Cultural Transmission and the Linguistic Construction of Reality
by Hugh M. Lewis
An alternative theory of symbolic linguistics is proposed herein. It is beyond the scope of this chapter to fully elucidate this theory, and so this task is saved for a companion piece. Here it is important only to outline the basics of this theory, and to demonstrate unequivocally how this theory ties in with a larger framework of the cultural construction of reality. I deliberately make some radical assertions about normal language structure that are in a sense anti-paradigmatic. This has been primarily in counter-reference to Structural Linguistics as this has been paradigmatically preponderant in the American academic contexts especially. Structural theory was based upon a mathematical conceptioning of human language that was not fully sufficient with the actual natural patterning that language takes. Furthermore, it alienated the description of language patterning fundamentally from the social and dialogical contexts of its natural production and articulation. Finally, it provided little room for tolerance of what linguistic relativities may occur in language, and therefore had no tolerance for any possible variation of language pattern that did not fit its own narrow parameters of transformational grammar. Therefore it also had no tolerance for any linguist who did not practice a strict orthodoxy of grammatical transformationalism. Since the inception of structuralist theory, other grammars have arisen, particularly beyond the academic context in which structuralist theory has had a strong hold (I should say strangle hold). These alternative grammars and theories may be more or less amenable to the theory of symbolic linguistics offered herein. I construct this theory derived from my own anthropological background and research, therefore it reflects strongly a tradition of anthropological linguistics particularly as this was founded by Edward Sapir and later developed by the Chicago School of Descriptive Linguistics.
Language is central and indespensible to the symbolic functioning of the human brain. It permits a form of associational patterning that links outer world and inner experience with an iconographic "thing" (the morpheme) that acts as a symbolic bridge and touchstone of that experience, and can also serve as a substitute for experience.
This refers to the "nomic" function of language in symbolization, which serves to disambiguate meaning in figure-ground pattern recognition. The word becomes a symbol conferring form and definite shape and function to thoughts and meanings, and serves simultaneously as a marker for a presumably external reference (though in many instances it may actually point to nothing external in the world.) It is in this function of the word, of naming, that enables the gestalt pattern to be achieved as a sudden sense of holistic integration. It entails most often the precipitation of the "correct" or suitable name that confers holistic meaning to a pattern. It enables as well, and no less importantly, the hypostatization and concretization of constructed, otherwise implicit only, understandings.
Thus the nomic function of language in the symbolic organization, mediation, construction, and articulation of human cultural reality must be taken fully into account in our theories of human systems, as in a sense it constitutes a core part of these theories.
Language works in another way in the nomic processes of externalization and internalization of information, by way of their affixation of a term to a definition, and the connotative accretion of differentiated meanings. Such words achieve a nomic concretization that gives them solidity and objectivity of meaning in the external world and a sense of coherence and subjective validity in the internal world. Thus words serve to fix and precipitate meanings out in fairly patterned ways. Empirical evidence supports the contention that the internal codification of words and meanings is fairly hardwired and critical to the cognitive organization of the mind by which meanings and associations can be referenced and called up by means of the vocabulary and external stimuli to which they are linked.
The linkage of words into ordered systems by principles of constraint enable the construction of elaborate systems of rationalization by which we seek explanation and understanding of reality. Our generalizations and theories and ideologies are all symbolic systems of nomic ratiocination. The differentiation of language that is determinative of an average or modal level of speech competence is thus directly tied to cognitive styles and degrees of sophistication of modes of thinking and psychologizing about the world. These are naturally and culturally in terms that are more-or-less well differentiated and permit relative levels of symbolic articulation of reality.
The central question of world-view is that of the critical role and status of language in the symbolic mediation, integration and construction of cultural reality. Language underlies how we view our world and make sense of it both daily and in grander senses, as well as encodes our socially based knowledge of the world. Cultural construction of the world, as we have come to know it, would not have been possible without language. But language happens paradoxically to be as constrained by cultural patterning and cognition as it is predetermining of these human informational patterns. These are feedback processes of mutual, nonlinear constraint. To look for simple cause-effect determinations is to be misguided in our inquiry.
Any anthropological theory of language in relation to the symbolic structuring of worldview must solve basic problems: the question of language change and stability, language structure and patterning, and the central and strategic social functionality of language in the human construction of reality. Language is holistically integrated as a symbolic system that communicates meaning though its achieved integration is always limited and never complete. Symbolic integration of reality is necessary for any language to function effectively as a communication system, and, paradoxically, a language must function effectively in order for symbolic integration of reality to happen.
The function of language is that of accomplishing cultural transmission, which is the same thing as symbolic communication of cultural patterning and information. The structure of language patterning is therefore predetermined by and derivative of that function. Language structure is fundamentally arbitrary, and therefore culturally defined by conventional constraints and not universal.
Similarity of linguistic structure between any two languages is indicative of either historical relationship between the languages or of some convergence of pattern that has happened by chance or borrowing, which is not uncommon.
It is not incorrect to claim a holistic sense of linguistic-cultural-cognitive relativism, or what can be called the relativity of worldview. This form of relativity is neither deterministically strong, nor weak, as this is conventionally construed. It is merely a general statement that people interpret things in different ways, and language serves to parse and construct the view of the world in different ways. To the extent that interpretation is inherent to perception, we can say that people "see" things differently, but we must say that pure mechanical perception is universally similar. Because people talk about what they see, and probably need to talk about it in order to interpret it in any significant degree, therefore we can claim that different linguistic patterns lead to different views of the world.
On the other hand, there is great categorical and emblematic and even propositional consonance and overlap between different language structures, and this great overlap can be accounted for by many reasons. All languages serve similar functions, and function similarly. The biological substrate of human language is universal to human identity, though each voice or particular linguistic system may be uniquely patterned to each individual human being. The categories and meanings that language comes to parse, symbolically embody and integrate to some extent are natural or shared cultural categories--hence the facility and frequent similarity between linguistic system. Some evidence suggests furthermore that all extant languages today may have ultimately been derived from a single ancestral proto-language or precursor. In a logical sense, just as with the evolutionary development of life, so also with the historical cultural development of languages. All languages ultimately had to have come from a single ancestor. Thus there is some minimal presupposition of structural similarity operating in all current languages.
Language cannot develop or evolve in a complete linguistic vacuum. A person who is born and raised in absolute deprivation of any social intercourse, will not spontaneously produce a language, and cannot be expected to produce a coherent language alone. It is impossible, because the essential requirement for such production, that of having someone to talk to, is missing. If two people were isolated together, in context to one another, without a language, then it is possible and to be expected that they will try to communicate and would work out some kind of language pattern between them in time.
Language is therefore fundamentally a social process of interaction and communication. This inherent sociability of language as symbolic communication entails that language structure must be historically and culturally arbitrary, and precludes the possibility of its innateness in human psychic structure.
The minimum requirement of linguistic structure is therefore patterned agreement, or rules of agreement that are part of an implicit social contract that two people will work by and follow the same rules of agreement in their communication. Agreement rules constitute implicit structures that underlie both the syntactic, semantic and pragmatic patterning of language.
Agreement rules lead to rules of performance that are the conventional forms of expression of agreement rules in language patterning. Performance rules can be interpreted linguistically as descriptive performance rules or else they can be interpreted by some set of grammatical standards for any one language, that can be referred to as prescriptive performances rules. In any given social context, probably both forms of rule patterning are operating simultaneously. Thus, the entire bulk of language patterning, in whatever form or shape it occurs in, can be interpreted as derivative variations of fundamental usage rules that have been agreed upon as an implicit language contract within a coherent speech community.
There is no need therefore to impute or superimpose an abstracted model of some innate or universal mathematical structure of usage. This does not mean that abstracted models do not apply in the description of linguistic structures, and are important to the understanding of universal language patterning, but we should not confuse the abstraction with the actual patterning of language itself nor with its developmental history.
Violation of agreement rules, whether accidentally or deliberately, is any noticeable kind of error that makes communication impossible or misunderstood (inefficient) in any context. Both descriptive and prescriptive rules exist that determine the degree of patterned constraint exhibited by any language. Generally, all languages have a minimal degree of patterned constraint necessary for sufficient and efficient communication to be achieved. To a large extent, this system of constraint is variable, dependent largely upon situational contexts within which it is applied.
Thus there is substantial variation between communities of speakers, or sub-communities, especially when they have had time to work out their own differential calculus of these agreement rules. The purpose of agreement rules is to maintain some standards of communicative efficacy in the encoding of language. I believe there are usually at least two sets of standards operating simultaneously, and there is some implicit optimal standard between these two that is approached but rarely perfectly achieved.
The first set of rules can be called core performance rules and these govern the minimum requirements necessary for sufficient communication to occur at all. Violation of these rules is basic and leads to obvious ambiguity of pattern in language.
The second set of rules can be referred to as derivative or elaborative performance rules, and these govern the maximal requirement of producing effective communication. These rules are more variable and hence more flexible and violable than core rules. Usually, strict prescriptive standards imposed on a language are attempts to ossify the secondary derivative patterning of linguistic expression into some restrictive framework. More natural and descriptive examples of this patterning occur in many of the pragmatic functions of language, as for instance, in creating and maintaining in-group/out-group or asymmetrical interpersonal identity in competitive relations, which seems like a common and pervasive patterning in linguistic usage.
The optimal standards are standards, I believe, that are achieved in everyday linguistic praxis within whatever context language gains expression. This can be defined as implicit rule patterning guiding an optimal attainment of efficient communication. It is neither necessarily over-elaborated as in derivative rules, nor so parsimonious as in basic communicative demands. It is perhaps best evident in the rhetorical and common logical standards that writers and orators strive for in linguistic communication, since these types of people have much invested in their linguistic performance. This has a great deal to do with style and concision of expression--saying the most with the least.
There are both direct and indirect cultural constraints or implicit sanctions against violation of rules of agreement at whatever level or in any context. We generally know different kinds of rules by their violation, and we learn these rules by their repetition and practice, and by their application and misapplication. Children are not so much automatic acquisition devices, with built-in sets of basic rules in their brains, as they are naturally adept at learning these basic rule patterns in language as they are applied within the life-world contexts. They are daily, by continuous trail and error, working out these patterns in their rapidly developing brains. This is clearly evident especially with children of about 4 to 7 years of age. Important in this process are models and significant others who can serve to correct the patterning. Most language learning is learned through modeling and through interaction. Children therefore pick up the dialects and socio-lects of their parents and their peers, and the linguistic mannerisms of the media. This is to be expected as a natural outcome. It also means that what a child learns will be fairly discrete to a give place and period.
The argument that a Russian child adopted into an American household upon birth, will learn some discrete variety of American English as the mother-tongue because that child carries the same rule structures that lead to transformations in either Russian or English is a ridiculous and inherently unfalsifiable deduction to make. Furthermore, it is unnecessary because it makes greater sense and is more parsimonious to assume a more direct explanation that a child learns the structure of whatever language he or she is born into within the natural context that language is articulated in. All languages have and create their own contexts, otherwise languages are dead or were never even born.
If such a priori transformational rules existed in the human brain, then we should expect that once we worked them out, we could easily apply them in the inter-translation between Russian and English, and this would be perfect or practically perfect in every way. So far, in foreign language study, I have not seen a set of universal rules that permit us to accomplish inter-translation between any languages. Translation remains a difficult process of working out the similar rules of agreement and performance that occurs in a foreign language, and then, just like a child, to relearn how to apply new rules in the second language. This is why adult language learners almost never entirely give up their native accent when working in a foreign language, and rarely gain "native speaker fluency" in the competent performance of the second language.
And this gives us a clue about primary acquisition and native language development. Children's entire language apparatus, including the brain, and the plasticity that occurs with these organs, becomes developed in primary language acquisition around certain rules of usage in a particular language, and not any language. Their organs are developed around this patterning, including even their organs of speech, such that its production seems natural and automatically reflexive, and in a sense, it is, as this is what children become even in an organic sense as they grow up.
Rules of agreement follow lines differentiated along certain functional syntagmatic and paradigmatic categories within which all words are placed, like a table structure or complex matrix. This is a minimal paradigm for language patterning. Generally, syntagmatic categories are ordered by placement rules, of what can come next, and what comes first. Paradigmatic categories are ordered by associational and analogical rules that are largely culturally or idiosyncratically defined, of which logical associations are a subset. These rules define what things are similar or different, and what things are complementary or contradictory.
The distinction between syntagmatic and paradigmatic may be something of a spurious dichotomy in fact. All words, and all instances and variations of all words, always have both syntagmatic and paradigmatic aspects about them, and these are dependent upon the implicit relational context in which they occur in natural language patterning. It therefore makes more sense to say that words in differing contexts form "natural sets" that are defined along both syntagmatic and paradigmatic conventions. In general, it can be said that though the same word may be used in many different ways, in a specific context in a specific set a specific word, and kind of word, can take only one distinctive form.
There is a sense that on a very basic level syntagmatic categories are in a sense paradigmatic categories as well. Because all words always take some inherent syntagmatic dimension of meaning, we may say with equal validity that all paradigmatic categories become syntagmatic as well. Syntagmatic categories merely point to the functional syntactic integration of words of different paradigmatic categories, whereas paradigmatic categories merely point to the functional semantic integration of words occupying different syntagmatic categories. We know in a minimal sense that syntagmatic categories are usually sequentially arranged in time, or in a temporal patterning of what comes next and what follows in a feed-forward process of language production. We know as well that paradigmatic categories are usually spatial arranged in some virtual or possible sense, of what alternates may possible occur at any point of articulation. This kind of dilemma arises from the inherent design feature of language known as duality of patterning and is the basis for the symbolic transformationalism of language.
In general, we can say therefore that syntagmatic patterning governs the particular sign structure or mechanical patterning or form that a word will take in any given context. These are in a strict sense syntactical and morphophonemic in character. They govern the production, reiteration and parsing of sound patterns and the linking of sound patterns in string structures.
Paradigmatic rules govern the particular contextual or configuration value of words, and therefore are in a strict sense, the symbolic value of meaning that is attached to words. They govern what I will call the relational value of sound signals.
All words and sounds produced in a language must have both syntagmatic and paradigmatic value simultaneously. Sounds that lack either or both are essentially meaningless. This is why, when the sounds of foreign languages fall on one's ears, they sound meaningless. Essentially, the patterning of foreign languages that are not "understood" falls upon deaf ears.
Generally, I would say, that core performance rules govern the syntagmatic patterning that a language takes, while derivative performance rules govern the paradigmatic aspects of a language. I would say that optimal or stylistic performance rules operate upon both levels simultaneously, and always serve some implicitly pragmatic function. It is like trying to write good poetry. Stylistic rules are in a sense strategic of game rules, and always involve pragmatics in the selective use of language.
The tripartite structure of language patterning, replicating the tripartite structure of the human brain as a symbolic organ, is therefore implied. The syntactic patterns govern syntagmatic construction of languages, paradigmatic patterns govern semantic and symbolic relations of language particles and phrases, and stylistic rules and patterns govern pragmatic functions of language production in social life.
Rules of performance and agreement contain another subset of rules that I will call transformational rules and transition rules. In their most minimal form, rules of transformation and transition govern the discrete identity of words as they may occur in different contexts, and the switching that occurs between contexts. (This is not to be confused with code-switching that is an interesting example of the operation of such rules in a mixed system.)
Transformational rules work like this: for such and such a word to be used in this string in this context, this particular thing must also be done. Transition rules work like this: to switch from this current context to some other, some indication of the switch must be made that points to the new context and defines the relative value between the previous and next context. This can be as discrete and mechanical as a pause or break, or a changing intonation of sound, or it can be some signal that indicates the difference between a question or an exclamation. Both types of rules are what I would call "marking" rules as they govern patterns of "marking" or alteration of basic word patterns to give differential emphasis upon these words.
Transformational and transition rules would seem to imply some basic change process in language. While in a direct sense, we can say that such rules in their consistent application are not directly the sources of change in language as they promote coherence in language. But it can also be said as well that indirectly they permit variation of pattern that leads to language change.
I would also say that there is another variety of performance rule that operates normally in all language contexts, and these are what can be called "selection" rules. Selection rules govern the selection of words and the ordering of words in strings, and, on another level, the selection and ordering of strings themselves. To impose selection rules on performance and agreement may seem somewhat functionally tautological. All articulation in performance naturally entails some kind of selection. Thus, in a larger sense, all agreement rules are necessarily selection rules.
But I believe that selection rules are of a special species of agreement rule pattern that are different fundamentally from transition and transformational rules. They are rules in a special and restrictive sense that governs the range of possible alternation of structural patterning in on-going linguistic construction. They define the range what is appropriate within any given instance, and what would be inappropriate. They are thus parameters that define the inherent variability of language pattern in any context. We really only learn these rules especially in their violation, as we learn quickly and immediately what is inappropriate and what might be possible within any given linguistic context.
I would add, in relation to selection rules, another kind of rule pattern that occurs, and this is a set of contextual rules, or rules that define the contextual relations of words as they occur or are expected to occur, and that define the relations between contexts. Like selection rules, any rule of agreement or performance can be considered to be a contextualization rule in that they help to delimit and define the context itself, and make up a part of the context. But again there is good reason to assume that in a strict sense there occurs a subset of rules that govern contextual relationships in some normal and expectable manner.
I would state also, that this structural patterning of language is inherently contextual in the sense that it is always context bound. Natural language patterning always occurs in some "con-text" with language of similar patterning following similar sorts of agreement rules. It is we who study language, who tape record voices and write in books, who have elaborated a reified and eidectic sense of language as an abstract entity separate from the framework of its natural occurrence and production.
Therefore, the structural patterning of language always has both internal context and external context, both of which are governed by rules of agreement. Distinguishing between internal and external context is somewhat similar to distinguishing between the inner and outer world of symbolic integration, but in symbolic linguistics this has a more precise definition.
External context is primarily and minimally a two-person context, or a dialogical context that fulfills the communicative function of language. We can say that this is usually a two way transmission process, but strictly speaking, we can only refer to it as a one-way transmission. Internal context has specifically to do with the relative internal coherence of the language signal at all its levels of patterning. We can say that language always creates its own context both internally and externally, and we are not far from the truth in saying this. We can elaborate an extended model of internal and external linguistic context to embrace many paralinguistic features of the world. And in a sense, the interior reality can be considered to be paralinguistic to the extent that, at least from a human perspective, we see the entire world is fundamentally articulated, mediated, and symbolically precipitated in terms of language.
So far, distinguishing between different types of rules on a functional basis alone does not serve a model of linguistic generalization that would aim at simplifying our understanding of language pattern. Presupposing so many different kinds of rules would seem to make language an over-determined system when in fact it is naturally quite the opposite. Natural language is inherently open, variable, nonlinear in description, and therefore occurs as an underdetermined system. If it weren't then it wouldn't change as rapidly and pervasively as it does. We can state that the structure of all languages is therefore minimally determined by rules of agreement and performance that are culturally constrained.
We can of course attempt the formalization of rules of agreement and performance upon whatever level we encounter them. Thus we have rules of embedding and object agreement, or rules of normal order, or, for English, passifying of active structure. Generally, to expand our rules to encompass every describable instance in every possible context, either our rules are become overloaded with exceptions that they quickly become unwieldy, or else our generalizations become so rarefied that they are essentially empty and useless to almost any situational application.
To hope to arrive at a single, universal set of such rules that apply equally to all languages, is, I believe, a futile and virtually hopeless task, as any such abstracted system is bound to be spurious and supercilious. But this is not the same thing as saying that we cannot have a systematic scientific understanding of the patterning that all languages take.
The structural patterning of language is therefore both symbolic and contextual, and in this sense, it incorporates a kind of natural logic that is entire linguistic. Linguistic "logic" can be understood as the functional expression of differential rules of agreement through linguistic patterning in varying contexts of its expression, to enable cultural symbolisms to be transmitted in a minimally coherent manner. This logic is always at least implicit to the patterning of language that is meaningful. I call this fundamental linguistic logic a form relational logic because I believe that term best describes its structural patterning.
This relational logic is universal to all languages and makes possible the coherence of language patterning in any context. I do not believe it is inherited by people, but that people have inherited genetically the capacity for such patterning as this is realizable only through alternative modes of human language. To call it logic is to risk its being misconstrued as something that is connected directly to classical truth-value logic, or to some exotic new computer logic. This is to risk imposing certain constraints upon it that may not otherwise exist. I would call it a natural logic, because, though it is constructed only through and by means of human language, which implies a degree of arbitrariness, its expression is inherent to the functional and material organization of the brain itself.
Before elaborating a model of natural relational logic, it must be reiterated that rules of agreement and performance, in whatever form they make take, are not the same as this, and are not derived in any necessary way from relational logic. Relational logic I hypothesize underlies all language structure, because this patterning is symbolic and is tied to the organic functioning of the brain. Relational logic makes possible the translatability between languages. It puts all people, as human beings, in the same symbolic playing field, regardless of the differential rules they apply in playing the game. This relational logic has little to do, I would state, with rules governing the actual mechanical articulation of language. This is only variable and ultimately arbitrary within the boundaries of natural constraints. I would state, on the other hand, that it has a very great deal to do with meaning and the natural constraints of human meaning as this is symbolically expressed in human cultural patterning. Thus relational logic is universal to meaning, and not directly connected to language as a sign system of communication. It lies behind this system that serves therefore as a vehicle of its symbolic function, expression and transmission, but it is not the substance of what is expressed or transmitted, rather the organization of that substrate of meaning.
Thus, linguistic structure is nowhere a priori in an intrinsic sense, in that it is conventionally constructed in the world between people, and therefore has a history marked by its repeated violation and change. The universal translatability of all languages arises from the common ground upon which meaning is constructed in all cultures, and upon which our language systems always work and hoe and manipulate. Languages point to the same or at least very similar sorts and sets of things in the world albeit in different ways, no matter if we are doing our cultivating work in China or Spain or North America. This is the basis of what I would call the propositional-relational logic that underlies the symbolic integration of human reality. Linguistic logic, whatever form of expression it takes, is but one set of variants, albeit central to its articulation and development.
Relational logic serves the purposes of the ultimate reasons for linguistic communication. Meaning in the mind of the speaker must resonate clearly in a corresponding way within the mind of the listener, in a one way model. Relational logic in language is about how the meaning inscribed into linguistic signals by the speaker can be transmitted and reconstructed with some degree of symbolic efficacy, fidelity and consonance, in the mind of the listener. It describes the "meeting of minds" made possible only through language.
Syntagmatic, paradigmatic and pragmatic categories are interrelated systematically in strings on the basis of transformational rules that govern the linkage patterns possible between words. For a particular word of a particular kind to fall into a certain category, a discrete kind of transformational operation must be performed first. This transformational operation is largely a system of "marking" of signals. It therefore has a denotative and indexical reference function.
There is nothing innate about transformational rules, either syntactically or semantically defined. They are rules of agreement, and therefore are implicit conventions that are imposed upon a language system. The transformation of active to passive in English is nowhere obligatory or necessarily indicative of anything inherently universal or deep about language. Passive form is a sort of reversal of syntactic ordering that carries certain referential values contrasted with the more normal active form, and therefore requires marking that the active form does not normally carry. In general, the unmarked form is the more basic and the more prototypical of a set of alternates.
Closely related to transformational rules are implicit rule patterns that govern word change in a language. Word change is described as morphophonemic conditioning, and part of the agreement structure of any language is what forms of morphophonemic condition are allowed, and what are disallowed, and under what conditions. Morphophonemic condition involves several forms of word change processes.
The string structure of languages is really its basic phrase-structure. Sentential structure is one form of string structure. This phrase structure is also defined by cultural convention and implicit rules of agreement, and are part of the linguistic chunking process. The normal conventional chunking of linguistic code into small sets and subsets are a means of imposing a high level of constraint and predictability upon natural language processes.
I propose symbolic linguistics as an alternative study of natural human languages as symbolic systems that serve specific and general functions in the integration and articulation of human reality. Symbolic linguistics comes to focus upon the word as a prototypical symbolic form that takes on various meanings in different alternative structures. Words as sound symbols occur in systemic contexts of speech frames that are sentential constructions, more or less, implicit or explicit. A word as a speech event has little value outside some kind of socially derived con-text or cultural framework of meaning. Gesticulations as sounds are not uncommon, but they serve only a very narrow and limited function of making noise perhaps in a way that is tagmemic or indexically marked.
The paradox of language is that it is rooted in biological-based apparatus that makes speech and linguistic understanding possible, and is at the same time an inherently social phenomenon. We cannot have any language that is not at least a two-person process. This makes the patterning of language both inherently and deeply structural at the same time that it makes the same patterning in its articulation in social life a pattern of social process. To emphasize one or another of these issues in understanding language organization and process is to miss the point that these are part of a feedback system that describes the central process of human symbolization and cultural transmission.
The human mind depends upon language centrally to achieve coherence and integration of the world, and the individuals as members of some social group depend upon this linguistic based construction of the symbolic world in order to adapt and function in the world together successfully. The patterning of linguistic structure is therefore based upon this central functioning, and can be described as naturally symbolic in design.
To look for a central language acquisition device underlying a universal mathematical structure of all languages is to centrally miss the point that the entire human brain is largely organized around the problem of language. Hence a great deal of the brain is devoted centrally or peripherally to the production, acquisition and organization of linguistically encoded information.
To attempt to construe an innate and fundamental universal structure, as in transformational grammar, for all languages is to miss the point as well that the structure of language is necessarily and fundamentally symbolic. Therefore cultural patterning intrudes even upon very basic levels of language parsing, as this is a two-way process of communication. It involves normally both production and reception, which are fundamentally different but interdependent processes. In other words, language is not just about speaking or speech production. It is about human communication. The fact that languages can and have been culturally built without the central utilization of speech organs demonstrates, among other things, that oral language structure is nowhere innate and naturally obligatory.
No structure of language has to be genetically innate. Only the mechanisms for production and reception of language are preprogrammed in their biological functional design. Structure must be agreed upon, worked at and arrived at in social contexts. Any structure in language is nowhere obligatory, or else there would be no variability and language would not change at all.
Structure is arbitrary within the biological parameters set down for spoken communication between human beings. All languages are therefore constrained and prestructured by the same sets of biological design parameters. All languages, as functional systems of symbolic communication, are more or less sufficient to the tasks of such communication. All human beings have the same approximate capacity for language, whatever historical form language may take.
Like language itself, collective rule agreement does not start off as a deliberate agreement between people. Perhaps in the very beginning, there was only a single proto-language that a group of people had worked out. Perhaps this gave them such success in surviving that they quickly spread out and transmitted their language to others. Perhaps others borrowed their sounds and built their own systems. These early language systems were probably not very efficient and fairly context bound such that they did not travel so well. Probably only a few rules were required for their mastery and their vocabulary was very restrictive.
From this first break-through on, there have been few clear instances of language being created from a complete linguistic vacuum. I would say that once language was established and caught on among hominid ancestors, it developed rapidly, and led to a basic shift in selection regimes between different populations. It would have probably been more or less a one-time occurrence, not needing to be repeated too many times to make it stick.
Thus the evolution of language is not so unlike the evolution of life itself, such that once it began, it took hold and differentiated and refined itself in ever greater degrees. Thus, for almost any language group extant today, there has been more-or-less the same length of historical development from the very beginning. Probably many groups and their languages became extinct along the way.
There is a sense that when we produce speech, we listen to our own language and monitor what we say, but this always comes after the fact of its production. A different mental structure guides the production of speech, in the selection of words and their arrangement syntactically, as a feed-forward process with many closed loops, than what appears to guide the reception and parsing and evaluation of language as something that is aurally rendered.
For the most part, these processes have not been clearly distinguished, and have been treated as nearly one and the same sets of things. Thus it is somewhat misleading, like using natural selection to refer to chance processes in evolution, to talk about native speaker intuition and really be referring to native listener intuition, as if speaking and listening were necessarily the same thing. I would claim that reception is largely a process of hearing and listening, bolstered by visual and other kinds of sensory cues. I would claim that speech production and speaking is largely a separate kind of process made by the organs of speech, and does not necessarily involve audition as self-audition or self monitoring. What we hear resonates our thoughts of what we were intending to say. That we are sometimes embarrassed at ourselves by saying what we do not intend to say, is a common indication that audition does not guide speech in the same way that it guides reception of speech.
What is the point of this? The rule structures and content patterning characteristic of these two modalities of language are complementary in many respects, but are not necessarily isomorphic in all respects. Speech production is a symbolic process that is largely a pattern of systematic and spontaneous responses to diverse sets of stimuli. Speech is largely triggered by aural recognition processes that themselves are linked to intentionality and relational structures in the brain. That hearing and listening come before and underlies speech production is evident in natural patterns of language processing and acquisition by children, and reflects the symbolic development of the mind.
Of course, language as a natural symbolic system is one that is based on a finite set of design principles. A structural paradigm for such a system must be founded on the idea that there are certain inherent constraints in speech production and processing that must be accounted for and cannot be violated.
Also, language systems have an inherent equilibrium about them, such that they put premium values on issues of efficiency of communication and expressiveness of what is communicated. Efficient communication demands a certain degree of regularity and predictability of patterning, which implies agreement and well worked out performance rules of ordering in speech production. These rules of ordering are derived from basic semantic dimensions of functional classes of words as signal devices. Rules are implicit, and are only worked out by grammarians after the fact of their natural realization and gradual social reformulation.
We can say that expression of meaning is a second important value of linguistic equilibrium. There is no point in uttering a sound if that sound carries little significance, and it follows that we should like to attach as much significance as possible to the sounds that we make.
There appears to be a basic trade-off and principle of allocation in language patterning, such that efficiency of communication can only be maximized past some optimal level by means of reduction of sound patterning and the inherent expressiveness of a language. Expressiveness of language can only be increased beyond some threshold at the sacrifice of communicative efficiency.
Native speaker intuition is perhaps better referred to as native listener intuition. Both terms would point to different functions. Native speaker intuition would refer to the capacity for a speaker to produce coherent strings on the fly or off the top of one's head. It would refer to the speaker's ability to speak coherently without having to think about it before hand. If we had to think about how to say what we wanted to say, then it would prevent us from being able to spontaneously say what we think or think about what we say. Native speaker intuition would refer therefore to the ability to automatically fill in the gaps between imperfect thoughts and impure speech so as to render both reasonably correspondent with one another. In this regard, we seek some minimal degree of consonance and correspondence between our inner states and the expression of our words.
Native listener intuition, to split hairs, would refer to the ability of the average respondent to be able to put contextual closure and to make sense of the full meanings of the speaker's intentionality via limited and often imperfect means of speech. Both forms of linguist intuition are based on the same symbolic-cultural integrity that we build in our shared world. It is obvious that in foreign language contexts such intuition breaks down on any level, to the point of rendering the non-native speaker/listener somewhat debilitated in basic communication processes.
This sense of linguistic intuition points back to the issue of relational logic which I have hypothesized as universally underlying the meaning structure of all language. It highlights the systemic and informational aspects of linguistic patterning in the sense that Gregory Bateson spoke of, in terms of the basic predictability of whole pattern from only part-relationships. Such linguistic intuition can be considered as a measure of the native and natural symbolic integrity of pattern achieved by language in any particular context.
The hypothesis of relational linguistic logic relates language centrally to the worldview problem, which takes on qualities as the symbolic mediating system between internal cognitive worlds and external cultural contexts. In this sense, we can refer to meaning as embedded in language processing as inherently "propositional." The same patterning can be found in the structure of our cognition and in the structure of our cultural understandings relating to social relationships.
The functional aspect of such propositional organization of human information has to do with the fact that we must derive meaningful symbolic inferences about otherwise only implicit or unknown variables based upon what we do know. Our inherent experience of reality is ambiguous in a way fundamentally different than how an animal experiences ambiguity. A dog will sniff its world to find out what it doesn't know. We search symbolically with our minds.
Many of the propositional inferences that we must derive about our world have to do with the acceptance or plausibility or credibility of an hypothesized propositional relationship as implicitly or explicitly stated in our language. Even many of our propositions may be mostly implicit.
Underlying the propositional structure of relational logic can be found a reticulated quaternary structure of analogical associations, a pattern of loose substitutive association that arises fairly early and is very basic to more constrained systems of symbolic relation. Understanding of relational symbolic logic in language or in cultural or cognitive pattern begins with understanding this basic aspect of symbolic chaining and associations.
Relational linguistic logic proceeds from an understanding of the possible or potential value of any sign. The nomic function of language is based on the possibly value of any word as a "symbolic variable" that may be rendered into linguistic propositional equations about the world or in relation to the world. It arises from the fact that a sign is a symbol in the human sense that can carry an almost unlimited range and amount of meaning. If we speak of intentionality structures, and complementary frames of expectations, in the minds of our speaker/listener, then we can also speak of plausible inference structures that underlie and guide the patterning of attributable meanings in a person's life.
The central hypothesis of this theory of symbolic linguistics is what can be called the linguistic construction of reality. The symbolic function of human cognition is biologically tied to the human linguistic apparatus and organization of the brain for speech production and language pattern recognition. The evolutionary acquisition of language was rooted in and provided the symbolic foundation for the cultural construction of human reality.
Symbolisms, primarily as these are linguistically encoded and expressed, have what can be called a plausible or weak inference structure that is both determined on a sign level and undetermined on the symbolic level. It is variable in its restriction and directionality, guided by principles of relative contextuality and relative non-contradiction.
If we look at the fundamental patterning of the symbolic integration of reality, we will see that it is primarily linguistic in its relational structure, and that it ties together in a total framework of understanding that can be called a symbolic network. We call it worldview. Relational dynamics develop in the formulation of worldviews that leads to internal interference of symbolisms and to a form of dissonance and internal symbolic ambiguity within such systems. Of course, one's worldview may not be the same as another, and the differences between points of view are liable to be expressed linguistically in many basic ways, if there is any communication at all.
It can be seen that language is fundamentally implicated as an integral and central component of the structural-functional organization of symbolic reality. This central role of language defines its own symbolic structure, function and contextuality, and is implicated with the symbolic-cognitive process of externalization and internalization of reality and with socio-cultural processes of symbolic communication.
To a great extent, the internal structure of language is independent and arbitrary of its function in these symbolic processes of mediation (any language would more or less suffice) except that language becomes conventionally bound and constrained within the normal and expectable cultural contexts of its occurrence in important ways.
Much spoken language leaves the symbolic meaning of the message implicit, subject to linguistic intuition only, and as such is bound within the speech setting in which the message is situated. Language organization also comes to reflect the order and patterning of the knowledge in which it becomes contextually defined. Cultural categories may be left highly differentiated or remain relatively unrefined. Language delimits, denotes, and defines symbolic meaning as knowledge and understanding, helping to give it shape and form in the world in a way that is anchored to the actual experience of the communicative event.
Human propositional thinking is inherently relational in nature in that it is concerned foremost with:
1. The status of relationship of one thing to another thing, and of the part to a whole.
2. The ascription of identity of a particular instance or thing to membership to one or more distinct categories (x is a kind of Y or x is a member of Y).
This kind of thinking is intrinsically symbolic, and is intimately related to the construction of categories and symbolic representations of the world. The possible relations and kinds of things are numerous and vary between basic dimensions such as "stative/active", positional/temporal, size/count, etc.
Propositional understanding that is concerned with the relationships between things in the world is to some extent dependent upon the definition and ascribed categorical identity of those things being related. The identity of parts conditions the relationships of the whole. At the same time, the inferred relationship between the parts has an influence upon and conditions to some unknown extent the ascriptive identity of the parts. Thus our propositional understanding of reality cannot be gained independently of our understanding of either the categorical construction of the world or its articulation with actual cultural realities.
Propositional understanding is primarily concerned with the comprehension and credibility of a stated or implied relationship between two or more things or categories in the world, or of the ontological or categorical identity of one or more things in the world. Such propositional understanding can be considered to be a necessary form of reality testing that is pivotal to the feedback process between internalized schemas and models and actual experiences in the world. Such reality testing occurs continuously at the inter-sentential level of discursive strips in even casual conversations in which a topic of discussion is proposed, statements are made and contested, and the topic is revised or put away, while conversation shifts to another topic.
But such reality testing can even be construed at the sentential level of the normal phrase structure of language itself. Each phrase that has some kind of stop or end marker forms an inference frame that fits in which a larger system of inferential-referential construction.
A more complete description and analysis of string production, or symbolic string theory in language production at this level will not be undertaken in this chapter. Suffice it to mention in passing that it is based on a kind calculus of confidence that permits progressive and more restrictive determinations to be made within a field that is loosely underdetermined. The sentence is the level at which syntactic sign patterns articulate with meaning patterns in the production of linguistic symbolism.
Each word as a symbolic variable must be determined sufficiently or optimally in a restricted relational sense within the context produced by the sentential string or phrase structure. This determination defines its point syntactic, semantic and pragmatic value, or particularistic value within that string. The sentence as a whole takes on a propositional value that is synergistically more than the sum of its component parts. The string as a whole automatically borrows the point value of the word part to create a larger articulatory symbolic structure. Symbolic value is thus lifted to another level of meaning and pattern.
Thus the meaning of the individual word elements is critically determined within the internal sentential context, which is itself semi-determined within the external inter-sentential or conversational context that is itself embedded within some larger behavioral setting and dialogical framework that is culturally constrained. Holophrastic utterances and one word tagmemes can be understood therefore as part-whole sentences the structure and value of which are left underdetermined, embedded as these remain within the context of the expression. Such one-word expressions can be easily and readily expanded into longer and more explicit string structures.
In a sense, sentential string structure in language serves to make optimally explicit what would otherwise remain implicit in the communication process between two speaker/hearers. It provides a system of communicative transmission that allows the symbolic polynomial expansion of meaning within a sentential, and by extension, inter-sentential framework of understanding. All rules of agreement performance are in a sense the system for explicating meaning in minimal frameworks of expectation and intentionality.
As such string structure constitutes a kind of symbolic and cybernetic feedback system, not only between two speakers ideally, or rather a speaker and a listener, but also possibly between a speaker with oneself as listener. This feedback system is constrained internally by minimum and maximum standards of efficiency and expressiveness, respectively, at the level of sentential articulation. It is this constraint which confers to language its dynamic stability and flexible structuration.
The semantics of language at the sentential level can be understood in propositional terms (as can the semantics of language at the lexical level of the word and the phrase-structure level be understood in paradigmatic terms of categories.) In general, we may state that a basic sentence contains at least one noun (sometimes left implicit, as in "(you) Run!" as subject and at least one "relational value" (like a verb, also sometimes implicit, as in "You (are (it))!"). A question or a statement posed as an open remark invites a determination of the propositional truth-value of a subject-topic. A remark may be a corollary conditional statement or an auxiliary statement serving to qualify, demarcate and further contextualize a main subject-topic such that a central inference might be made.
String theory in language points to consideration of linguistic structure. It makes greater sense in this symbolic theory to speak of such structure as a process of structuration, and, more exactly and fitting, as "constructing" the world linguistically in terms of strings of set length. It has been a central point of this theory that such structuring of language is context bound, conventional, defined by arbitrary rules of performance agreement, and seeks certain kinds of functional equilibria in communicative efficacy. These are inherent to the natural design features of human language as a system of symbolic communication, and there is no reason to presume noumenal, a priori, mathematical structures in this kind of process to be embedded in the neural functioning and patterning of the human brain.
It follows that there can be no precise linear mathematical description of symbolic language. In any mathematical equation, there is always only one correct solution for any set of values. We thus can only approach natural language, and by extension, human intelligence, by means of more complex non-linear programming models, as have been found appropriate to other natural information systems.
In language, there are always multiple alternative possible solutions for the same set of values in the same string. Even if variables in a mathematical equation are inherently undefined and ambiguous, the explicit relation between variables is never left undetermined. In symbolic languages, not only are the values themselves left ambiguous, but the relations between values are also inherently ambiguous, therefore there can be no "determined" consequence of such a string.
The challenge of symbolic language, unlike the challenge of mathematical equations to determine the correct solution, is to disambiguate the meaning of the string in a sufficient manner to make communication efficacious. Disambiguation takes the place of systematic determination in mathematical models. This disambiguation is always relative, both to the internal context of the string itself, and to the external context in which the string is embedded, albeit the total cultural context in which the string is articulated.
Thus, string theory and relational logic determine a kind of phrase-functional grammar that permits some degree of systematic regularity to be superimposed upon the string, primarily through a system of marking rules at different levels of string analysis, or parsing. These marking conventions permit and facilitate the subsequent parsing of the string by the recipient in a manner that has fidelity to the original intentionality structure of the producer.
The philosophical implications of this critical distinction separating a mathematical description from a symbolic linguistic description are important. In the former case we can have an implicit one-to-one correspondence between the sign and the thing(s) it signifies. It indexically points to one thing (or one set of things.) There is an implied equivalence between sign and signified. In symbolic language structure such a presupposition does not hold. This entails that the relational logic underlying symbolic language is fundamentally different from the kind of two-value logic underlying a mathematical construction of the world, as there is not clear or necessary correspondence or equivalence between the sign and the signified. Therefore, it follows as well, that mathematically derived descriptions or models of language, while possible, are inherently insufficient to the description and analysis of linguistic structure. They approach this patterning by means of an integral calculus, but these remain improper integrals.
There is no room in this digression to fully develop more fully than this the theories of relational logic, propositional understanding and constructive string theory that is deserved in a full treatment of symbolic linguistics. It is necessary here primarily to understand that these are an intrinsic aspect of human language as a dual processing symbolic system that permits a level of symbolic patterning of information that is detached from the linear level of the sign and its production in the act of communication.
Symbolic linguistics is made possible by the fact of indirect displacement of reference from the sign to remote stimuli, and hence the resulting structure of inference that systematically relate the sign to its remote possible values.
A symbol, unlike a sign, is capable of incorporating an infinite range of meaning in its value, along with its own identity and the negation of its own identity. Unlike a simple sign, a symbol is therefore inherently ambiguous. Thus its expression requires a system that serves to efficiently disambiguate its value within the apprehended or implicit framework and contexts of its articulation.
Language plays a key active role in the process of the symbolic differentiation of the phenomenal field, and language is held to influence this process, as well as being a medium of expression. This leads to several points about language:
1. "Naming" or the "nomic" function of language is an integral part of symbolization that connects the object-sign to the effective response and the cognized signification. At the same time the named word serves to precipitate, reify and embody this symbolic relationship between inner and outer worlds.
2. Language serves in symbolically embedding experience in a coherent, intelligible way. Names and symbolisms become associated and networked with other names and connections, such that the patterning of language production and structure is a naturalistic expression of the symbolic organization and differentiation of human experience.
3. Linguistic acquisition and ethno-semantic organization of knowledge mirror the patterning and organization of the symbolic differentiation of experience, and can be utilized as an effective means for analyzing and mapping this development.
Language structure therefore influences and is influenced by this symbolic stratification of meaning. It comes to reflect this embedding/embodying process of symbolization, and the organization of both the internal and external patterning of symbolisms. It comes to reflect its duality of patterning as well, and is the principal mechanism of the symbolic articulation of reality that leads to this stratification of meaning in both our inner subjective and outer objective worlds.
Consideration of linguistic contextuality proceeds from an understanding of the relative dependency that linguistic encoding has upon the context, both internal to its own construction, and external to the world to which it points. The internal structure of language is relatively independent and arbitrary of its function in the symbolic processes of mediation, except on the surface level of its discursive and mechanical iteration. In the normal speech settings and contexts in which it regularly occurs and in its pragmatic and functional effects, it becomes bound within such contexts in important ways. A language and its structuration, dynamics and directionality of change are bound within and tethered inextricably to the dialogical contexts in which it arises and takes shape. It remains thus partially non-arbitrary to the extent that it is externally constrained by socio-cultural convention and context.
A great deal of spoken language leaves the symbolic meaning of the message implicitly bound within the speech setting in which the message is situated. Language organization also comes to reflect the order and patterning of the knowledge in which it is contextualized and its meaning given shape. Culturally defined categories may be highly elaborated or left relatively undifferentiated.
Language function must thus solve two interrelated problems. The first is the efficiency of communication that is linked to the disambiguation of meaning or the reduction of noise in the system. This is tied to the carrying capacity and sense of equilibrium found within such systems. The other problem is the issue of the expressiveness of the meaning in conveying the meaning intended. It can be considered to be the channel capacity of the string, or the informational loading or effective value it can carry. Efficiency and expressiveness are overlapping but dissimilar problems, and at some point the problem of efficiency of the communication leads to a direction of language structure that begins to be at odds with the problem of increasing meaning and load of the message.
Thus there is a tendency when the problem of efficiency, or reduction of ambiguity, toward the simplification and reduction of signal length to the minimum, for differential trade-offs to the opposite tendency of the enhancement of expressiveness toward the elaboration and extension of signal length towards the maximum.
There are definite speech settings and socio-cultural contexts in which one kind of function will consistently take precedence over the other. This occurs in inter-lingual speech settings where there will be a premium placed on the disambiguation and simplification of the signal. With in group symbolic communication events, especially those that are highly inbound and frequently oriented toward maintaining an exclusive group boundary within a context, will tend toward the expressive elaboration and extension of messages towards a greater degree of achieved coherence.
Languages vary considerably as well in the degree to which the lexicon is elaborated and the degree to which its lexicon and syntax are bound within the context of the normal speech setting. Languages thus are contextually relative. There is a relationship between efficiency and expressiveness and context boundedness and relative contextual independence. In general it can be said that messages situated and tied to the content of the communication event have the disambiguation of noise as an implicit aim, while expressive elaboration will tend towards relative contextual independence.
Languages that tend to be more contextually bound will thus have less highly differentiated categories and will thus reflect cultural knowledge domains in a different way. They will tend to be more concrete and practically oriented toward everyday life, hence they will appear more basic in the sense of relying heavily upon basic, simply encoded terms. Within such languages, the elaboration and expression of certain types of information, especially of a more fine-tuned range of discriminations, will tend to be relatively unavailable and not readily encoded within the language. A language encodes its own cultural categories of knowledge, rendering access to expression and concretization of symbolic understanding outside of those categories relatively problematic.
Highly differentiated languages and areas of language will tend to be less context-bound, in an external sense, and therefore more relatively context independent. They will contain a greater frequency of marked and elaborated terms and messages, and will permit more facile access to a wider range of discriminations and generalizations. Elaborated languages tend, on average, to embody and carry its own context. Any language, furthermore, will exhibit an internal range of variation within which there might be considerable variance in terms of its relative differentiation, elaboration and contextuality. Thus we can speak of a relative, variable level of average speaker competence of a language that is linked to a speech context.
It can be expected that syntactically in terms of basic dimensions, languages will also vary substantially in a similar manner. Some languages being more well developed syntactically reflecting its relative contextual independence compared to languages that are less syntactically constrained as demonstrated by systems of overt marking of covert categories of meaning. Thus less highly differentiated languages will tend to show fewer systems of over marking of such dimensions, and thus its syntactic systems will be less constrained tan those that have greater marking and more differentiation.
The question of inferring an average speaker competence in any language for any given area of discourse refers to a performance competence and this leads to some inherent ambiguity of meaning. In general, linguistic performance refers to the parole of speech production and praxis in everyday speech events, while linguistic competence refers to the capacity of the average speaker to function in any context with equivalent results. The presumption of linguistic competence refers to the notion of Langue that brings us back to a priori, idealized structures of language.
At this point, we must see linguistic structure as inherently and inextricably embedded in social context of the discursive speech event. We may properly speak of the phenomenological process of linguistic construction that is guided by certain indirectly implicit and embedded conventions rather than as a "structure" per se. The rules that guide linguistic structure and its redundancy of pattern in construction in the maintenance of its communicative efficacy, are ultimately social products of shared constraints. They are "performance" rules that meet certain conventional criteria of social functionality.
We must separate the question of linguistic performance from the notion of linguistic competence, and tacitness, the notion of the parole of the actual speaker-hearer versus the langue of the ideal speaker-hearer. This kind of dichotomy is a measure of the differential between what a language could have said, and perhaps should have said, and what it actually did say, in any given setting.
Because the only way to measure competency is to judge performance it is not clear to me if this is a very useful and non-trivial kind of dichotomization to make, unless of course we are intending to build our cloud houses of eidectic structures. Of course, anyone could have said anything, and intended to say or mean almost anything else. The fact of the inherent sociality and social situation of all language as performance precludes the possibility of language being primarily a psychological phenomenon of some solipsistic, idealized structure.
Competency and Langue are rooted in the inherent, biologically based and evolutionarily rooted human capacity for speech production and language itself. It is a capacity that all human beings, by definition of their nature, share, just as we normally have two hands, two eyes, two feet, etc. From this fact, and from the common constraints placed upon all natural languages, there arises the notion of virtual equivalency of all languages to meet, more or less well, the need of human communication. Certain design features of language, such as openness, productivity, symbolization, duality of pattern, prevarication, are inherent and implicit in the definition of any and all human language--as a system of communication cannot be without these features and be regarded as a fully functional human language.
We must note and take exception with the idea of the ideal Structure of language, especially when this structure must somehow be genetically hard-wired to the brain of the ideal speaker-hearer. We are led by this notion to search for a mathematically correct universal grammar that sees language as a kind of perfect communication machine. This is the point at which we must discriminate between the critical difference of our own generalized and formulaic constructions of abstracted language and the informal and naturalistic constructions of functional language.
A deeper level of language patterning and structure linked to the problem of language production can be said to have less connection and linkage to the contexts of its performance, which are thought to occur more upon the surface patterning. Such a deeper level of structure can be said to be linked to the inherent design features of human language, as a symbol communication system, and are for the most part shared by all languages. Nevertheless, at this more implicit level two patterns of context and constraint can be found to exist, and they may influence the directions in which a language will change.
When we refer to the dynamics of language change, we can recognize the implicit structure of language by the systematic nature of these changes, and we must acknowledge both exogenous and endogenous sources for such change, and the partially structured nature of such sources. Languages change from without in contact and relation to other languages, and from within in relation to internalized variations of language pattern and the reverberations that context and changing contexts may have upon its implicit patterning.
Language change and variation tends to be systematic, even to the point of being scientifically describable and predictable. It largely arises from inter-speaker variability of language that leads to differential rule patterns underlying language. Conventions of agreement surrounding linguistic configurations shift in their balance over time. This determines and is determined by what catches on in everyday usage of language and what dies a natural linguistic death.
Language change must therefore be construed from the standpoint of language stability. In this, we can describe a kind of equilibrium of language that describes its homeostasis about some optimum level of performance. This defines the range of variability of pattern as well. Societies have a certain investment in the traditional stability of their language. They can tolerate so much language change, but not so much as to cause great ambiguity of its patterning, especially not at multiple levels. To rapid linguistic change becomes disruptive for a cultural system, and is a sign of disruption of such a system.
Structure and change in language are caught in a dynamic dialectic around the problem of increasing coherence and decreasing entropy. In human language this becomes critical in spoken discourse when we can properly speak of the phenomena of linguistic transmission of information. Spoken discourse must effect a resonance not only at the linguistic level of a string of signs, but upon a meta-linguistic level of the symbolisms that these signs point to in the world. Thus at the very moment of speech production, language must not only be about itself, but about something in the world and about the world that something is within.
Languages must therefore maintain a flexibility of pattern that is a key to their functional adaptability and their boundless productivity. All languages function best at an optimum level of structure. Too much overloads and ossifies a system, too little renders the system chaotic and incoherent. Internal changes tend to increase the coherence of a language system, but a system can become too constrained and thus compromise its external functional flexibility in the world. Thus there are built-in limits upon the amount of structure and degree of variation possible within human language. These limits are defined by the limits of human cognitive processing itself at various levels.
Language must therefore effect an optimal trade-off between internal coherence and external consistency in the encoding of the world. This is an allocative trade-off that is accomplished by balancing the structural constraints at the level of the sign with the expressive freedom at the meta-linguistic and symbolic level of meaning.
It accomplishes this dynamic equilibrium of an optimal level of functioning by means of a transformational calculus marking the permutation and modification of linguistic components at various levels, and a mechanism of mediation marking the transition from one statement to the next.
Language change describes an inherent natural selection process that to some extent resembles biological evolution. We can speak of changes in words in morphophonemic conditioning processes in much the same way as we can speak about genetic mutation. Language change is entirely dependent upon social selection processes. In this it is basically arbitrary and not entirely random.
At the same time, language patterning had to have conferred relative fitness and a counteadaptational influence upon human evolution leading to the progressive development and elaboration of human language organs, including the brain and speech apparatus and nerves and muscles connecting these things together.
Also, we cannot see language existing outside of some kind of well defined cultural context. All languages have culture and come attached to things cultural like little nametags. Thus, human groups have had very long and deep origins in working out their shared rules and implicit conventions regarding language agreement.
Language patterns are intrinsic to cultural patterns, thus, as with gene-culture evolution, we can also speak of gene-language evolution in much the same way. Many reconstructions of histories of language families replicate the ancestry of people very well, though this is not without important exceptions of borrowing and contact.
If we look at language change processes, we can see that they are continuous and still occur all the time, such that the English spoken today is fundamentally different in some ways compared to how it was used a century or two centuries ago.
Historical processes of language change suggest that there are two forms of linguistic variation--drift that comes from endogenous language change and borrowing that comes from exogenous language change. Linguistic drift appears to be continuous and fairly rapid, and also reflects other kinds of social changes that occurs. Basically, languages tend towards dialectical variation, to the extent that it reflects the functional integration of social reality. In state societies the processes of natural dialectical divergence of everyday speech are more or less counteracted by the superimposition of some standard that is usually enforced in schools and via the media.
Thus endogenous changes are influenced by changes in paralinguistic contexts that are not inherently linguistic, as well as by exogenous change events and processes, not just linguistically defined. To a great extent, endogenous changes are described by drift and the accretion of both meaning and new words within a language, as well as by the consistent morphophonemic conditioning of words within the language.
Another form of language change process occurs in linguistic history and this is the process of exogenous change that comes mostly through borrowing of words, a sign of a larger framework of acculturation. More dramatically but less commonly through processes of creolization, or the superimposition of a foreign language, or through complete or partial replacement of one language pattern by another, or some combination of these patterns.
In general, it can be said that exogenous patterns reflect not only contact between cultures, but, even more importantly, structural differentials between cultural groupings, such that the more powerful may impose their linguistic changes and wills upon the weaker. This tends to affect patterns of linguistic diffusion even when there is no direct power relationship between societies. Standards are imposed not only within societies to forestall dialectical differentiation, but standards become imposed between different societies to superimpose some degree of constraint and order between these societies.
A language is composed of many elements upon several interrelated levels of functional organization. Change may occur among any of its elements, along any of its dimensions, and upon any of its levels, but changes may be more likely to occur among some elements, areas or dimensions of a language than others, while some may be relatively resistant to change. Such differentials will vary widely with languages. It can be said that in general, core, basic, highly constrained patterns are less likely to change than more derivative, non-basic and less constrained patterns.
The basic mechanism of linguistic change is held to be the stylistic variation of human speech. The possibility for this variation arises in the intrinsic openness and productivity of language as a symbolic system of communication. It is the source of its adaptability and flexibility to produce an infinite variation of meaning, as well as of its intrinsic ambiguity and incorporation of contradiction. The mechanism of stylistic variation of language pattern drives both internal and external processes of change, or what can be called endogenous and exogenous change.
Though stylistic variation is the primary mechanism of language change, most stylistic variation fails to catch on and take hold in a speech community. The determining factors of whether a variation catches on are balanced by our two principles of communicative efficacy, the increase in the communicative efficiency or reduction of ambiguity and the increase in the expressiveness of language, or the augmentation of ambiguity. Furthermore, incorporation of new elements or variations of old elements may produce reverberations throughout a language that may result in other changes of other elements and aspects of a system, or create the potential for alteration in other places within the system.
Stylistic variation is motivated by numerous social and psychological factors, not the least of which are the desire for individual self-expression and the empowerment this brings, and the creation of social speech boundaries that mark off status differentials between people. The short term variation of language thus obfuscates the long-term stability and its actual rate of development a language may take.
Endogenous change has two contradictory consequences. First it drives a language towards greater internal coherence, but at the cost of external consistency and variability. Secondly, it accomplishes what may be called evolutionary linguistic divergence and fission, as two groups with a common parent language drift apart to become two entirely separate languages. At first glance, these two consequences of endogenous change appear to be contradictory, but on closer scrutiny we can see how they may in fact be complementary processes in which the drive towards increasing internal coherence will tend to create linguistic boundaries between speakers where not existed before. This will result in linguistic separation and isolation, and in processes of linguistic schismogenesis, or co-evolution.
We can see exogenous linguistic changes as having an opposite set of effects upon a language. Exogenous change will tend to increase the variability and reduce the internal coherence of a language. This function will be an inevitable response to the ambiguity of two different speakers attempting to understand one another. Though this increases the internal ambiguity of a language, the overall effect of this kind of change is one of linguistic fusion and convergence. Sometimes such convergence will dramatically alter the borrowing language, or an entirely new language may emerge from a process of linguistic amalgamation. One aspect of exogenous change is that unlike endogenous change that appears quite steady and constant in the long run, exogenous change appears quite erratic and discontinuous in the structure of the long run.
We can understand the history of language change then to have been one of an on-going dialectic between endogenous and exogenous factors of change. This dialectic revolves around stylistic variability in the application of rules of agreement, that is functionally motivated and which is an inevitable consequence of the trade-off between internal coherence and external correspondence.
We have come to a basic model of symbolic linguistics underlying process of cultural transmission and the linguistic construction of human reality. Language remains a centrally defining feature of human intelligence and human cultural patterning. We cannot have one without the other. In the penultimate chapter, I must return to some initial questions and close the story of human systems about the construction of symbolic worldviews.
Hugh M. Lewis
Blanket Copyright, Hugh M. Lewis, © 2005. Use of this text governed by fair use policy--permission to make copies of this text is granted for purposes of research and non-profit instruction only.
Last Updated: 03/17/05