispute. No consensus obtains about the future trends and
modifications that linguistics should undergo. In such a state of affairs, we
cannot merely wait to see what develops in day-to-day research and discussion.
We need to draw up the theoretical balance sheets of past investigations.
Surveying the major issues and problems of the discipline through their
treatment in the discourse foundational works can be an inaugural step in
planning for future research on a truly comprehensive and organized scale.
All linguists share at least one special predicament: they can get evidence only
from their own encounters with language, with and within some mode of discourse
(cf. 13.1, 48). The system never steps forward to be ‘observed’ in some
concrete selfhood; and data are not data until they have been understood as
language. In consequence, linguists deal with data in whose constitution and
interpretation they are always to some degree involved, at least behind the
scenes. Since language is so extraordinarily sensitive to how it is used, it may
assume different appearances depending on how it is grasped. We therefore need
to expand our scope from ‘looking at language’ to ‘looking at linguists
looking at language’ and in particular talking or writing about it. We cannot
eliminate the linguist's perspective, but we can scrutinize it by asking how
human beings, whether linguists or ordinary speakers, abstract systematic
knowledge from language experience and at the same time apply systematic
knowledge in order to relate experience to language (cf. 13.44).
That you must ‘know language’ to ‘understand language’ and vice versa is
a truism, but by no means an insignificant one. We seem to confront a peculiarly
vicious circularity enshrouding the question of how we might approach language
from the ‘outside’: how children or linguists or anybody else can reach the
‘critical mass’, the stage of ‘knowing’ the system
behind or beyond the individual uses of language (13.38). Much of that
knowledge is concealed from conscious awareness during everyday discourse, and
the prospects for making it conscious and explicit are by nature precarious
(13.49). To observe yourself observing language, to watch or hear yourself
thinking, to grasp your own understanding -- all these acts are easily beset by
paradox or infinite regress. We can, however, subject the discourse of those
engaged in such acts to steadily more circumspect and integrative scrutiny,
thereby adding fresh emphasis to our perenniel insistence on the centrality of
language (cf. 13.22).
My survey accordingly proceeds by arranging and presenting the discourse, the
statements and arguments, of representative theorists in linguistics of this
century, sticking as close as is feasible to their actual wordings, especially
where major points are expressed. By this expedient, I hoped to restrict my own
role in increasing or complicating the mediation between linguistics and
language, as I would have had to do had I paraphrased and summarized the sources
in my own words. Though admittedly laborious, this method may help to reanimate
the complex flow of the discourse in the gradually emerging discipline, to focus
on characteristic moves, and to retrace the key terms as they gain or lose
currency. Proceeding by author rather than by ‘school’ may help to
accentuate individual views, voices, and personalities, and thus to
re-experience some of the momentum and perplexity of repeated confrontations
with the recalcitrant problems that the study of language necessarily raises.
Due to this gallery of problems, a general book on linguistics tends to have the
character of a performance, raising and responding to typical questions , such
Where does linguistics stand among the other disciplines?
aspects of language deserve to be put in focus, and which ones are of lesser
means or methods are recommended or rejected?
do linguists gather data, and how can they check their own estimation of it
against other sources?
are examples brought to bear on theoretical issues and abstractions?
are the fundamental units and structures of language?
is the theoretical status of traditional concepts such as ‘word’,
‘phrase’, and ‘sentence’?
shall be seeing quite a spectrum of potential answers, some explicit, others
merely implicit. Few of the answers will seem definitive, since they depend on
the goals and aspirations of the particular theorist, and these are by no means
uniform (cf. 13.58, 60ff). Still, considering such a spectrum assembled in one
volume may shed light on the nature of the questions, whatever the eventual
answers we may yet select.
It was rather agonizing to decide which ‘fundamental works’ should be used,
given the unmanageably large number worthy of inquiry. My selection was guided
by two major criteria. First, these works were influential in the general
development of theories or models, as attested for instance by frequent
citation. Second, these works propound such a wide range of positions and issues
that we can profit by bringing them into explicit interaction with each other.
My treatment is only roughly in chronological order, because the works and their
spans of influence sometimes overlapped in time, and because some influences
emerge more clearly through direct follow-ups, e.g. Bloomfield to Pike,
Hjelmslev to Chomsky, and Firth to Halliday. However, similar arguments and
conceptions also appear where we cannot trace such influences, or at least none
that the authors acknowledge. Conversely, demonstrable influences do not
necessarily promote agreement, and successors may differ from their predecessors
or teachers on major issues.
Obviously, my selection could have been different or larger. But the approach
proved to require such detailed attention to each work and theorist that I
lacked the space to include more of them. For motives of size, I regretfully
deleted a chapter on Terry Winograd, a major thinker both in linguistics and in
artifical intelligence. I also deeply regretted not being able to deal with such
undeniably influential linguists as Emile Benveniste, Dwight Bolinger, Wallace
Chafe, Simon Dik, Charles Fillmore, Charles Carpenter Fries, Hans Glinz, Joseph
Grimes, Z.S. Harris, Roman Jakobson, Daniel Jones, William Labov, George Lakoff,
Robert E. Longacre, Aleksei Leontev, Nikolai Marr, Andre Martinet, Vilem
Mathesius, Ivan Meshchaninov, Nikolai Trubetzkoy, or Leo Weisgerber. Also, I
would have liked to include such precursors and pioneers as Franz Bopp, Jan
Baudouin de Courtenay, Samuel Haldeman, Wilhelm and Alexander von Humboldt,
Hermann Paul, Rasmus Rask, Henry Sweet, Dwight Whitney, etc. And major figures
from neighbouring disciplines also deserve such attention: semioticians such as
Julia Kristeva, Jurij Lotman, Charles Morris, Charles Peirce, Thomas Sebeok,
etc.; or language philosophers such as John Austin, Jacques Derrida, Michel
Foucault, Paul Grice, Martin Heidegger, Alfred Korzybski, Jacques Lacan, John
Searle, Ludwig Wittgenstein, etc.; or logicians such as Rudolf Carnap, Max
Cresswell, Richard Montague, Janos Petofi, Alfred Tarski, Lotfi Zadeh, etc.; or
psychologists and psycholinguists like Philip N. Johnson-Laird, Alexander Luria,
William Levelt, William Marslen-Wilson, George Miller, Charles Osgood, etc.;
sociologists like Basil Bernstein, Erving Goffman, Harvey Sacks, Emmanuel
Schegloff, etc.; anthropologists such as Edmund Leach, Bronislaw Malinowski,
Claude Levi-Strauss, etc.; or analysts of narrative and literary or poetic
discourse such as Roland Barthes, Algridas Greimas, Tzvetan Todorov, etc. Though
I had to exclude all these figures, I glean some comfort from the fact that I
have made use of their work in my previous writings, and from the hope that I
may give them more attention in the future.
here to return to Linguistic Theory Main Page
Click here to return to Linguistic Theory Main Page