Although modern society has been very keen on new language techniques and technologies, the split between the mundane tasks of language learning and the elevated speculations of language theoreticians is still rather wide.

At the same time, the history of linguistic thought shows that linguistic inquiry has always been trying to justify its existence by a task of facilitating language learning. One can find abundant evidence of this even in the long titles of mediaeval, Renaissance and Enlightenment treatises which, being actually summaries or outlines of the books, always included some reference to language learning. General or philosophic grammar was supposed to ‘serve as a foundation for language learning’: pour servir de fondement à l'étude de toutes les langues (Beauzée 1767). By the 20^th century the titles became shorter, and the acting linguistic subject (the speaker/hearer, the languaging individual, etc.) seemed to have disappeared from the linguistic writing. Maybe that was one of the reasons why the lay public as well as practicing teachers and textbook authors turned away from language theory.

Today, both the theory and the practice have changed again, and there is a tendency to revise the scientific and methodological foundations of language learning, borrowing ideas not only from contrastive and monolingual research of concrete language materials, but also from old and new philosophies and methodologies.

The linguistic science itself, the more it progresses forward, the more interest it seems to show in its own past. Any researcher compares what is going to be done with what has already been written by the predecessors. And, of course, there are figures whose scientific heritage is paradigmatic in the sense Thomas Kuhn attributed to the term paradigm (Kuhn 1970: 175-185), and in a broader sense.

This broad sense of the term paradigm can be seen if we consider what we really mean when we speak about outstanding historical figures in the history of linguistics or any other science. Do we attribute the name to a particular person who lived, got education, married, lectured, wrote, died in this or that year? In this and only this sense we can speak of the real Humboldt, the true Saussure, the genuine Bakhtin etc. Anyhow, truth is nothing other than a historical corrective to a persistent error, as Gaston Bachelard once stated (Bachelard [1934] 1983: 172).

In all other cases we have to admit – frankly and overtly – the fact that our interpretation of the writings of predecessors is nothing more than an interpretation. We create a kind of mythology, and the outstanding personality is a myth.

A myth is a system of beliefs to organize our behavior and attitude. It is not the real X who or what needs this organization (i.e. the object of our research), but we (the subjects of inquiry). We actually want to read out of the lines written by the classics the arguments and proofs for what we have already believed in.

There is a Russian joke (which may lose its intertextually induced charm in a foreign language environment): we are not readers, we are writers. This, anyhow, corresponds to Roland Barthe's idea that the author of the text dies in the process of reading, and reading is actually a process of writing, or, re-writing.

To admit this and to confess that we actually re-write the classics every time we read them in a new historical paradigm, is being frank and honest. To seek the ever-evading ghosts of things past (the ‘real’ classics), is a delusion.

A support for our views comes from Humboldt himself: Humanity and Nature cannot be grasped intellectually... one can only get somewhere near them actively. An active reading of Humboldt's works will help us get closer to his views.

And the views are there in abundance. Humboldt's ideas have been feeding researchers and thinkers of different scientific denominations for about two centuries. We would like to concentrate upon three issues of importance in Humboldt's writings.

A primary question, though not always explicitly asked by a practicing teacher, is the ontological status of language and language phenomena. However, there are certain presuppositions, or even biases that can be seen through their influence upon personal learning technologies and personal theories of language learning (Dufva 1994: 40-41) and language in general. These pre-conceptions can be called implicit beliefs or mythologemes (Many take it for granted that language exists, that there are certain units that a researcher has to investigate and a student has to learn one by one (words, or sentences, sounds, or syllables, etc.).

This leads us to a mythologeme peculiar to professional as well as to naive linguistic thought. Language is regarded as an existing thing, or it is dismembered into separate units (words etc.) which are regarded as things.

Chosisme (or object fetishism), as Gaston Bachelard called this attitude (Bachelard 1983: 39, 107), seems to be peculiar not only to linguistic philosophy, but also to natural sciences, such as physics. Anyhow, even there it bears the mark of the mechanistic world view of the previous century. After paradigms turned from Newton to Einstein, in microphysics it was also impossible to distinguish between object and motion (Bachelard 1983: 64-70).

The subject–object opposition in post-Cartesian thought presupposes some fundamental difference between the two. The object has to be different from the subject, its has to be passive and singular. Everyday practice of human existence in a physical world did not give many counter-examples.

Humanitarian research seemed to borrow the beliefs from those of physical sciences. The scientific ideal was to study separate objects, or structural combinations of such objects. The essence of humanitarian objects, at the same time, is quite different.

Maurice Merleau-Ponty once inquired: Where exactly is the painting I behold? If the answer is: in front of me, on the wall, in the museum or the like, then the painting is equal to the canvas and the colors on it. Anyhow, we do not get paintings only by joining colors and canvases (Merleau-Ponty 1964: 17). Neither do we get language by joining paper and printing toner. Even if we reduce it to its primary oral form, language is not sounds or air vibrations.

Physical existence, in its pure sense, can be attributed only to the humans, the ‘speaking bodies’, in fact. What we call ‘words’ and other linguistic units are not objects in the primary sense of the term, their existence is different from the existence of objects in physics or chemistry. What we call words are patterns of movement that speaking bodies produce, moreover, they are constantly repeatable patterns for large groups of speaking individuals, patterns of ‘consensual behavior’ (Maturana 1978: 50) that somehow are able to organize their behavior as a whole:

Linguistics is sometimes in the trap of a myth of ‘objective reality’. Once singled out and designated metaphorically as things (one of Lakoff's ‘metaphors we live by’), words are often conceived as such. And what a linguist does in his research is closer to disentangling dead bodies than to studying life within language, but:

Humboldt drives his reader away from this persistent stereotype of positivist objectivity:

If we regard language not as pure ergon, but as energeia, we can drift from the pathologoanatomic point of view to the live reality. Humberto Maturana, in the 20^th century, coined a very good word for this: languaging. He compared language activity, or languaging, to dance, calling it a dance or coordination of behaviors.

Another German thinker, a language philosopher of the so-called neo-Humboldtian trend, Fritz von Mauthner, at the turn of the two previous centuries also used to repeat in his writings that language is movement (Sprache ist Bewegung). He writes:

The idea of antinomic relations of thing and movement was not alien to other branches of science in that period as well, it was coordinated with the new scientific spirit of the time. Henri Delacroix later added to this list such specifically human activities as art and religion (Delacroix 1930). Humboldt wrote:

Fritz von Mautner compared language with the bed of the river (Fluß, Strombett). The movement of the flow is languaging and the bed itself is the residue of this flow. Recurrent schemes of activity act like attractors (a mathematical term), they shape the everyday languaging in continuous echoing and re-echoing. They emerge like meanders, the recurrent paths of flowing rainwater on the glass, and might die as soon as the water extinguishes. Anyhow, the residues of the prehistoric rains can be found in stones and in year-rings of the trees, just like the petrified languaging movements of deceased civilizations are observed in texts.

Humboldt is thus against ‘mechanical learning’, he advocates for the ‘development of linguistic power’ in the learner:

This finds reflection in another neo-Humboldtian's work. Ernst von Cassirer speaks of language and any other semiotic activity of humans (animal symbolicum) as of an intermediary. Language constitutes an intermediary world between man and nature, between humans themselves: ein neues mittleres Welt. This, again, ends up in a positivistic metaphor. There is no intermediary world as a thing, as an object. The intermediary world has no other ontology but its emergence in the actions of those between whom it is an intermediary, it is ‘incorporated into the act of its actual generation’ (Humboldt). Mauthner adds in a poetic manner: “Sprache existiert in der Luft zwischen den Menschen” (Mauthner 1982: 3-4).

Writing about the essence of language, Humboldt presents his views in antinomies, one of them being the antinomy of the objective and the subjective:

Language acts as an intermediary between the languaging subject and an object, but not only there can we observe its intermediary function. The subject needs another subject to understand his utterances:

But is it not quite natural that everything in this world exists in antinomic collisions? And what we owe Humboldt is the ability to grasp the unity of the individual and the universal in language activity:

In linguistics and information theory, there has been a controversy of late between two approaches to human language activity. The one, borrowed from Shannon's mathematical model of communication, can be called a transmission model: information is transmitted in words (electrical signals or the like) from the source to the receiver. It is in evident compliance with the reification bias in the study of language and ‘language units’, Bachelard's chosisme: things can be transmitted, received and stored. It is also in compliance with the ‘tape-recorder myth’ in naive language users (and learners of a foreign language): some of their preferences in studying the language betray their personal theory of learning a language through receiving a number of ‘correct samples’ of the target language, preferably in the form of an authentic recording. Their mind, in their opinion, is equal to another tape-recorder.

Another tendency in regarding the process of communication and learning turns from passive perception and reproduction to the ‘echo’ metaphor. Humboldt wrote:

Nothing is actually transmitted or received, no material is transferred in communication, it is produced by the receiver according to the echoing pattern. Individual and social actions unite in a transient equilibrium of communication, they create the human sphere which inevitably becomes objective environment for every new individual to enter this world:

Humboldt's Selbsttätigkeit is close in its essence to autopoiesis in Maturana's conception. Autopoietic systems create and re-create themselves in recurrent patterns of activity (Maturana 1995). Two hundred years ago, long before any ideas of cybernetics or general system theory were formulated, Humboldt considered languages to be ‘self-regulating and developing sound systems’ (Über den Einfluß des verschiedenen Charakters der Sprachen auf Literatur und Geistesbildung, 1821). The idea of echoing, or re-creation can also explain the phenomenon of understanding. And here again against the transmission model of communication:

There can be some parallel between the antinomy of ergon and energeia, and Saussure's langue/parole, although in each opposition the bases are different, and Roman Jakobson once warned against this simplification. Pierre Bourdieu underlines the development of a langue/parole antinomy into competence/performance with Chomsky, although the latter is also an opposition on a different ground (Bourdieu 1991: 4-5).

Natura sane nationes non creat sed individua

Spinoza.

Another field of persistent inquiry is the problem of similarities and differences between languages and within languages. If language as a whole is an emergent phenomenon, which does not exist in the material sense of the word, then how is understanding, or similar action or attitude, possible?

Each human being, then, uses his own system of ‘linguistic movements’. Communicating ideas and understanding occurs as far as these systems intersect. Humboldt's metaphoric explanation uses the idea of two cones. Understanding is the more appropriate, the deeper the tips of these cones plunge one into the other. Complete understanding is thus an illusionary ideal.

Individual linguistic acts form the domain of joint linguistic actions. It is this domain that we traditionally call language. Language is thus intersubjective:

Using modern terminology, which has become popular after Barthes' and Kristeva's interpretation of Bakhtin's work, the consensual domain of language activity can be regarded as an intertextual activity. What we use in speech has already been used by previous speakers in the history. We learn the meaning of words from our parents, they, in their turn learned the meaning from their parents, and so on. Even innovation is intertextual, since every new reading of a text is a new interpretation of its meanings.

At the same time, no one understands the word in the same manner as his predecessor, as his conversation partner. Following this idea in Humboldt's work, A.Potebnya suggested a graphical interpretation of understanding. If two lines intersect, they form two equal angles, but if we imagine two triangles (or cones) meeting in the point of intersection, or ‘point of understanding’, we have to admit that all the rest is ‘infinitely variable’ (Potebnya [1913] 1993: 94). Maturana's ideas of consensual domain sound like an echo of this idea of intersection.

Intersection, consensus, similarities in individual linguistic activity form a common language. Language is thus a domain of similar linguistic activities of a group of individuals:

The differences in individual linguistic activity lead to differences in group activities, and territorial and social borders help establish different domains of activities of different languaging groups

On the world-wide scale we thus appear to have different languaging domains, ‘language circles’, in Humboldtian terminology:

Thus, language can be described as a social, intersubjective, intertextual phenomenon:

But it is only one side of language being an ‘intermediary’. The fact is that language is an intermediary not only between the individuals in a community and the whole humankind, it is also an intermediary between man and the world, the universe. “Language is constituted in the triangle: man – reality – real time” (Yurčenko 1994: 26). A variant of these ideas expressed by Humboldt can also be found in M.M.Bakhtin’s assertion that culture does not have its own territory and is situated on the borderlines. Language has no territory of its own and is also situated on the borderlines, the frontiers of the world. The borders of language are the frontiers of his struggle for survival in the world which is mainly hostile. Survival is a common task for all humans, and so “We have to talk in order to survive” (E.Rosenstock-Huessy):

Although one of the founders of dialogism, M.M.Bakhtin, treated Humboldt as a representative of what he called ‘individualistic subjectivism’, the following words written by Humboldt, could serve as a motto of dialogism, since they proclaim the other as the touchstone of the individuality of the self:

As well as people complement each other in linguistic interaction, so do the nations. The universal human nature displays itself in linguistic variation. Thus, unity and complementarity are two sides of the human language. National languages are closer to individuals and emerge from their consensual interaction:

And the ‘universal human language’ (no matter how hypothetically the term might sound) is a limit towards which the interpretative activity of humankind is striving, and the final destination of this strive is nature:

Thus, language is antinomic also in the following respect: it is individual materially, i.e. it is effectuated only through the linguistic movements of individuals; and it is social referentially, i.e. individuals develop a system of common reference through correlated linguistic actions. One can say, following Humboldt:

Language learning technology is always in search of a better way to mastering a foreign language. Whatever new techniques appear, there is always criticism from the naive learner that they are still not adequate, they do not help this individual personally. It was about two centuries ago that Wilhelm von Humboldt wrote the following words:

Has much changed within these two hundred years, or language learners are still bound to wandering in an uncoordinated manner among the foreign words ‘thanks to the methods of teaching’?

One of the respondents in our own study of personal language development, told the following story (it was an average student, far from being very fluent in English):

This evidence shows two major biases of modern language learning. First, what is studied is not language, but language units (‘objective entities’). Second, these units are learned as ‘the outer world’, while one's own language constitutes a world of itself. A foreign language, thus, is considered to be a ‘subject’ of the curriculum to be studied, not as activity to be developed and practiced.

A learner of a foreign language treats his own language on a different basis. It is something ‘one lives with or in’ – a definition given by a naive user in a reply to a questionnaire, which runs in total compliance with Humboldt's ideas about language and the world of the speaking subject.

A foreign language, on the other hand, is treated like an almost physical material consisting of separate unites to be studied. Knowledge about (objective knowledge) is substituted for knowing how (procedural knowledge). But how many people who seem to know that in the 3^rd person singular there should be an ending -s in the English verb, actually use this ending whenever it is time to act so?

This is what Humboldt describes in his terms in the following passage:

Dialogue is equal to learning. Understanding each other is also achieved in learning to act or feel like the other. Learning languages is learning how to act linguistically, how to produce linguistic body movements. And linguistic body movements are movements that organize other movements of ourselves and the others.

A very short article written by Humboldt in 1801-1802 and dedicated to language learning contained an outline for an elaborate program of linguistic studies. The program is so broad that it is purposefully called an encyclopedia:

Actually, there had been encyclopedic attempts in the history of linguistic thought before Humboldt, but what is most specific and attractive about his plan, is that it again was trying to embrace two branches in an antinomic unity. The encyclopedia was going to be more than just a scientific treatise to be read by language scientists. Humboldt's intention was a great deal more far reaching:

Humboldt criticizes methods of teaching contemporary to him for their lack of systematicity and coordination. But who could say that these tasks have been solved by now, and that the humankind has found today the easiest and effective way to study languages? Humboldt writes:

But criticism gives no positive fruit. Humboldt goes further than doubting the efficiency of learning techniques. He expresses the idea which inevitably follows the main principles of his humanistic language philosophy: learning is not reproduction but re-creation, the product of man's free will.

Learning is also communicating, and the process of learning a language is a dialogical interaction, where the hearer is ready to understand the speaker due to his inherent capacity:

The learner is the best teacher for himself and there are moments when he feels it:

Like understanding, which is the feeling of communicational satisfaction, informational comfort, the moment of acquiring a new world outlook through language, the moment when one feels he has crossed the line and now has the sense of the language he has been studying is a feeling of intellectual revelation and achievement. Languages are not taught, they are learned, or acquired step by step until the learner reaches the turning point, where the foreign language ceases to be an object to be studied, and becomes part of his identity:

Acquiring a language (whether a foreign one, or the mother tongue) is not a mechanical process of reproduction:

Humboldt's language acquisition theory is in parallel with his ideas on education and human social behavior in general. The idea of linguistic freedom lies in the basis of his antinomy of freedom and rules in language:

Some major ideas that can be derived from Humboldt and further developed in a new post-Cartesian or in a holistic paradigm seem to be as follows: intersubjectivity of language; processuality of language activity; language rhythm; language as environment and intermediary; language as a world view; unity and diversity in language and in languages; humanistic nature of linguistics; humanistic nature of language and what not.

So, as we have tried to show, Humboldt's ideas are a paradigm on their own. Following his thought more adequately, linguistics might have made a different turn in the late two centuries. We have seen that Humboldt's major issues show a different range of basic concepts.

First, as for the ontology of language. Language does not exist as a thing, it occurs, happens, proceeds etc. in the linguistic actions of individuals. Language existence is dynamic. The final reality in language activity belongs to the acting linguistic subject, not to the structure of language system. Language as a system is only a system of patterns emerging from linguistic behavior of individuals. Behavior occurs, though, only in reaction to some environment, i.e., it is in its historical essence dialogical by its character. An individual needs another individual to behave. So, language ontology is dynamic existence.

Second, as there are differences between linguistic behaviors of different individuals, there are also differences in the behavior of groups of such individuals. These group similarities form a wide range of coincidental behavior: from mutual understanding in a dialogue, to coexistence of dialects within ‘one language’ and translation between national languages. Thus individual merges with social and vice versa. An individual needs a group to create a language. So, there are no language units but language actions. And there are no clearcut boundaries or frontiers in language activity.

And last, but not least, learning a language is not passive or mechanical transmission or reproduction. One has to learn to act linguistically in one's own or another nation's pattern. An individual needs to act to form new patterns of behavior. So, there are no language users but language actors and creators:

There were followers who implemented Humboldt's ideas in theories of their own (Potebnya in Russia, Cassirer in Germany, to some extent even Chomsky, etc.). The mainstream research, anyhow, was mainly directed in search of a positivist myth of ‘objective reality’ in language. History, time, activity and the human spirit itself were almost eradicated. That is why linguistics dealt mainly with ‘linguistic autopsy’, and not with real languaging. But the ‘objective reality’ of linguistics and any humanitarian branch of knowledge cannot but be subjective by its nature. Linguistic research studies languaging subjects, it is a dialogue of self-conscious individuals, each of them being in turn an object for the other.