The Linguistics Website
Homepage About Me My Papers and My Essays Selection Academic Papers Bibliography Computational Softwares Glossary of Linguistics Links
Main Links


Linguistic Theories

1. Structural Linguistics


Ferdinand de Saussure is the originator of the 20th century reappearance of structuralism, specifically in his 1916 book Course in General Linguistics, where he focused not on the use of language (parole, or talk), but rather on the underlying system of language (langue) and called his theory semiotics. This approach focused on examining how the elements of language related to each other in the present, that is, 'synchronically' rather than 'diachronically'. Finally, he argued that linguistic signs were composed of two parts, a signifier (the sound pattern of a word, either in mental projection - as when we silently recite lines from a poem to ourselves - or in actual, physical realization as part of a speech act) and a signified (the concept or meaning of the word).

This was quite different from previous approaches which focused on the relationship between words and the things in the world they designated. By focusing on the internal constitution of signs rather than focusing on their relationship to objects in the world, Saussure made the anatomy and structure of language something that could be analyzed and studied.

Saussure's Course influenced many linguists in the period between WWI and WWII. In America, for instance, Leonard Bloomfield developed his own version of structural linguistics, as did Louis Hjelmslev in Scandinavia. In France Antoine Meillet and Émile Benveniste would continue Saussure's program. Most importantly, however, members of the Prague School of linguistics such as Roman Jakobson and Nikolai Trubetzkoy conducted research that would be greatly influential.

The clearest and most important example of Prague School structuralism lies in phonemics. Rather than simply compile a list of which sounds occur in a language, the Prague School sought to examine how they were related. They determined that the inventory of sounds in a language could be analyzed in terms of a series of contrasts. Thus in English the words 'pat' and 'bat' are different because the /p/ and /b/ sounds contrast. The difference between them is that the vocal chords vibrate while saying a /b/ while they do not when saying a /p/. Thus in English there is a contrast between voiced and non-voiced consonants. Analyzing sounds in terms of contrastive features also opens up comparative scope - it makes clear, for instance, that the difficulty Japanese speakers have differentiating between /r/ and /l/ in English is due to the fact that these two sounds are not contrastive in Japanese. While this approach is now standard in linguistics, it was revolutionary at the time. Phonology would become the paradigmatic basis for structuralism in a number of different forms.

Source: (T.R. Quigley, 1998)

Structuralism and Poststructuralism: Background Summary and Analysis [1]

1.1. The Kantian Background - Review

1. What defines the form of human experience?

a. Space and Time (a priori forms of Intuition).
b. Categories (concepts of the Understanding).

2. For Kant, these concepts are fixed and universal, i.e. ahistorical.

3. Problems: Kant's categories seem arbitrary and their universality is merely assumed by Kant, not proven.

4. In a post-Darwinian world, it seems more likely that such concepts and categories of human experience are historical, i.e. subject to change - contingent.

5. In response to this shift in emphasis, Husserlian phenomenology demands that we look and see what the status of such categories are independent of our theoretical presuppositions.

6. Social scientists, who approach this issue empirically through observation and prediction, suggest that there may be significant variations in conceptual frameworks culturally and historically. But the evidence is not entirely conclusive. So, from a scientific standpoint, the issue remains open.

1.2. Summary of Saussure's Structural Linguistics

The French linguist Ferdinand de Saussure studied language from a formal and theoretical point of view, i.e. as a system of signs which could be described synchronically (as a static set of relationships independent of any changes that take place over time) rather than diachronically (as a dynamic system which changes over time).

According to Saussure, the basic unit of language is a sign. A sign is composed of signifier (a sound-image, or its graphic equivalent) and a signified (the concept or meaning). So, for example, a word composed of the letters p-e-a-r functions as a signifier by producing in the mind of English-speakers the concept (signified) of a certain kind of rosaceous fruit that grows on trees, viz., a pear.

According to Saussure, the relation between a signifier and a signified is arbitrary in at least two ways. First, there is no absolute reason why these particular graphic marks (p-e-a-r) should signify the concept pear. There is no natural connection or resemblance between the signifier and the signified (as there would be in what Saussure calls a symbol, i.e. an iconic representation such as a descriptive drawing of a pear). After all, it's not as if the word "pear" looks or sounds anything like a pear! In fact, a moment's reflection makes it clear that the connection between the signifier and the signified is due to a contingent historical convention. It didn't have to happen the way it did. In principle, the word "pare", "wint", or even "apple" would have worked just as well in associating a word with the concept pear! But given that the word "pear" has come to signify the concept pear in English, no one has the power to simply change it at will. In other words, the relationship between a word and a concept is arbitrary in one sense (in terms of its origin) but not in another sense (in terms of its use).

Saussure makes a second point about the arbitrariness of the sign. He points out that the relation between the sign itself (signifier/signified pair) and what it refers to (what is called the referent, i.e. the actual piece of fruit-the physical object) is also arbitrary. This claim is less plausible than the former. For example, one might object that the concept in the mind of the speaker is formed, either directly or indirectly, by actual pears. Ideally then we would expect it to be the case that the properties of actual pears would be causally related to our concept of a pear-that the characteristics of pears produce in one's mind the concept of a pear either directly through experience with pears, or indirectly through pictures of pears, descriptions, or some such thing. Thus, the concept pear might be thought of as some basic information and set of beliefs about actual pears, e.g. what they look like, how they feel and taste, what they're good for, etc.

Saussure's way around this obvious objection is to say that his interest is in the structure of language, not the use of language. As a scientist, Saussure limited his investigation to the formal structure of language (langue), setting aside or bracketing the way that language is employed in actual speech (parole). Hence, the term structuralism. Saussure bracketed out of his investigation any concern with the real, material objects (referents) to which signs are presumably related. This bracketing of the referent is a move that enabled him to study the way a thing (language and meaning) is experienced in the mind. In this sense, his motivation was similar to Husserl's. And in the end, Saussure never offered a method for investigating how language as a system hooks up to the world of objects that lie outside language. As we shall see, this was to have far-reaching effects.

Thus, according to Saussure's structural linguistics, each sign in the system of signs which makes up a language gets its meaning only because of its difference from every other sign. The word "pear" has no meaning in itself or in the intention of the speaker, but only due to the fact that it differs from other possible graphic images such as p-e-e-r, p-e-a-k, f-e-a-r, b-e-a-r, etc. In other words, it doesn't matter how the form of the signifier varies, as long as it is different from all the other signifiers in the system (langue). To the structuralist, meaning arises from the functional differences between the elements (signs) within the system (langue).

An economic analogy helps to illustrate Saussure's theory of meaning. The signs of a linguistic system are like the coins of a monetary system or currency. Thus, a system of signs (words of a language) is analogous to a system of values.

A quarter has a certain monetary value determined by its exchange value. Quarters can be exchanged for other things because they have a designated (but flexible) value. Quarters can be used to buy goods or commodities. But they also have a fixed value in relation to other coins. So, for example, a quarter is equal to two dimes and a nickel; it is more than a penny; it is less than a dollar, etc., etc.

Linguistic signs also have values in relation to other signs. For example, the word "bachelor" can be "exchanged" for the term "unmarried man". This is, in many ways, an equal exchange. That's what it means for words to be synonymous - they have the same meaning or linguistic value. They can be substituted or exchanged for one another just as the quarter can be exchanged for two dimes and a nickel.

1.3.The Significance of Structuralist Theory

The first thing to notice is that, according to structuralist theory, meaning is not a private experience, as Husserl thought, but the product of a shared system of signification. A text is to be understood as a construct to be analyzed and explained scientifically in terms of the deep-structure of the system itself. For many structuralists, this "deep-structure" is universal and innate.

If we consider the application of structuralism to art and extend the monetary analogy, we can think of paintings as comprised of many languages or sets of conventions that play a role in the exchange of signs. For example, the language of western academic painting can be contrasted with the language of African sculpture or Japanese brush painting. Just as one word in the English language is paired with a concept, so a visual image, icon, or symbol is paired with a concept or idea that it is said to "express". Such a study of signs in the most general sense, whether visual or verbal, is called semiotics. In the West, art schools are the institutions that have the function of passing on these visual conventions. [2]

Second we should note that in structuralism, the individual is more a product of the system than a producer of it. Language precedes us. It is the medium of thought and human expression. Thus, it provides us with the structure that we use to conceptualize our own experience.

And third, since language is arbitrary, there is no natural bond between words and things, there can be no privileged connection between language and reality. In this sense, reality is also produced by language. Thus, structuralism can be understood as a form of idealism.

It should be clear from what we've just said that structuralism undermines the claim of empiricism that what is real is what we experience. It can also be seen as an affront to common sense, esp. to the notion that a text has a meaning that is, for all intents and purposes, straightforward. This conflict with common sense, however, can be favorably compared with other historical conflicts (e.g. Copernicus' heliocentric system). In other words, things are not always what they seem. Thus, the idealist claim of structuralism can be understood in the following way: Reality and our conception of it are "discontinuous". [3] This view has important implications, as we shall see below.

According to structuralist theory, a text or utterance has a "meaning", but it's meaning is determined not by the psychological state or "intention" of the speaker, but by the deep-structure of the language system in which it occurs. In this way, the subject (individual or "author") is effectively killed off and replaced by language itself as an autonomous system of rules. Thus, structuralism has been characterized as antihumanistic in it's claim that meaning is not identical with the inner psychological experience of the speaker. It removes the human subject from its central position in the production of meaning much as Copernicus removed (de-centered) the Earth from its position at the center of the solar system. And since language pre-exists us, it is not we who speak, as Heidegger was to say, but "language speaks us".

1.4.From Pre-Structuralism to Structuralism to Post-Structuralism

The shift from a pre-structuralist to a structuralist theory of language and the implications drawn from it by poststructuralists is represented in the following diagrams:

A. Pre-structuralist theory assumes that there is an intimate connection between material objects in the world and the languages that we use to talk about those objects and their interrelations.

B. As we saw above, Saussure puts this connection between the material object and the word in brackets, i.e. he sets it aside in order to study the very structure of language. Thus,

According to Saussure's structuralist theory of language, the meaning of a term (a word or expression) does not begin and end with the speaker's experience or intention (as it does in Husserl's theory). The act of speaking and intending presupposes a language already in place and upon which the speaker must rely in order to say anything at all. Concepts or meanings are picked out (signified) because of the differences in the network of words (sound- or graphic-images) that make up the language (langue). Thus each word-each structural element of the language-finds its own relative position or node within the network of differences.

In other words, the meaning of a particular term in a language is due to its relative difference from all other terms in the language. A signified, i.e. a concept or idea, is properly understood in terms of its position relative to the differences among a range of other signifiers (words with different positions in the network (langue) and, hence, different meanings).

C. Poststructuralist theory denies the distinction between signifier and signified. According to the poststructuralist, concepts are nothing more than words. Thus, signifiers are words that refer to other words and never reach out to material objects and their interrelations. To indicate this shift in theory, the French philosopher Jacques Derrida introduces the word "différance" to indicate the relation between signifiers as one of difference and deferral.

If a word's meaning is solely the result of its difference from other words, then the meaning (the concept or signified) is not an additional thing "present" in the sign itself. On the contrary, "meaning" (if it can be called that at all) is the ever-moving play of difference from signifier to signifier; a slipping from word to word in which each word retains relations to ("traces" of) the words that differ from it.

Thus, according to poststructuralists such as Derrida, the specification of meaning is an infinite and endless process! Meaning, to some extent, always escapes one's grasp-it is always just out of reach, ungrounded, with no origin in the intention of the speaker, contrary to what Husserl thought. In other words, when a speaker uses certain words ("This is a pear"), then according to the theory she does not have a nonlinguistic object or concept in mind-there is no additional thing or "object" outside of the language (i.e. no "meaning") that could be transmitted or made "present" to her listener or reader. There is nothing there in her speech but language, i.e. a network of signification.

Thus, "meaning" is the result of a play of différance-a movement which brings about both difference and deferral. (It may help here to bring in the traditional distinction between the denotation and the connotation of a term. The connotation may be thought of as the aura of suggestion, the echo or trace of other words to which it is related by such things as association, common usage, similarity, etc. The denotation, the relation (reference) between the word and the actual thing denoted by the word, from structuralism on, is bracketed and never brought back. Its absence, however, leaves its own "traces" in the form of problems for a poststructuralist theory of language. (See below.)

So the poststructuralist draws the following consequences from the study of language:

1. Meaning is never fully present in any one signifier, but is infinitely deferred or suspended.
2. Meaning is contextual, i.e. affected by related words.
3. There is always an excess of meaning.

But there is another, more radical, consequence that can be drawn from our analysis. If the meaning associated with an expression is not present in the expression itself, and if the speaker must make his own presence felt by communication through words, then it follows that the speaker is never fully present in the act of using language. And if, as a human being, I can only think and experience a world through language, then "I" and "my presence" are as much deferred as the meanings I attempt to grasp when I try to understand and explain myself. In other words, I am never present even to myself. Rather, it is language that speaks, not a unified and autonomous ego or self. (How is this related to Kant's theory of knowledge?)

One final note. On p.60 of Literary Theory, Eagleton makes use of the following argument:

1. All experience depends on language.
2. Since, to have a language is to be part of a whole form of social life, there is no possibility of a private language.
3. Therefore, all experience is social experience, i.e. there are no private experiences.

This argument presupposes the notion in Saussure (and Hjelmslev in Prolegomena to a Theory of Language) that language is constitutive of experience. [4]

Notice the central role played by the premise that experience itself "depends on" or is structured by language. Without this assumption, the slide into the de-centered self is not so easily motivated. (Cf. Heidegger's notion of the de-centered self. Derrida himself says that consciousness is an effect of language.) This poststructuralist view of language undermines the theories of Descartes, Husserl and most of western metaphysical thinking about the primacy or centrality of the subject and reinforces the notion of the "decentered self" as characteristic of the human condition.

What alternatives can we imagine as a challenge to the poststructuralist position? One strategy would be to start by agreeing with Kant that we must have categories or concepts of some kind to organize human experience. But we might also disagree with Kant over the nature and a priori character of those concepts. In doing this, we could borrow from Heidegger the view that the categories of human experience are historical in nature and potentially in flux-not fixed and universal. But then we might question Heidegger's emphasis on the linguistic nature of these concepts by drawing on Gestalt psychology to argue for the existence of certain "structural" and hard wired components of human perception and thought of a prelinguistic nature. This is just one tentative direction one might take in challenging the view presented by the form of poststructuralism that we've been considering.

Other problems are raised if we consider language not simply as an object but as a practice. Suppose I say to you, "Open the window" in a situation where there is no window in the room. You might ask, "What do you mean?" This would be to question my "intentions" - what am I trying to accomplish by saying what I've said? Perhaps I am making a point about the fact that there is no window in the room. My paradoxical statement - inexplicable in Saussure's structuralist terms - might be meaningful to you in another practical sense. This is because understanding is recognizing what effects one might seek to bring about through the use of certain words. My obscure command might be a request that we move to a room that has a window.

In other words, speech is not just an object, it is a form of behavior, and as such it can only be understood contextually, i.e. in a situation. This realization of the pragmatics of language signals a shift from language to discourse, and a concomitant change in emphasis away from a text's meaning to its function.

In the end, we may want to say not so much that reality is linguistic but that language is real, and not necessarily all there is to human reality and experience.


1. See also: Saussure, "The Linguistic Sign" in Innis, Semiotics; Barthes, "From Work to Text", Art in Theory, 940-46. [return]
2. For more on the application of structuralism to painting, see Francis Frascina, "Realism and Ideology: An Introduction to Semiotics and Cubism", in Primitivism, Cubism, Abstraction: The Early Twentieth Century, Harrison, Frascina and Perry, New Haven and London: Yale University Press, 1993, 87-183. [return]
3. Eagleton, Literary Theory, Minneapolis: Univ. of Minnesota Press, 1983, 108. [return]
4. See Innis, Semiotics, 26. [return]

2. Sytemic-Functional Linguistics


Systemic-Functional Linguistics (SFL) is a theory of language centred around the notion of language function. While SFL accounts for the syntactic structure of language, it places the function of language as central (what language does, and how it does it), in preference to more structural approaches, which place the elements of language and their combinations as central. SFL starts at social context, and looks at how language both acts upon, and is constrained by, this social context.

A central notion is 'stratification', such that language is analysed in terms of four strata: Context, Semantics, Lexico-Grammar and Phonology-Graphology.

Context concerns the Field (what is going on), Tenor (the social roles and relationships between the participants), and the Mode (aspects of the channel of communication, e.g., monologic/dialogic, spoken/written, +/- visual-contact, etc.).

Systemic semantics includes what is usually called 'pragmatics'. Semantics is divided into three components:

* Ideational Semantics (the propositional content);

* Interpersonal Semantics (concerned with speech-function, exchange structure, expression of attitude, etc.);

* Textual Semantics (how the text is structured as a message, e.g., theme-structure, given/new, rhetorical structure etc.

The Lexico-Grammar concerns the syntactic organisation of words into utterances. Even here, a functional approach is taken, involving analysis of the utterance in terms of roles such as Actor, Agent/Medium, Theme, Mood, etc. (See Halliday 1994 for full description).


  • Halliday, M.A.K. 1961. Categories of the theory of grammar. Word 17. Reprinted in Bertil Malmberg (ed), .... . Abridged version in Halliday (1976).
  • Halliday, M.A.K. 1994 Introduction to Functional Grammar, Second Edition, London: Edward Arnold.
  • Martin, James R. 1992 English Text: system and structure. Amsterdam: Benjamins.

3. Generative Linguistics

source: (7 March 2008)

Generative linguistics is a school of thought within linguistics that makes use of the concept of a generative grammar. The term "generative grammar" is used in different ways by different people, and the term "generative linguistics" therefore has a range of different, though overlapping, meanings.

Formally, a generative grammar is defined as one that is fully explicit. It is a finite set of rules that can be applied to generate all those and only those sentences (often, but not necessarily, infinite in number) that are grammatical in a given language. This is the definition that is offered by Noam Chomsky, who invented the term [1] , and by most dictionaries of linguistics. It is important to note that generate is being used as a technical term with a particular sense. To say that a grammar generates a sentence means that the grammar "assigns a structural description" to the sentence.[2]

The term generative grammar is also used to label the approach to linguistics taken by Chomsky and his followers. Chomsky's approach is characterised by the use of transformational grammar – a theory that has changed greatly since it was first promulgated by Chomsky in his 1957 book Syntactic Structures – and by the assertion of a strong linguistic nativism (and therefore an assertion that some set of fundamental characteristics of all human languages must be the same). The term "generative linguistics" is often applied to the earliest version of Chomsky's transformational grammar, which was associated with a distinction between the "deep structure" and "surface structure" of sentences.

[1] Chomsky, Noam (1957,2002). Syntactic Structures. Mouton de Gruyter, 13.
[2] Chomsky, Noam (1965). Aspects of the Theory of Syntax. MIT Press.




Generative grammar: algorithmic (phrase structure grammars)

            Transformational grammar (1960s)

            Generalised phrase structure grammar (late 1970s)

            Head-driven phrase structure grammar (1985)

            Principles and parameters (1980s)

            Lexical functional grammar

            Categorial grammar (lambda calculus)

            Montague grammar

            Dependency grammar (Lucien Tesnière 1959)

            Link grammar

            Cognitive grammar / Cognitive linguistics

            Construction grammar           

                        Fluid Construction Grammar

            Word grammar

            Stochastic grammar: probabilistic

            Operator Grammar

            Functional grammar: usage-oriented (behaviorist)

            Danish Functionalism

            Systemic functional grammar

            Role and reference grammar