下载掌阅APP，畅读海量书库
立即打开
The question“As What Should Language Be Studied?”is a very essential one to settle if we are to conduct a comparatively thorough investigation of language and language in use, because what attitudes or views linguists or philosophers of language take towards language will greatly affect approaches they will adopt, objects they will set, or even paradigms they are likely to use in their study of language.Language is a multi-dimensional entity, so any different perspective to it can reveal something intrinsic to its nature.Halliday（1978），for example, views language as meaning potential, studies it as instrumentality instead of as autonomy.He takes a socio-functional approach to language and thus centers upon the inter-organism perspective rather than the intra-organism perspective which Chomsky and his followers have so far devoted themselves to.In the case of paradigm, Chomsky, on the other hand, adopts a mathematic one, rejecting the structuralis t 1 empirical method and claiming that empiricism cannot explain the creativity of language.In what follows then our attention, on the basis of Saussure’s seminal work, will be focused upon the discussion of whether language ought to be differentiated as a medium from the expression that involves it or even the content expressed in it.If so, what is this medium, what is its nature, and its meaning?What is the difference between them?And finally the discussion will go on to the question of whether language should be studied as an autonomous system.Along with it are some enumerations of the objections, criticisms on autonomous linguistics and, more importantly, our own attitudes and evaluations regarding the problem in question.
At the beginning of this century, Ferdinand de Saussure, the generally acknowledged father of modern linguistics stipulated the task of modern linguistics—the analysis of language—as“a system of units”. Or put differently，“to do linguistics is to define the units of a language, the relations between them, and their rules of combinations”（Culler 197（8）6：94）.And he consequently proposed in his posthumous Cours de Linguistique Generale（1916）a series of distinctions： 2 langue vs.parole, synchronic vs.diachronic, paradigmatic relations vs.associative ones, internal vs.external, and form vs.substance.It dawns upon us that the most central concept in his groundbreaking work is the dichotomic separation of langue from parole.Saussure himself claims that he is not content with the linguistics that he inherited and knew.For him, linguists like his predecessor s 3 deviated from what they should do.Due to influences by the triumphant positivism in the Western thought in the 19th century, Saussure observes that they“muddled up faits de langue with faits de parole[……]so little progress had been made towards establishing linguistics as a science”（Harris 1988：37）.In actuality what they were doing was to find the causes of historical changes or at best the linguistics of individual act（see Barthes 1964：13）.Such a prospect, Saussure claims，“never attempted to determine the nature of the object it was studying, and without this elementary operation a science cannot develop an appropriate method”（Saussure 1959：3）.In order to make explicit the nature of the object to be studied, the initial and essential step that ought to be taken is to revolutionalise the traditional view of language, the then common but confusing term, for linguistics, Saussure points out, unlike other sciences, has no ready-made object of study given in advance.Saussure subsequently initiates a novel point of view to language, the then complex and heterogeneous phenomenon, contending that in linguistics，“it is the viewpoint adopted which creates the object”（Saussure 1983：23）.
Through observation of the speech circuit, from which the multiform and heterogeneous mass or unclassifiable reality result（see Barthes 1964：13），Saussure means to differentiate system and use and deductively postulates the conception of linguistic system, namely, in his own term, langue, which by definition is the“well-defined[homogeneous]object in the heterogeneous mass of speech acts”（Saussure 1959：14），thus making good sense of the“confused mass of heterogeneous facts”. In separating langue from parole， 4 Saussure brings about a momentous distinction between the underlying linguistic system which is common to all and the various instances of individual linguistic behaviors, alternatively its actual manifestations.The distinction between langue and parole is of fundamental importance because it does not only separate“what is social from what is individual, what is psychological from what is material, what is essential from what is ancillary or accidental（see Saussure 1983：14），but also isolates a suitable and analyzable object of study for linguistic investigation.Such an object is homogeneous, excluding everything else besides itself.
With reference to its acquisition as a common property, Saussure claims that langue is what individuals assimilate passively as they learn a language as a set of forms in a community. Or in his own terms, as a“hoard deposited by the practice of speech in speaker who belong to the same community，[as]a grammatical system, to all intents and purposes[that]exists in the mind of each speaker”（Saussure 1959：13-14）.In this sense, it is also a social product or, to borrow two terms from Whitney and Durkheim, a social institution or a social fact, which can be neither created nor modified by any individual.It is also a“collective contract that must be accepted in its entirety in communication”（see Barthes 1964：14）.
Saussure makes it explicit in the Course that language is a system of signs and it should be included in the domain of semiolog y 5 and be taken to be a special semiological system at that. There are two layers of significance in these assertoric remarks.To begin with, the notion of system, or to be precise, structuralis m 6 promotes a novel conception of language and exerts considerable influence in the field of linguistics in general and even commences a way or a trend of thinking that generates inspiration to other fields of science.As seen from the Course, the traditional view of language was dominant at his time because it was commonly held that the world consists of independently existing objects capable of objective observation and classification.So in the field of linguistics, such positivist perspective leads to“a notion of language as an aggregate of separate units called‘words'，each of which somehow has a separate‘meaning'attached to it, the whole existing with a diachronic or historical dimension which makes it subject to observable and recordable laws of change”（Hawkes 1877：19）.
Saussure, however, bluntly rejects this substantive view of language and adopts a holistic or structural view instead. Hence he sets forth a revolutionary perspective towards language.As is made explicit in the Course, Saussure holds that it is not sufficient to study language only in terms of its constituent parts and diachronically, but also in terms of the relations between the individual parts and moreover synchronically.That is, it ought to be a structure consisting of a series of substructures, i.e.，its current structural property, in which constituents of the system acquire their value through opposition with their counterparts in the same structure or system.So, one of Saussure’s originalities, which has given rise to revolution in linguistics and in semiology（see Tobin 1990：xii），as Jameson（1972）points out, is the novel viewpoint that“language as a total system is complete at every movement, no matter what happens to have been altered in it a moment before”（1972：5-6）.The structural or holistic perspective of language, unlike an aggregate one in which every component part has an independent existence, exhibits that language does not construct its formation of words by reference to the patterns of reality， 7 but on the basis of its own internal and self-sufficient rules（see Hawkes 1977：17）.Therefore it should be studied as a system, autonomous and unified.Alternatively, contrasts or oppositions within the system alone determine the value of its linguistic sign（Harris 1988：43）.“A language is a system of which all its parts can and must be considered as synchronically interdependent”（Saussure 1983：124）.
In addition to the holistic perspective to language, which is frequently regarded as one of the two pillars in the Suassurian linguistics, another one, more scientific and more pivotal, is the semiological approach that Saussure advocates adopting for a profound understanding of the communicative instrument. He says：
“Is it not obvious that language is above all a system of signs and that therefore we must have recourse to the science of signs if we are to define it properly?（Saussure 1966：16）.
He adds further：
“Language is a system of signs that express ideas and is thus comparable to the system of writing, to the alphabet of deaf-mutes, to symbolic rituals, to forms of etiquette, to military signals etc.. It is but the most important of these systems”（Saussure 1966：16）.
Such a perspective, in reality, is the strengthening of the revolutionary structuralist view of language that has changed the way language is studied, for semiology, the science of signs presupposes arbitrariness of the sign. Such unmotivated properties in signs in turn yield as logical and necessary corollarie s 8 （see Culler 197（8）6：39-45）two essential notions.One is the concept of system, which interrelates individual isolated signs as a whole.Hence, in the case of language, it becomes a system of linguistic signs that are mutually related in value and hence acquire their respective values by virtue of opposition or difference, functional in essence, within the boundary.Consequently, language, either at the levels of phonology, morphology or syntaxs, is a relationally formal and autonomous system.
The other one then is that of synchrony， 9 by which Saussure means that language is studied without reference to time and as a currently sustaining system. Such an insistence on atemporal approach to language is equally profound since it suggests the（metaphysical）existence of a language as a system with currently adequate properties in addition to its historical nature.Given this consideration, language is not an aggregate or a nomenclature as the Biblical account of Adam naming animals or as the Platonian philosophy of natural nomenclaturism indicates, because to treat language as such is“to isolate words from the linguistic system to which they belong, and simultaneously, to isolate the language-user from the linguistic community”（Harris 1988：17）in which the user is a member.Moreover in fact, the nomenclaturists'language does not count as a language at all, but only as loose marks that lack systematicity.Therefore, language, instead of being an aggregate of names, is an organic system, though before its construction nothing is given in advance.
Early in the Western linguistic tradition, language is taken to be of divine origin and separated from thought and thus“peripheral to our grasp of the world”（Harris 1988：6）. Language, as is evident in the Biblical story and Plato's philosophy of language, is an aggregate of separate names, or words or units.These words or vocal labels have meanings because they are used to surrogate for something such as objects, or properties given prior to language.Similarly, Aristotle and the eighteenth century philosopher Locke（1706）alter a bit but still believe that thought comes before language.They hold that, words, unlike the Platonian conception of standing for either objects or concepts, have meanings for they stand exclusively for“mental affections”or signifies idea s 10 which stem from the pre-existing objects in the world as felt by the senses.
However, Saussure, both as an anti-nomenclaturist and an anti-surrogationalist takes to task the erroneous idea of pre-existing thought independently of language. He insists that language is the upshot of human social interaction and hence central to the understanding of the world.According to him, language is by no means strings of names matching with a series of things, but a set of structures that segment the outside world.In his own words，
“Psychologically, setting aside its expression in words, our thought is simply a vague, shapeless mass. No ideas are established in advance, and nothing is distinct, before the introduction of linguistic structure.”（Saussure 1983：155）On the basis of such simultaneity of thought and
language, together with an element of truth that he sees in the nomenclaturism and surrogationalis m 11 —the dual nature of linguistic units：sound and concept, Saussure, puts forth his own distinctive theory of signs which are themselves the union of a series of sound-images matched with a series of mental entities, i. e.concept s 12 This unified duality theory of the sign is of paramount importance， 13 since on it, a word or a sign, to be more precise, no longer has to be“explained by reference to the thought it allegedly expresses nor the thought in turn to be explained by reference to some‘object'or feature of the external world which it mentally‘represents'”（Harris 1988：29）.To demonstrate the indivisibility of the two sides of the sign, Saussure invokes two such analogies as the recto and verso of a sheet of paper（Saussure 1983：57）as well as the contact between air and water（Saussure 1983：155）.Thus according to his framework of value， 14 the two sides of the coin, the phonetic side and the conceptual side, separately speaking, have their own mechanisms capable of assigning meanings to the constituent parts in their respective systems or subsystems.Such devices, as the structuralist or semiological view brings to light, operate on oppositions and differences.Specifically speaking, language as a system segmenting the amorphous mass out there and the shapeless and indistinctive thought（the vague and uncharted nebula）is meaningful itself in that linguistic items, either on the phonic level or on the morphemic level get their meanings through their oppositions or relative positions with others in their respective systems.For instance, the phoneme/a/acquires its value in relation to/e/，/i/，/o/and/u/in English and the word red obtains its value through opposition s 15 to other terms like yellow, green, blue etc.in the spectrum.Their meanings are defined, as Saussure argues, negatively, i.e.“not through their intrinsic value but through their relative position”（Saussure 1959：118）.So for red, whatever distinguishes itself from others constitutes it, namely, red is what others（such as yellow, green, blue）are not.To quote Saussure again，
“In language there are only differences without positive terms（stress is Saussure's original）. Whether we take the signified or the signifier, language has neither idea nor sounds that exist before the linguistic system, but only conceptual and phonic differences that have issued from the system.”（Saussure 1959：120-121）
So from the conclusion that Saussure draws, language evidently is a sum total of relations that create meanings. The most typical relation types as Saussure names them are associative or paradigmatic relations and syntagmatic ones.The former, characteristically absent in the linear presentation, refers to the type of relations in which elements can be replaced and differentiated by one another to acquire meaning；the latter is the linear combination of signs, each of which derives its meaning from the pattern in which it is positioned—i.e.by reference to what precedes or succeeds it.
As is common to all, language as a medium used in communication is necessarily the collective property to all members of a particular speech community, meanwhile it is an individual's command at his disposal. Therefore, in this section, we will confine ourselves to the discussion of how Saussure's distinctive characterization of language makes it possible that language readily lends itself to being such a common property.As made explicit in the Course, Saussure contends that the（community members'）common internalisation of the same system is possible due to a social fact—being membership of the same speech community.That is two or more individuals internalize the same system in a passive fashion.To use his own words：
“One's langue is acquired willy-nilly from one's social surroundings. Speakers have no choice.Language is a product of passively registered by the individual.”（Saussure 1916：30）
Besides this social explanation—the same language is imprinted in the brain of every speaker in the community because each speaker's membership in the community, Saussure goes deeper by exploring what inherent properties that a language ought to possess to facilitate the process that members of a speech community could not choose but internalize or be passively imposed the identical system used in their community. He asserts at length that：
“If language is to be an effective vehicle of communication, it must take the form of a totality of imprints in everyone's brain……Thus it is something which is in each individual but is more or less common to all. At the same time, it is out of the reach of any deliberate inference.”（Saussure 1916：38）
This, as may be argued, would appear vacuous unless Saussure could furnish a set of characterizations of language, the social institutional and psychological fact, illustrating how it could be simultaneously mastered by each member as the same as any other and escapable from“the control by the will, whether of individual or of society”（Saussure 1916：34）. Therefore，（an）intrinsic feature（s）of language that can make it capable of its equal imposition on the collective psychologies of speakers and independent of individual agent seem（s）the most appealling and hence the search for and depiction of them（it）appears fundamental and of paramount importance.
The primary one that Saussure provides and believes to be capable of guatanteeing a common internalization is the feature of arbitrariness in language. He regards it as“the organizing principle for the whole of linguistics”（Saussure 1916：100），upon which a system of signs builds itself as an organic language.This arbitrariness, as shown in the previous sections, finds its reflection specifically in signs（the connection in each of them between two constituents）comprising a linguistic system.And each sign according to Saussure is a duality of signifiant（acoustic-image）and a signifie（concept），or a“combination of form and content”（Cheng 1993：32）.The connection between the components in the sign is arbitrary.So no individual has any reason for preferring one sound-image to any other forms for a particular concept，“[for]a language as a system of arbitrary signs，……there is no fair ground for discussion”（Saussure 1916：107）.In addition, each side of the union, like each side of a sheet of paper, is internally related to the other side and to the identity of the whole.“[A]concept（signifie）becomes an identifying characteristic of a certain sound（signifiant），just as a given sound is an identifying characteristic of the corresponding concept”（Saussure 1916：144-5）.Hence, to change either the signifiant or the signifie leads to the alteration of the identity of the sign as a whole.Evidently, as Saussure maintains, knowledge of the sign means knowledge of each of the two and linguistic fact of their connection.It is therefore implausible for one to know the sign without knowledge of its meaning.Nor is it possible for two or more people to disagree about the signified（concept）of a signifier（sound-image）thanks to the close interdependency of the two sides of the compound.Thus the unified duality（unite a deux face）（Saussure 1916：45）of the sign make it possible for the unity to escape the voluntary control of the individual speakers, to“elude control by the will”or more precisely to be one psychological reality collectively imprinted on each individual in the speech community.
Furthermore, the structural holism, the logical construct（cf. Culler 1976：15；Taylor 1992：87）arbitrariness in the sign, also adds to the plausibility of a linguistic intersubjectivit y 16 in the sense that negativeness epistemologically facilitates perception and cognition.Saussure demonstrates cogently that any linguistic element is determined by its place in the holistic system.In other words, it is language as a whole which determines the sign and its value, because“the sign by itself would have no meaning of its own”（Saussure 1916：79）and“language is a system which admits no other[external]order than its own.”（Saussure 1916：43）And in reference to each of the component sides of the union, according to Saussure, neither sound image nor concept is positively decided as having an independent character or content, yet each is negatively defined by their respective difference from all the others in their respective system or as Saussure himself puts it：
“What characterizes each[sound-image or concept alike]most exactly is being whatever the others are not”
（Saussure 1916：162）.“A language system[a form not a substance]is a series of phonetic differences matched with a series of conceptual differences……”（Saussure 1916：167）Given this fact, a linguistic sign composed of a sound-image
and a concept, each of which is differentially defined, is an identity（a unified duality）determined by the language as a whole rather than by any external force as an atomistically defined element.Neither is it capable of being circumscribed, as discussed in the last section, in light of substantive properties of its dual constituents：sound-image or concept.Therefore, it follows from this that the identity of a sign is only determined holistically in the system to which it belongs as a member；and that the sign can only exists as a“unified duality”.To put an end to this chapter, we hold as it follows from the above consideration that linguistic meaning is formal in nature rather than substantive.And Saussure’s insistence on the compound of sound-image and sense produces two layers of value.One is that the dual nature of the signs extricates language that are made of them from the identity problem that perplexes the naturalist code model theorists（for details see Chapter III, Part II）.The other is that incorporation of meaningful propertie s 17 into the language itself makes it“a system of signs that express ideas”（Saussure 1958：16）and a communicative tool with its own form and meaning（content）whose acquisition, as Saussure indicates, is through formal relations between the signs.As a result，“Language is form, not substance”（Saussure 1959：112）or content which it is a surrogate for.[And as to the critical issue of whether the signified is identical to concept, we will leave it until the discussion of the function of linguistic signs in communication in Chapter VI.，Part III]In the first place, like naturalist type of code theory, structuralist code theory also takes it for granted that communicational understanding does occur and that its occurrence depends on the use of a shared（internalized）language code.And, unlike naturalist who assumes that the explanatory foundation for linguistic intersubjectivity to be found in nature, thus attempting to identify those facts of human nature which guarantee the common internalization of the same code, the structuralist, however, starts with the assumption that it is because of a social fact—being membership of the same speech community—that two or more individuals internalize（in a passive way）the same code.As Saussure has made it explicit：
“One's langue is acquired willy-nilly from one's social surroundings. Speakers have no choice.Language is a product of passively registered by the individual.”（Saussure 1916：30）
And：
“If we wish to demonstrate that the rules those in a community accept are imposed on them, and are not freely agreed to, it is a language which offers the most striking proof.”（Saussure 1916：104）
Besides this social explanation—the same language is imprinted in the brain of every speaker in the community because each speaker's membership in the community, Saussure goes deeper by exploring what inherent properties that a language ought to possess to facilitate the process that members of a speech community could not choose but internalize or be passively imposed the identical system used in their community. He thinks this perspective is more relevant and necessary because attention paid to the contingent properties of language as the naturalists have done will distract linguists from the proper task entrusted upon them.With this spirit in mind, Saussure asserts at length that：
“If language is to be an effective vehicle of communication, it must take the form of a totality of imprints in everyone’s brain……Thus it is something which is in each individual but is more or less common to all.At the same time, it is out of the reach of any deliberate inference.”（Saussure 1916：38）
This would appear vacuous unless Saussure could furnish a set of characterizations of language, the social institutional and psychological fact, illustrating how it could be simultaneously mastered by each member as the same as any other and escapable from“the control by the will, whether of individual or of society”（Saussure 1916：34）. Therefore，（an）intrinsic feature（s）of language that can make it capable of its equal imposition on the collective psychologies of speakers and independent of individual agent seem（s）the most appealling and hence the search for and depiction of them（it）appears fundamental and of paramount importance.The one that Saussure provides is the feature of arbitrariness in language, which he regards as“the organizing principle for the whole of linguistics”（Saussure 1916：100），and upon which a system of signs build as an organic language.
This arbitrariness finds its reflection specifically in signs（the connection in each of them between two constituents）comprising a linguistic system. And each sign according Saussure is a duality of signifiant（acoustic-image）and a signifie（concept），or a“combination of form and content”（Cheng 1993：32）.The connection between the components in the sign is arbitrary, so individual has any reason for preferring one sound-image or any other form a particular concept，“[for]a language as a system of arbitrary sign……there is no fair ground for discussion”（Saussure 1916：107）.
In addition, each side of the union, like each side of a sheet of paper, is internally related to the other side and to the identity of the whole.“[A]concept（signifie）becomes an identifying characteristic of a certain sound（signifiant），just as a given sound is an identifying characteristic of the corresponding concept”（Saussure 1916：144-5）. Hence, to change either the signifiant or the signifie leads to the alteration of the identity of the sign as a whole.Evidently as Saussure maintains, knowledge of the sign means knowledge of each of the two and linguistic fact of their connection.It is therefore implausible for one to know the sign without knowledge of its meaning.Nor is it possible for two or more people to disagree about the signifie（concept）of a signified（sound-image）thanks to the close interdependency of the two sides of the compound.Thus the unified duality（unite a deux face）（Saussure 1916：45）of the sign make it possible for the unity to escape the voluntary control of the individual speakers to“elude control by the will”or more precisely to be one psychological reality collectively imprinted on each individual in the speech community.This is clearly a contradistinction to Condillac's accusation of the sign's arbitrariness of resulting in failure of communication.
Whereas this arbitrariness in the sign proves to be essential for Saussure for it, along with its logical construct（cf. Culler 1976：15；Taylor 1992：87），the structural holism, demonstrate that any linguistic element is determined by its place in the holistic system.In other words, It is language as a whole which determines the sign and its value, because“the sign by itself would have no meaning of its own”（Saussure 1916：79）and“language is a system which admits no other[external]order than its own.”（Saussure 1916：43）
Moreover, in reference to each of the component sides of the union, according to Saussure, neither sound image nor concept is positively decided as having an independent character or content, yet each is negatively defined by their respective difference from all the others in their respective system or as Saussure himself puts it：
“What characterizes each[sound-image or concept alike]most exactly is being whatever the others are not.”（Saussure 1916：162）
“A language system[a form not a substance]is a series of phonetic differences matched with a series of conceptual differences……Although signifie and signifiant are each, in isolation, purely differential and negative, their combination is a fact of a positive nature.It is, indeed, the only order of facts linguistic structure comprises.For the essential function of a language as an institution is precisely to maintain these senses of differences in parallel.”（Saussure 1916：167）
Given this fact, a linguistic sign composed of a sound-image and a concept, each of which is differentially defined, is an identity（a unified duality）determined by the language as a whole rather than by any external force as an atomistically defined element. Neither is it capable of being circumscribed in light of substantive properties of its dual constituents：sound-image or concept.Therefore, it follows from this that the identity of a sign is only determined holistically in the system to which it belongs as a member；and that the sign can only exists as a“unified duality”.
In the second place, as to the content of communication, Saussure, though like Condillac, starts his argument with the premise of ordinary occurrence of communication. Nonetheless, he has come out with not only a novel semiological definition of language（as formal differential system of relations）capable of being imposed equally on the collective psychologies of a community of speakers, but also a distinct notion of what communication is as a result of that unique characterization.
According to Saussure's exposition of his speech circuit, closed in effect，“communication seems to be“a mechanical process of“transmission”and“association”[Saussure's associtive relation]”[and]this process of communication is complete when the same mechanism process in B's head has run their course and the corresponding‘concept'has been produced：B then understands what A said to him.”（Farrow 1992：135；see Harris 1988：114）.
Wittgenstein（1953）offers an example of communication between a builder and an assistant. He sets his criterion of success in communication as the agreement in definition as well as agreement in judgment i.e.practical consequence（action）following speech.According to this, an instance of successful communication is that the assistant sets off and brings back the required object if the builder asks his assistant to fetch a piece of block in terms of uttering“Block”.However, in accordance with Saussure’s criterion, once the identification by the master and his helper of the sign is set up, communication is already complete—well before the assistant’s satisfaction of the request via fetching the material.
Saussure and those in the Saussurian line of thought, i. e.those who take a structuralist and semiologcal prospect, including Chomsky the post-Saussurian structuralism， 18 view language as“a self-contained whole”（Saussure 1959：9），the homogeneous object extracted from the confused mass.This is, in reality, the logical and necessary consequence of the structuralist and semiological prospect taken, because structuralism inherently entails the idea of wholeness or internal coherence and self-regulating.The explanation of it makes no appeals to things outside the structure itself.One of the great contributions that Saussure makes to mankind is the anti-atomistic, anti-Positivistic and holistic perspective to the world, which encourages people to“go behind the actions or objects themselves to the system of rules and relations which enable them to have meaning”（Culler 197（8）6：19）.And in the case of linguistic study in particular，“Saussure's great insight is that at the heart of language lies a structured interrelationship of elements charaterizable as an autonomous system”（Newmeyer1986：4）.Such an approach has remained dominant ever since then and is even strengthened by the Chomskyan autonomy of syntax.This trend is so influential that the central hypotheses in the study of language remains that“it is productive to view the main core of language as self-contained SYSTEM（emphasis in the original），which can be‘characterized independently of the society, culture, personalities, beliefs etc.of the speaker of the language'（Moore and Carling 1982：166，216）”（Newmeyer 1984：123）.
However, such a long period of dominance on the part of the autonomous linguistics is not without resistences, criticisms or challenges. As a matter of fact, from the inception of the theory onward, there have been criticisms, attacks of various kinds.The following part will be devoted to some of controversial issues in relation to the Saussurian and Chomskyan tradition.
The first sort of criticism or reaction against the autonomous standpoint stems from those functional or sociological linguists who pose questions regarding what they think the precarious divisibility of, possibility in or necessity for what Saussure calls the confused mass. Among them, the most influential one that come to mind first is Martinet, the French functionalis t 19 who is usually grouped in the Prague School, though he was never affiliated in that group.Martinet, by regarding communication as the basic function of language, broadens the scope of linguistics, suggesting that what Saussure had left out of account be taken into account.He contends that the object of linguistics should not only the norm underlying toutes les autres manifestations du langage, that the true and unique object of the discipline ought to be human language activity, to be accurate in Saussure’s term, langage， 20 from and through which linguistic system emerges as a result of analysis of linguistic facts that are determined through function filtering（criblage functionnel）rather than by criteria of langue.
Similarly, Jakobson and Halle， 21 when redefining the area of linguistic description, consider as the object of the inquiry the Saussurian langage，“a complex duality of both system and process operating in a social context and subject to variation and change”（Widdowson，1979：112）or, to use their reiterated wording，“language in all its aspects, language in operation, language in drift, language in the nascent state, and language in dissolution”（Jakobson and Halle 1956：55）. The third one is Firth（1957），whose objection to Saussure's notion of language goes at length as follows：
“The multiplicity of social roles we have to play as members of a race, nation, class, family, school, club, as sons, brothers, lovers, fathers, workers, church goers, golfers……involves also a degree of linguistic specialization. Unity of language is the most of fugitive of all unities, whether it be historical, geographical, national or personal.There is no such thing as une langue une and there never has been”（Firth 1957：29）.To be precise, Firth finds little use in those dualities, stating that“it[is]unnecessary to operate with mind/body, mentalism/mechanism, word/idea and such dualism.”（cited from Halliday 1978：51）
The association that Firth's above assertion arouses leads us to Hockett（1968），the fourth one to cite. He, in a similar vein, along with other American structural linguists, finds it difficult to impose a clear-cut dichotomy on the basis of observable graded data.Collectively, they simply disconfirmed the langue/parole separation, regarding the former as no more than a set of‘habits’deducible directly from speech behaviour.Such an approval of the distinction results in the central issue separating the autonomous linguists, Chomsky in particular, who has, as Hockett claims，“reduced natural language to a well-defined derived system, an artificial language”（cf.Widdowson 1979：113）from Hockett himself who insists that“the things left out of account to achieve an approximatio n 22 of this particular sort are just the most important properties of human language, in that they are the source of its openness”（Hockett 1968：62）.Hockett further notes squarely that“since language is ill-defined[not well-defined]，mathematical linguistics in the form of algebraic grammar is mistaken”（Hockett 1968：61）
Another opposition to the classic dichotomies of langue conceived of as a static system（Widdowson 1979：113）or idealized form of natural language（Halliday 1978：88）and parole as well as competence conceived of as the ideal speaker-hearer's knowledge of such a system（Widdowson：ibid.）and performance, finds its illustration in Halliday's theory of language（1978）. In his theory, language or linguistic behaviour is seen as a component（and an essential one）of the social semiotic system, being subject to social-cultural environment.Halliday contends that the only distinction, if there has to be one， 23 ought to be drawn between meaning potential（what the speaker can say）and its particular actual choices from the linguistic resource in the system of network in a given context.Halliday, unlike other functionalists who take an instrumental approach to language with main interest in the use of language to explore something else, i.e.the relation between actual meaning, text and context（see de Beaugrande 1991：225）instead of in the nature of language itself, supports a complementary perspective of the two, one being taken by Chomsky to“study language for the purpose of throwing light on language”，and the other being taken by functionalists to“study language for the purpose of throwing light on something else.”（Halliday 1978：51）
Still another objection to the autonomy of language is over the failure in consideration of language users and context of situation. The first of this is concerned with language users, for example, Moore and Caroling（1982）accuse Chomsky of setting up a deductively formulated theory with common underlying principles explaining apparently disparate phenomena（see Moore and Carling 1982：2）.Such an opinion, they hold, is attributable to the Saussurian conviction that“Language can be separated from local instances of language in use and viewed as a system, self-contained and common to all language users.”（Moore and Carling 1982：64）They forcibly argue that“language is inextrinsically bound up with users and their experience, expectations of the world and that language is an epiphenomenon on the accumulated and generalized experience of its users”（Moore and Carling 1982：215）.Language thus, for its inseparability from its users'data store—knowledge assimilated—is an imperfect instrument, serving only as catalyst or trigger in communication.And the role that they take language to play in communication, as Newmeyer puts it, is to“initiate within understanders a complex series of processing mechanism s 24 which are bound up with their states at the time of processing.”（1984：124）
On the basis of such realization, they view language as one not idealizing away from language users. And hence the principal condition for the successful operation of language on any occasion is that“interlocutors have a similar perception of the particular area of experience or concepts of the world to which the linguistic exchange relates”（Moore and Carling 1982：172）.They even go so far as to assert that language conveys nothing.It acts only as a medium whereby users“cause another to access his own store of the accumulated and generalized knowledge of experience to locate what appears to make sense of the sounds he hears”（Moore and Carling 1982：165-169）.As a result of this analysis, they conclude that if language is believed to enable its users to“gain access in part to what they already know, however imperfect, then knowing more about what they know is clearly relevant to the study of language that makes crucial use of the‘complex framework of the knowledge experience, expectations, attitudes and beliefs that language users have and to a limited extent, share'（Moore and Carling 1982：103）.”（Newmeyer 1984：123）
Another aspect of this objection is related to contextual situation and is offered by those context-oriented linguists such as Dietrich and Graumann（1989）who, pay special attention to the influence of context of situation on the structure of language used as a tool of expressing ideas.They observe that the linguistic means used to express some particular ideas invariably vary with parameters in situation.And this reality has been overlooked by the structuralist theorists, who concentrate solely on the idealized system formed at an idealized point of time in history.Such a paradigm, they argue, risks exhausting something of essential importance, i.e.something relevant to language though this something does not belong to the well-defined object, because the structuralist description of linguistic items or strings of signs, as Saussure demonstrates, is confined to, or centers upon, that of the constituents with a particular system or subsystem.
To summarize for the present, it follows from the above illustration and analysis that the central idea in objections to the linguistic autonomy is over the issue of whether language at large is divisible into dichotomic parts, or whether such division is necessary or whether it ought to be distinguishable in terms of system and its use, the latter of which involves factors of various sorts such as the speaker's encyclopedic knowledge as well as contextual information. 25 Or put differently, the question at issue is between whether language used in the sense of langue by the autonomous linguists can be studied independently of parole or whether language used in the sense of langage by the non-autonomous linguists can be studied as a means of communicatio n 26 between people or“as an adaptabl e 27 instrument of human interaction”（Widdowson 1979：114）.Or whether language itself can convey what its users desire to express.
But as far as we are concerned, the above-mentioned critics are descriptively in the right as regards the autonomy of language, but explanatorily mistaken and not to the point, because nearly all of them are blind to the fact that if language, the abstract linguistic system cannot be extracted from the confounded mass of speech facts, they cannot obtain a clear picture of what language really is like in the strictest sense of the term. That is, in what sense language is a tool, as loosely and randomly selected forms of words or as an organic system?In this sense are they not likely to be thought of as committing the similar error like Saussure's nineteenth century predecessors did?Such erroneous commitment or rather nebulously confounded idea of language derives from, in our opinion, their inability to be aware of the perplexing phenomenon that language as a system“has no concrete existence of its own, except in the piecemeal manifestations that speech affords”（Hawkes 1977：20）.To be fair, all of us, except the few insightful figures, are prone to be at a loss about this puzzle, because as Cheng's（1991）ingenuity reveals，
“With language, the differentiation of the tool from the artifact, and the differentiation of the thing we say from the thing we say about are difficult to make—because the tool（language）is not anywhere to be found outside of the artifact（utterances）and very often the thing we say about（as in the case of ideas）is not to be found outside of what we say（again utterance）.So naturally utterances are easily taken to be language.”（Cheng 1991：125-126）
Making things worse, language, being a collective treasure of individual imposition，“remains incomplete at the level of each isolated individual：a language does not exist perfectly except in the speaking mass”（Barthes 1964：16）. Or, to use Hawkes'metaphor, language is the larger mass of the iceberg and speech is only the tip；thus language never appears in its totality，“but only in the incomplete performance of part of the repertoire by individual speakers'（1977：20）.Worse still, a more troublesome aspect is the conflation in the practice of parole of the linguistic meaning of the medium and the content expressed in this medium. 28 In spite of its intangibility or having“no natural form of expression except in its product—utterances”（Cheng 1991：126），there is no denying the existence of such a tool, even if in a metaphysical form, which“provides the basic operators and rules of operation for speech as arithmetic does for calculation”（Cheng 1991：126）.The incompleteness and metaphysical existence of the linguistic system in its users'instances of actualization highlight Saussure's（1916）deduction and Cheng's（1991）lucid and cogent account of that puzzling data, which are a great indication of their creativity and originality.
In the following we will commit ourselves on the issue of linguistic self-containedness since we take the afore-mentioned discussion to be off the point.Our approach to the problem is at a rather fundamental level—i.e.in terms of the working mechanisms of sub-systems within the language system.As is evident, Saussure’s structuralist view of language stipulates the closed, self-sufficient or self-defining nature of it, which is characteristically inherent in structuralism.In other words, a description of the system must be carried out inwardly, i.e.with reference to itself not to what lies outside of it or beyond it.Thus, language, being a system as Saussure reinforces, consists of“interdepending terms in which the value of each term results solely from the simultaneous presence of the others”（1959：114）.And the interdependent terms, to be precise, signs are themselves, the union of interdependent signifier and signified.The relation between them is arbitrary. 29 This is, as generally acknowledged, one of Saussure’s profound views of language.The principle of oppositions equally applies to the phonological level, the morphological level and the semantic one.This being the case，“what characterizes each[signifier and signified]most exactly is being whatever the others are not”（Saussure 1916：162）.Therefore，“a language is a system of a series of phonetic differences matched with a series of conceptual differences”（Saussure 1916：167）.Given this stipulation, the number of signifiers on the expression side of the sign, if we grant as Saussure intends that there is a one-to-one relation between the two sides, will have to be indefinitely large.And this enormity will give rise to a series of conflicts that escape Saussure’s attention and examination and hence threatens the structuralist coherence of the perspective—i.e.linguistic autonomy.One of the conflicts is that if language is to be the common property of all speakers, adults and children alike, namely if it is to able to be imposed on them as imprints within a given speech community, it cannot choose but to be within reach of their cognitive abilities such as memory, aural perceptibility, etc.Even if it is within the reach of everyone as desired, it has to be further“simple and immutable enough”（Cheng 1995）in terms of both the number and structure, lexical and syntactic alike, since language is not the only entity to which they apply their cognitive abilities during their limited life time.Therefore, the supposed requirement for a large number of what signifies（sound-images）appear difficult to be met.Otherwise, language, so far the most efficient communication tool, will fail to live up to what it is set up to.
Another conflict consists in the required amount of signifying items and the limitation of human beings'psychological attribute, to wit, audio production devices such as the vocal organs. Even though, one may argue, it is allegedly or even practically possible for human beings to produce much greater number of noises than actually needed in a particular natural linguistic system in which each of the chosen sound or sound pattern acquires its value in opposition to others, it is still impossible to satisfy the potentially immense needs.Finally and most significantly, the clash also lies in the natural psychological or physical inertia on the part of the language users who always desire to reduce energy cost by using fewer or more general linguistic units for the unlimited needs to express or communicate.Such economy-oriented tendency indicates that even if human beings are able to produce as many as possible distinctive units or phonemes, to use Martinet's（1960）term, that could come into the linguistic system, they tend to refrain from doing so for fear of complicating the otherwise simplistic system which rightly guarantees their ordinary communication.Consequently, homonymous and equivocal elements both at the morphological or lexical levels or at the syntactic level, as is the usual case, abound in language.And the basic distinctive units are limited in number in a given language.Though limited, they can produce a rightly needed amount of significant unit s 30 through double articulation, which operates fairly well and can satisfy the communicative needs of speakers of that given language.
Then to solve the homonymous and ambiguous factors rife in language?Or, to be more exact, how to settle the difficulties aroused by the oppositions between phonemes as are used in Martinet’s sense, which poses menace to the structuralist perspective of language and which, as linguistic facts exhibit, brings forth much hindrance to communication via language?
For this, if low level phonological and morphological oppositions are to no avail, Saussure and other formalist linguists seem to resort to another level which is supposed to work equally efficiently, though Saussure himself dedicates little to it.Saussure’s explicit claim in his book is to the effect that“the entire linguistic system can be reduced to and explained in terms of a theory of syntagmatic and paradigmatic relations”（see Culler 197（8）6：61）and specifically“the clearest assertion of what may be called the structuralist view of language[is]not simply that a language is a system of elements that are wholly defined by their relations to one another within the system，[……]but that the linguistic system consists of different levels of structure；at each level one can identify elements that contrast with one another and combine with other element to form a higher-level units, and the principle of structure at each level are fundamentally the same.”（Culler 197（8）6：61）
This level hierarchical in nature is called syntagmatic relations in which elements preceding or succeeding a given homonym may be drawn upon in terms of grammatical means such as word class or collocational features. Suppose for instance, in phrases（1），（2）and（3），
（1）saw a saw（our own construction）
（2）in the bank
（3）on the bank（see Cheng's manuscript）
In（1），the first saw is a verb, belonging to a different paradigmatic system than the nominal saw, thus causing no confusion；whereas in（2）and（3），though both occurrences of the bank are nominal, the prepositions preceding it settle the ambiguity, for normally in implies that the bank is the financial institution and on, the margin of a river. Such hierarchical approach first appears to serve some and good purposes, and indeed it does.Yet it does not fare so well before lexical and syntactical homonymous instances like（4），（5）and（6）：
（4）He went to the bank.
（5）She couldn't bear children.
（6）We need more realistic philosophers.
Without doubt, the problem of whether bank in（4）refers to a financial institution or to a margin of a waterway or even a slope of hill, or bear in（5）means put up with or give birth to, or more in（6）modifies the noun following it an adjective or the adjective immediately after it as an adverb, is not soluble within the boundary of the structure. That is the case where linguistic knowledge fails to work.
Additionally, there are other cases where linguistic knowledg e 31 about certain critical units fulfills no purpose. For instance，（7）and（8）；
（7）We didn't get down from the elephant.
（8）He watched her duck.
Evidently, no clues can be found to help determine whether down in（7）is a noun or an adverb, or her in（8）is a genitive case or an accusative one. But the solubility of such instances are rather great once some means of another kind is resorted to.Suppose if（7）and（8）are followed by（9）and（10）respectively，
（9）We got down from the goose.
（10）And it swam across the lake.
then everything is immediately clear for non-linguistic information derived from their respective fragment followed help to settle the equivocation involved. What is this effective means?Inference made on the basis of knowledge, of course.Moreover there are, due to stylistic variations, in language a large number of synonyms， 32 namely different sound patterns for identical or nearly identical concepts which purports to be an extra burden for human beings’mnemonic and psychophysical capabilities.
Up till now, following from these instantiations at least in this respect, we can come to a conclusion that language is not self-contained. Then how does it work so well as a communicative tool?As it will be seen from the discussion throughout this book, it is the intelligent users that make it so.This being the case, one may argue, does this imply that the inherent defect in the structuralism as reflected in the linguistic homonymy and ambiguity would put it in danger of being abandoned for a more viable prospective which has not yet been found?It is human beings'intelligence that salvages the structuralist approach from being invalid， 33 though structuralism may pertain quite suitably to other symbolic systems such as Morse code or high-way code which are either paradigmatically, syntactically or semantically simple enough.But for natural language such as English, which is like none of the above-mentioned systems, will not work if it is not comprised for or utilized by intelligent beings like humans.Imagine if it is meant for computers, however sophisticated they may be, it is bound to fail.This has been evidenced from findings in computer science and Artificial Intelligence.
Finally, as a complement to the present chapter or as a comment on Saussure's delineation of langue and parole, we will offer a few remarks worth making.It is now a fact that Saussure’s distinction between langue and parole results in more profound investigation in language, and his treatment of its study as a branch of semiology“for the first time[has]succeeded in assigning linguistics its place among the sciences”（Saussure 1959：34）.But Saussure’s separation also gives rise to some questions regarding the scope of language and parole.There seems to be something that belongs to language but Saussure leaves out for a purely differential and psychological definition.
Hjelmslev（1959）（see Barthes（1964：17-18）；Culler（197（8）6：95-97）），on the basis of accepting Saussure's conception, inserts in the intervening area between the two ends two more terms such as norm, usage and replaces Saussure's langue with his own tern, schema. As a result, Saussure's distinction becomes a four-way pattern：schema（langue），norm, usage and parole.By schema, Hjelmslev means the pure form in the Saussurian sense of langue, in which abstract structures, items are defined in formal relations and with no reference to phonic substance；norm, however, denotes a materialistic side in language, for instance, that/p/is a voiceless bilabial stop and/b/is a voiced bilabial stop is a fact in the English language.Such a fact is not up to individual preference；finally usage, as a set of habits prevailing in a given linguistic community, is statistically regular, and subject to individual choice（see Barthes（1964：17）；Culler（197（8）6：96））. 34 Culler（ibid.）cites various disputes about the nature of langue among different schools of thought.According to him, Hjelmslev, for instance, in his theory of glossmatics and his followers treat language as a purely abstract schema；linguists in the Prague School include schema and norm in their treatment of language, claiming that phonological distinctive features should be described in articulatory terms and that the distinctive oppositions set forth by Jakobson“are not only abstract features but norms of physical or phonetic realization”（Culler 197（8）6：97）；whereas Daniel Johns and his supporters regard language as the triple set of schema, norm and usage.By defining“the phoneme as a family of sounds they thus include‘usage within language’.”（Culler 197（8）6：97）
Additionally, by the same token, Culler himself thinks that Saussure's criteri a 35 for separating langue and parole are precarious because“they divide language not in the same way and they leave much room for dispute”（Culler 197（8）6：96）. He accuses Saussure of his abstract and formal concept of langue and his relegating sound to parole.Being in complete agreement to the Prague School linguists and Daniel Johns, Culler further stresses that other acoustic features have to be admitted to language，“since differences between accents and pronunciation have a psychological reality for speakers of a language”（Culler 197（8）6：96）.He demonstrates that the English will not essentially the same language as Saussure suggests even if its units were expressed in some other way.
Diver, one of the leading figures in the American Columbia Schoo l 36 （see Cheng 1986，1992（1982）：3）also mounts attacks on the pure psychological conception of the Saussurian‘sound-image'，and assigns physical property to the signifier, making a sign a combination of a sound（with physical feature）and meaning. Consequently from all the above discussion, Saussure's assertion：“English will be essentially the same language even if its units were expressed in some other way”will be held precarious since it is not comprehensive to define language solely in terms of formal relations.
Another major aspect of challenge to Saussure's purely differential and oppositional nature of language is the obscurity perceived by Chomsky, namely syntax.In contrast to his detailed discussion of the paradigmatic relations, Saussure is rather weak at the level of syntagmatic relation description in that he seems to be at a loss as to what belongs to langue and what belongs to parole.In defining parole，‘the executive side of language’Saussure refers to it as“the combinations by which the speaker uses the code of the linguistic system in order to express his own thoughts[and]the psychological mechanisms which permit him to externalize these combinations”（Saussure 1959：14）.It seems from this quotation that Saussure relegates combinations of linguistic elements to parole, which gives positive phonic and psychological exhibition.Actually, Saussure himself perceives some of the difficulties involved in his treatment of syntagm, saying that“there is no clear-cut boundary between faits de la langue, which are examples of collective usage, and faits de parole, which depend on the free choice of the individual”（Saussure 1959：125）.He also realizes that in language there exist some fixed syntagms such as idiomatic phrases and“groups of words built on regular patterns”（Saussure 1959：125）.The existence of such syntagmatic constructions on regular form which are not free to individual choice and the combinative freedom in speech perplexes Saussure.The source of the perplex, as Chomsky points out, is his consideration of language“as basically a store of signs with their grammatical properties, that is, a store of word-like elements, fixed phrases, and perhaps, certain limited phrase types.He was thus quite unable to come to grips with the recursive processes underlying sentence formation, and he appears to regard sentence formation as a matter of parole rather than langue, of free and voluntary creation rather than systematic rule.There is no place in his scheme for‘rule-governed creativity’of the kind involved in the ordinary every use of language”（Chomsky 1964：23）.Culler（197（8）6）argues in more relevant terms：“It is precisely because Saussure recognizes the creativity of ordinary language use that he was unwilling to include sentence formation in the langue”（Culler 197（8）6：99）.Thus according to him, Saussure is unable to reconcile the fact that we can produce totally new sentences with the fact that a language contains phrase types.Saussure fails to realize that language is more than a system of interrelated units.It also includes the non-differential linguistic facts, rules that are finite in number but can generate structural descriptions for an infinite number of sentences, which can be judged in terms of well-formedness by the hearer who may have never heard them before.
In our opinion, the crux consists in the Saussure's inability to discern language as a system, thought expressed in that medium and the social activities formed through using the system as a tool. As a result, Saussure's narrow definition of language in terms of formal oppositions and differences seems to be of necessity broadened to incorporate into it some non-differentiating linguistic elements such as rules and material acoustic properties for a complete notion of language.
1 Here structuralism is used to refer to the American structuralism as represented by Bloomfield and Harris.
2 Saussure has no intention to underrate the importance of the latter parts in the binary pairs listed here.
3 Saussure（1983：82）notes in the Course that in the 17th century, language was taken to be a picture of an image of thought by the Port Royal grammarians who sought through a study of language to discover a universal logic.Their grammar was wholly atemporal and Saussure posits that such a point of view is irreproachable, because Saussure believes“the 17th century grammarians had a well-defined object, knew what they were doing and did not confuse synchronic and diachronic studies, though their practice was wanting in many other respects”（Saussure 1983：82）.
4 The separation of the two, as Saussure notes, is only a theoretical requirement.Actually, they are dialectically related or alternatively in a relation of reciprocal comprehensiveness（see Barthes（1964：16）and Hawkes（1977：9））.
5 Barthes（1964：11）realizes the great influence of the Saussurian linguistic semiology.He thinks it is so important that he proposes that linguistics is not a part of the general science of signs but inversely semiology is part of linguistics.
6 Levi-Strauss first calls the Saussurian notion of system structuralism though Saussure himself never uses the term.Levi-Strauss also applies structualism to the investigation of anthropology and sociology.
7 If a language is studied as a structure in which everything is determined within itself and with no reference to what is outside the structure, then the language is a closed system.So Saussure should not be misunderstood as saying language is a closed system in terms of its number of signs available.In this regard, language is instead actually an open-ended system, in which there is no limit to the number of signs that might come into the system and form a part in the system.
8 The extraction of langue from parole is often misunderstood as“merely a stipulation”（Newmeyer 1986：3）for the sheer purpose of isolating an object for the linguistic study and establish it as an independent science.As it shows here, the linguistic system is a logical and necessary consequence of the arbitrary nature of signs which property has gained wide consensus among philosophers of language and linguists.
9 There is often a misconception that Saussure, as he is often accused of, ignores the study of the linguistic evolution in time.We think otherwise, for synchronic study also shed light on diachronic study.Diachronic changes, as Saussure himself holds, is the alternation from one synchrony to the other.
10 The identity of ideas or concepts for every language users in a community will be discussed in Chapter VI.
11 According to the nomenclaturist view, language is a list of names for a list of things.A name or a word has meaning because it stands for something else.And this something else can be either a vocal entity or a psychological one.Saussure draws upon this element of appropriateness and sets up the theory of sign duality.
12 Later on, Saussure insists on using another pair that he thinks is more appropriate, the signifier and signified in place of the former pair—acoustic-image and concept（see Xin Delin 1993）.The significance of the substitution will be discussed in Chapter V.
13 One of the significances is that it guarantees the identity of the signifieds for all speakers in a given speech community.This will be further discussed in Chapter VI.
14 Saussure does not use the term meaning.In its stead, he borrows the term value from economics to illustrate the phenomenon of linguistic meaning in signs in a given language system.
15 Roman Jakobson calls Saussure the greatest promotor of linguistic oppositions（le grand revelateur des antinomies linguistiques）（cited from Xu Guozhang 1983：12）
16 The linguistic intersubjectivity refers to the same language indispensable for communication as the term used by Condillac（1746，1947，1948 and 1981）.
17 By contrast, American structuralists leave meaning to psychologists.They concentrate on forms, hoping the solution of forms will shed light on that of meaning, the semantic component of the sign.
18 Due to the fact that Chomsky’s Syntactic Structure（1957）bases itself on the essential notion of structuralism, Chomsky can also be identified as a structuralist linguist.But this is called the European structuralism as against the American Bloomfieldian structuralism.In fact, Chomsky and his followers are often referred to as such by commentators from outside the linguistic field（cf.Newmeyer 1986：4）.
19 Martinet calls himself a realist actually.He does not regard himself as a structuralist.
20 Langage is another term used in Saussure’s Course which includes both langue and parole.It is often understood as the collective activities performed with the linguistic tool.
21 To our knowledge, Jakobson and Halle’s（1965）opposition to the Saussurian autonomous linguistics is largely concerned with the confinement of linguistic discipline to langue.In fact such an indictment is a little bit unfair to Saussure because, in his defense, though he is devoted to setting up the status of linguistics as a science, he is not oblivious to the importance of parole.This is evidenced by Godel’s（1957：119，181）and Mauro’s（1972：476）judgment that the closing line of the Course, the often quoted line by structuralist linguists is only the editors’imposition on the Course as one of Saussure’s central ideas.In reality, Saussure also shows much interest in social relations（see Xu Guozhang 1983：9）and moreover, as Wells（1958）understands him, Saussure after deciding the object of linguistic inquiry, intends to study it as well as its related aspects.Only in this way, that is，“by concentrating first on the synchronic systems which they form, then on the diachronic relations between these systems, and lastly on parole, can linguistics found a rational and unifying order whether or not it proves practical”（Wells 1958：15）.
22 The term approximation refers to Chomsky’s reduction of language to a well-defined object for his own purpose.
23 Halliday（1970：145）with respect to Chomsky’s distinction of competence and performance, believes that such a“dichotomy runs the risk of being either unnecessary or misleading：“unnecessary if it is just another name for the distinction between what we have seen able to describe in grammar and what we have not, and misleading in many other interpretations”（145）.And later on Utaker（1979）makes some more pertinent remarks, which we believe, can shed more light and more to the point.He believes that the separation is unnecessary if competence only signifies the grammatical model, and misleading since it presupposes another distinction between language and non-language——that ought to be made explicit.（1979：110）.
24 Moore and Carling unfortunately make no specification about the nature of these mechanisms.
25 Saussure describes parole as heterogeneous.He shows that he is aware of the variety of elements affecting language use.Chomsky（1964：52），too, in the same line of thought, notices the complex interplay of many factors involved in the actual use of language.he says among others the grammatical process is only one.
26 Communication has a two-fold implications, one being the phatic or emotive exchange, namely for establishing good rapport between people, the other being the true sense of the word, conveying ideas.it seems to us that the non-autonomous linguists listed here refer to it in the first sense.
27 By the term adaptable, to our understanding, the non-autonomous linguists refer to the subjection to the restrictions of social norms, rules or factors, for they hold that language use is a social behavior.It is not the sense we are using in this thesis.
28 This topic will be taken up again and dealt with more seriously in Chapter VI.
29 The significance of the signs being arbitrary is offered in Chapter VI.
30 American Spanish, for instance, has only 21 distinctive units and it produces, as
Barthes（1964：39）shows，100，000 significant units.
31 Other instances such as homographs are also phonocentrically troublesome, though grammatologically discernible.
32 Heteronyms phonocentrically causes no trouble in this regard.
33 This topic will be taken up again in greater detail and with special reference to the other side of the sign, namely the signified in Chapter VI.
34 Barthes cites the French/r/as it is pronounced in some regions for explication and Culler uses the charting of the frequency of different pronunciation or of other uses of linguistic elements as exemplification of usage.
35 These criteria refer to Saussure’s statement in separating langue from parole, the essential from the contingent, the social from the individual and the psychological from the material.
36 According to Cheng（198（9）2：1），linguists in the American Columbia School are more like the European structuralists than the American structuralists owning to the influences from Martinet, Jakobson who once worked in Columbia University.