Michael Halliday has discussed in some depth the relationship between what he refers to as the'"distinctively human semiotic"; the special form of dialogue powered by a system we call language' and the `two macrocosmic orders of which we ourselves are a part : the social order, and the natural order.' (Halliday 1987). Halliday argues that we have always used language in culture to construe, i.e. to understand or make sense of, these two macrocosmic orders, without necessarily being fully aware that we are doing so. He goes on to say that 'All this dialogic construction is, by definition, interactive. At the micro level, we get to know our fellow creatures by talking to them and listening to them; and they respond to us in the same natural language. At the macro level, our "dialogue with nature" (cf. Prigogine & Stengers 1984), is also interactive; but in another guise. When we want to exchange meanings with physical or biological nature we have to process information that is coded in very different ways, and that may need to go through two or three stages of translation before we can apprehend it.' (Halliday 1987) In other words, our understandings of natural processes and systems developed within the fields of science are mediated, not only by the languages we as a scientific community have evolved to describe and understand them, but also by the interpreting media and technologies that we use in order to quantify, and make particular the phenomena we are examining.
Halliday treats languages as dynamic open systems in very much the same kind of way as Sydney Lamb does in his stratificational model of grammar (Lamb 1966). Where Halliday differs most from Lamb is in his concern with incorporating into his description the development of the closely woven networks of interpersonal relationships that grow over time between users of language in culture. Halliday has in recent years begun to focus on the close relationship between the development of grammar and language, the development of intersubjectivity, and the types of interactions human beings and their cultures have with the broader horizon of the natural environment. Here, he draws heavily on work by amongst others Jay Lemke and Gerald Edelman. For Halliday, as with Peirce, there is no need to make a Cartesian mind-body distinction, and no need to make a human-nature distinction either. Halliday's approach is thus 'naturalistic' in one sense; it advocates an ecologically grounded, systemic understanding of language as just one aspect of an encompassing evolutionary process.
Halliday stresses continually in his writings that language and grammar are as much a product of evolution as we ourselves are; languages and their lexicogrammars are naturally evolving, and thus, at present, evolved systems, rather than designed systems. With reference to the theory of neuronal group selection developed by Edelman, he attempts to synthesize the development of language and grammar with the growing insights that are emerging from neurobiological research into the evolution and ontogenesis of the human brain, using Edelman's metaphor of a rapidly growing jungle (Edelman 1992). We must, says Halliday, concentrate on studying languages and grammars as they are; as naturally occurring phenomena; as one aspect of our situatedness as human beings within a larger natural order which has a unidirectional drive towards steadily greater complexity and diversity.
With regard to the relationship between language and science Halliday points out that we - through our 'investigative dialogues with nature' that we have over time have come to characterise as science - have learned to measure, experiment and try and understand the phenomena we observe around us. As a result of this evolutionary process we have developed metalanguages - languages of philosophy, mathematics and science. These metalanguages are, however, all in themselves extensions of natural languages, not totally new creations; and they remain intimately tied not only to one another, but also to the very phenomena that they are used for investigating and attempting to illuminate.
Although Halliday never refers specifically to Peirce or Husserl I find there are nonetheless an astounding number of similarities between his and their ways of thinking and investigating reality; his basic ontological approach is a phenomenological one. His fundamental concern with the evolution and development of language (and thus also with the languages of science) as they are constituted through sociocultural and interpersonal practices and as they are construed by us through our meaning-making practices within the horizon of a greater natural environment has resulted (in my humble opinion) in one of the most elegant descriptions of the development of a semiotic system in modern science.
It seems then to me that a phenomenological approach of this kind that is able take completely seriously the highly disparate and dynamic nature of "the phenomena as they present themselves for us", without the need for prior theoretical constructs or models to frame these experiences, and which at the same time attempts to situate these phenomena as semiotic systems grounded within an evolutionary horizon of a wider set of social and cultural practices; i.e. in terms of symbolic representations enacted and construed by a community of interpreters is the only viable way of investigating and describing the evolution and ontogenesis of text-norm systems in distributed virtual environments. This means too, that any technology which enables the creation of such environments by those participating in them also must be studied as situated within such a horizon of culture situated within an even wider ecological horizon of an encompassing nature.
Halliday has also discussed at some length how the languages of science develop (see Halliday 1987). Many scientific concepts are created through a temporally extended process of what he calls grammatical metaphor. Grammatical metaphors embody (or perhaps come to signify is a better expression to use here) in a concise form complex procedural descriptions that have been made systematically over time on the basis of individual researchers' observations and their intuitions about these observations. We need to let language alone, says Halliday, and he disputes the need to construct languages of science, as such as David Bohm's suggestion of the rheomode (Bohm 1980), precisely because language is itself a dynamic open system, which is constantly regenerating itself through the process of creating new meanings within the social semiotic of the scientific community. Ernst Cassirer has pointed out that the semiotic system of language continues own its trajectory within the evolutionary horizon of nature, culture and society, and the languages and symbols generated within science are constantly becoming interwoven into the greater set of language practices of culture as a whole, taking on new meanings as "animal symbolicum" devolves new forms of collective meaning out of the discoveries of the sciences (Cassirer 1945). Umberto Eco has also discussed this same theme in his discussions of the way in which the signs and symbols produced within the mass-media are taken up in and given new meanings by culture (see for instance Eco 1987).