[Fis] Information is a linguistic description of structures

[Terrence W. Deacon]

The language metaphor is so very seductive. I fear that our discussion is
too easily ensnared by its ubiquity in our lives. From Günther's response I
am also not clear whether he is defending using the language model as
generic or as a special case (though we do not agree in our assessments of
von Frisch or Tomasello).

To me, using language as the paradigm general case is like a biologist
studying mammalian hair from the perspective of porcupine quills and
defining all other forms of hair as simplified quills (quills minus some
properties).

The problem is, of course, the way we define the distinctive properties of
language (simple property lists vs more basic semiotic analysis). Language
is the most developed form because its symbolic capacity depends on and
grows out of complex lower-order iconic and indexical forms of reference.
This is why one can also represent WHAT simpler forms represent in a
language-like system (but not vice versa). But the difference matters.
Saying that one is sad and sobbing may communicate the same content, but
the difference is significant. These more basic modes of representing are
presupposed in the concept of language (but typically bracketed from the
analysis). As a result we can erroneously ignore the difference in HOW
these other forms represent, as well as how language representation itself
is constituted.

As is exemplified by coughs, smiles, music, and skunk odors (and thousands
of other forms), I consider mapping all communication onto language (even
in the form of e.g. language-minus-properties-X-and-Y) to be a truly
procrustean enterprise— as is defining these non-linguistic modes of
communication in terms of syntax, semantics, and pragmatics. And so I am
advocating that we not too hastily assume that an information theory made
axiomatically isomorphic with natural language is the best most complete
model for all uses. Such a model may be sufficient for analyzing many human
linguistically-based communication systems, but that usefulness can blind
us to what formal language models tacitly assume and leave unexplained
about information in general.

— Terry


On Tue, Sep 29, 2015 at 9:59 AM, Günther Witzany <witzany at sbg.at> wrote:

> Yes, I agree with Terrance, that language of humans is a highly
> specialized form used in communicative actions of various forms, although
> in the beginning the difference between our ancestors – great apes – and
> early hominoids was not so dramatically different (see M. Tomasello. The
> origin of human communication, 2008). As we learned with the success of the
> pragmatic turn thinkers which refuted the linguistic turn thinkers in their
> attempt to construct an exact scientific language ( mathematical theory of
> language) and to delimit exact science from metaphysics is that a natural
> language we can term any sign system that functions according semantic,
> pragmatic and syntactic rules.
>
> This means also body movements can express utterances, can be combined to
> serial content which has meanings according its pragmatic context.
> Interestingly decades before this, there was a controversy on the language
> of the bees between Karl von Frisch with several others in which they
> refuted v.Frischs termination of bee "language" just as being a metaphor
> not really a language. In an article v.Frisch proved the fact that it is a
> real language with all key features of real natural languages, including
> the fact of several dialects. 20 years later he was awarded with  the
> nobelprize for his research results. The dialects he detected by mixing
> Austrian and Italian bees with the result, that the same sign sequences
> (dance movements) expressed different meanings according the original real
> lifeworld of the bees where they socialized.
>
> If we follow Juergen Habermas, the most cited philosopher of the present
> we can term “communication” every sign-mediated rule-governed
> interaction, although he did not mean this outside humans. As I tried to
> demonstrate in my "biocommunication and natural genome editing" approach we
> can identify such sign-mediated interactions throughout all
> organismic kingdoms in coordination and organization processes between
> cells, tissues, organs and organisms. Interestingly on the very basics of
> life, viruses and sub-viral groups of RNAs the difference between signs and
> sign-users can change rather fast, which means the passive function as
> template for genetic replication or being active as catalyst using a
> colonized (former) catalyst as template may switch. But this is a dynamic
> process that drives evolution into a constantly continuing process since
> nearly 4 billion years. We can imagine how successful artificial genetic
> engineering will be to manage genetic parasites…
>
> Am 28.09.2015 um 07:13 schrieb Terrence W. Deacon:
>
> As exemplified in Guenther's auxin example, and Pedro's worries about the
> procrustean use of language metaphors in the discussion of inter- and
> intra-cellular communication, it is likely to be problematic to use
> language as the paradigm model for all communication, much less as the
> foundation upon which to build a general theory of information. From an
> evolutionary point of view, language is a highly derived human
> idiosyncratic form of communication that evolved only very recently in
> vertebrate phylogeny, in only one species, and is supported by a vast
> semiotic cognitive and social infrastructure. Communication in a more
> general sense is vastly older and far more generic. For this reason, it is
> wise to avoid talking in terms of the semantics of a cough, the meaning of
> a piece of music, or the syntax of a skunk's odor. The use of Carnap's
> approach to language semantics and various other uses of linguistic
> categories in information theoretic analyses needs to be understood as
> a special case, not the generic form. I would recommend that presentations
> and comments to them be framed with appropriate caveats, indicating whether
> they address such special cases of human information or are intended to be
> generic.
>
> On Fri, Sep 25, 2015 at 4:37 AM, Pedro C. Marijuan <
> pcmarijuan.iacs at aragon.es> wrote:
>
>> Dear FISers and all,
>>
>> I include below another response to Immanuel post (from Guenther). I
>> think he has penned an excellent response--my only addition is to
>> expostulate a doubt. Should our analysis of the human (or cellular!)
>> communication with the environment be related to linguistic practices? In
>> short, my argument is that biological self-production becomes "la raison
>> d'etre" of communication, both concerning its evolutionary origins and the
>> continuous opening towards the environment along the different stages of
>> the individual's life cycle. It is cogent that the same messenger plays
>> quite different roles in different specialized cells --we have to
>> disentangle in each case how the impinging "info" affects the ongoing life
>> cycle (the impact upon the transcriptome, proteome, metabolome, etc.) There
>> is no shortcut to the endless work necessary--wet lab & in silico. So I
>> think that Encode and other big projects are quite useful in the continuous
>> exploration of biological complexity and provide us valuable conceptual
>> stuff--but looking for hypothetical big formalisms (I quite agree) is out
>> sight. Molecular recognition which is the at the  fundamentals of
>> biological organization can only provide modest guidelines about the main
>> informational architectures of life... beyond that, there is too much
>> complexity, endless complexity to contemplate, particularly when we try to
>> study multicellular organization. Anyhow, this topic of the essential
>> informational openness of the individual's life cycle appears to me as the
>> Gordian knot to be cut for the advancement of our field: otherwise we will
>> never connect meaningfully with the endless info flows that interconnect
>> our societies, generated from the life cycles of individuals and addressed
>> to the life cycles of other individuals. Info sources, channels for info
>> flows, and info receptors are not mere Shannonian overtones, they
>> symbolically refer to the very info skeleton of our societies; or looking
>> dynamically it is the engine of social history and of social complexity.
>>
>> Well, sorry that I could not express myself better.
>>
>> all the best--Pedro
>>
>> Günther Witzany wrote:
>>
>> Dear all!
>>
>> What is the opposite of a linguistic description? a non-linguistic
>> description? Please tell me one possible explanation of a non-linguistic
>> description. So Im not convinced of the sense of the term "information".
>>
>> Concerning the "difference" of physical and semantic information: What
>> would you prefer in the case of plant communication. Does the chemical
>> Auxin represent a physical or a semantic information? Auxin is used in
>> hormonal, morphogenic, and transmitter pathways. As an extracellular signal
>> at the plant synapse, auxin serves to react to light and gravity. It
>> also serves as an extracellular messenger substance to send electrical
>> signals and functions as a synchronization signal for cell division. At the
>> intercellular, whole plant level, it supports cell division in the cambium,
>> and at the tissue level, it promotes the maturation of vascular tissue
>> during embryonic development, organ growth as well as tropic responses and
>> apical dominance. In intracellular signaling, auxin serves in
>> organogenesis, cell development, and differentiation. Especially in the
>> organogenesis of roots, for example, auxin enables cells to determine
>> their position and their identity. These multiple functions of auxin
>> demonstrate that identifying the momentary usage (its semantics) is
>> extremely difficult because the context (investigation object of
>> pragmatics) of use can be very complex and highly diverse, although the
>> chemical property remains the same.
>> Yes, mathematics is an artificial language. Last century the Pythagorean
>> approach, mathematics represents material reality, (if we use mathematics
>> we reconstruct creators thoughts) was reactivated: Exact science must
>> represent observations as well as theories in mathematical equations. Then
>> it would be sure to represent reality, because brain synapse logics then
>> could express its own material reality. But this was proven as error. Prior
>> to all artificial languages we learned how to interconnect linguistic
>> utterances with practical behavior in socialisation; therefore the ultimate
>> meta-language is everyday language with its visible superficial grammar and
>> its invisible deep grammar that transports the intended meaning. How should
>> computers extract deep grammar structures out of measurable superficial
>> syntax structures? In the case of ENCODE project (to find the human genome
>> primary data structures) this was the aim which got financial support of 3
>> billion dollars with the result of detecting the superficial grammar only,
>> nothing else.
>>
>> Best Wishes
>> Guenther
>> Am 24.09.2015 um 07:47 schrieb Emanuel Diamant:
>>
>> Dear FIS colleagues,
>>
>>
>>
>> As a newcomer to FIS, I feel myself very uncomfortable when I have to
>> interrupt the ongoing discourse with something that looks for me quite
>> natural but is lacking in our current public dialog. What I have in mind is
>> that in every discussion or argument exchange, first of all, the grounding
>> axioms and mutually agreed assumptions should be established and declared
>> as the basis for further debating and reasoning. Maybe in our case, these
>> things are implied by default, but I am not a part of the dominant
>> coalition. For this reason, I would dare to formulate some grounding axioms
>> that may be useful for those who are not FIS insiders:
>>
>>
>>
>> 1. *Information is a linguistic description of structures observable in
>> a given data set*
>>
>> 2. Two types of data structures could be distinguished in a data set:
>> primary and secondary data structures.
>>
>> 3. Primary data structures are data clusters or clumps arranged or
>> occurring due to the similarity in physical properties of adjacent data
>> elements. For this reason, the primary data structures could be called
>> physical data structures.
>>
>> 4. Secondary data structures are specific arrangements of primary data
>> structures. The grouping of primary data structures into secondary data
>> structures is a prerogative of an external observer and it is guided by his
>> subjective reasons, rules and habits. The secondary data structures exist
>> only in the observer’s head, in his mind. Therefore, they could be called
>> meaningful or semantic data structures.
>>
>> 5. As it was said earlier, *Description of structures observable in a
>> data set should be called “Information”. *In this regard, two types of
>> information must be distinguished – *Physical Information and Semantic
>> Information*.
>>
>> 6. Both are language-based descriptions; however, physical information
>> can be described with a variety of languages (recall that mathematics is
>> also a language), while semantic information can be described only by means
>> of natural human language.
>>
>>
>>
>> This is a concise set of axioms that should preface all our further
>> discussions. You can accept them. You can discard them and replace them
>> with better ones. But you can not proceed without basing your discussion on
>> a suitable and appropriate set of axioms.
>>
>>
>>
>> That is what I have to say at this moment.
>>
>> My best regards to all of you,
>>
>> Emanuel.
>>
>>
>>
>>
>>
>>
>>
>> --
>> -------------------------------------------------
>> Pedro C. Marijuán
>> Grupo de Bioinformación / Bioinformation Group
>> Instituto Aragonés de Ciencias de la Salud
>> Centro de Investigación Biomédica de Aragón (CIBA)
>> Avda. San Juan Bosco, 13, planta X
>> 50009 Zaragoza, Spain
>> Tfno. +34 976 71 3526 (& 6818)pcmarijuan.iacs at aragon.eshttp://sites.google.com/site/pedrocmarijuan/
>> -------------------------------------------------
>>
>>
>> _______________________________________________
>> Fis mailing list
>> Fis at listas.unizar.es
>> http://listas.unizar.es/cgi-bin/mailman/listinfo/fis
>>
>>
>
>
> --
> Professor Terrence W. Deacon
> University of California, Berkeley
>
>
>
>


-- 
Professor Terrence W. Deacon
University of California, Berkeley
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listas.unizar.es/pipermail/fis/attachments/20150929/4f5aef9b/attachment.html>