This is Lecture Seven

The Social Basis of Linguistic (and Propositional) Knowledge

From the Syllabus:
the origin and function of language; language and action in groups;
the social use of mental models; the populational character of language

In lack of time, this is going to be just a sketch.



Language and Action in Social Groups
In this section am using and also twisting around some stuff from
Csanyi, V.: Az emberi termeszet (in Hungarian), Vince, Budapest.

It all begins with social rites, an element of human group behavior.
A social rite is a course of actions performed repetitively, in a rigid order and stable manner
        (i.e. it is, what else, a mechanism).
Rites have roots in aminals:
        the case of dog
        the case of primates
Animals can have rites, in particular the dog, which is bred for social skills in human society
    (attention, friendliness, cooperation, etc etc - including schematized, repetitive behavior).
Rites possess the same structure as language (and exemplify stable narratives)
    sequential
    "symbolic"     - this should be taken with care, a better word is "arbitrary"
    constancy of form
    etc.

(A speculation:) the social machinery of early human groups is the first mechanism.
Remember that mechanisms are "friendly" systems - if that is true, rely on some "affordances";
    i.e. options ensured by the environment.                        ( I thank this remark to Takashi Ikegami).

So, where do we find them, in the first place? Here is the speculation:
The primary affording environment for mechanisms is the world of artifacts and social rules.
The human as "rule-following animal" - reliable behavior enacted by social constraints and cultural rites.
Mechanism and narratives as general tools of modeling may depend on that.
            There is a little discrpenacy here. Elsewhere I say that animals are good at handnlig episodes and mechanisms.
                Which of these is true... can they be true simultaneously?
                When I say animal, I mean primates or dogs... they are both close to having social machines and langauge
                 So maybe no contradiciton after all.

Language is the secondary use of mental models based on group action.
Animals don't have language because their social structure (and their memory) is not so well developed.
            Perhaps these two "definicencies" of animals are related or the same.


Language as a Social Phenomenon

The origin of language:
"Some people see language as arising out of the manufacture and use of
tools.  Others see it as arising out of a need for gossip and social
bonding.  Some see it as arising out of song, dance and play.  Others see
it as arising from gestures.  Still others see it as arising out of sexual
selection."

The grooming theory    (social comfort function - in accordance with Bonobo etc. theories)
According to the bonobo theory, (proto) human groups were possible to maintain (e.g. to
        limit intra-species aggression to levels that allow for large mixed male-female populations)
        by means of "bribery" (compromise) - by comfort and/or sex.
R. Dunbar (1996) Grooming, Gossip and the Evolution of Language.
        London: Faber & Faber

The function of langauge
Language is not for the communication of truth (or propositions in general).
A possible function is to establish and maintain contact with other species (group) members.
Another possible function (perhaps derived from the contact function) is to coordinate group actions.
Etc.

We do not need to take sides between such views here, as the point of the lecture is just to illustrate
language as a purely social phenomenon (as opposed to the widespread view that it serves
naming or thinking or communication in the sense of "exchange of information").
Even today most discussions of the origin of language (e.g. Pinker, Bickerton, etc) assume this view.

The communicative view is completely incompatible with the biological view that language has its
    origin in small (proto-) human groups and needs to have a biological function which is esential
    for individual (or group) survival. ---> recent group selection theories

It is also incomptible with Wittgenstein's analysis that language cannot be private, i.e. that
    language exists as part of a broader social interaction, from which it is inherited to the individual.
This view, besides "pulling out" language from the head, immediately solves a number of notorious
    problems, such as the problem of certainty - our simplest utterances are simply repetitions of
    others'; language is not our own; judgment is that of the community; "we can't be wrong" (Davidson).


The Emergence of the Propositional Use
Social and situtational use is fundamental.
    *Tyical sentences are social sentences (speech acts, expression sor relations, commands, etc.)
How does a propositional use arise and why?
Propositions are labels of (parts) of situations.

Based on the associative theory, suggested by Wittgesntein already (and others like Halbwachs etc),
    language is part of a broader social activity to which it is associated.

Words and other linguistic units emerge as part of this activity and have no function in isolation.
But they become parts of the mental models of the episodes that constitute the social situations.
Hence it is possible to use them as markers of entire mental models, or of components of MMs.

This propositional use of language is not independent from the underlying mental models, but
    since propositions (utterances) themselves consitute a target domain to be modeled,
    there can be mental models for them as well - and propositional language begins "new life".

In other words, there are two different situations:
    thinking based on mental models of an original target domain, accompanied by propositions        (primary, meaing is given)
    thinking based on mental models of propositions                                                                        (secondary, meaning is problematic)



The Populational Character of Language

This is a different but related topic.
Language (if anything) is usually thought to be categorical.
Words are types, and their definite meaning is also a type. And so on.

We already deconstructed this view, when discussing the error from writing -  the propositional mind.
In actuality, word meaning is
        flexible, multiple, and blurred;
        there is variability;
        meaning is a family (Wittgenstein).
Family resemblance is a concept inherited from biology.
The biological notion of species is based on this, family resemblance - a disjunction of conjuctions, or
    partial, non-transitive similarity.
Therefore (as we discussed) the Wittgensteinian suggestion is that language (in this respect) behaves like a species.

But as the Heraclitus paradox ("You cannot step twice in the same river") has two sides ("you" and "river"),
quite similarly, word and language populations have two different sides.

One is the populational character of meaning.                                            ---> flexibility, unsharpness of reference
Another is the populational character of word (and grammar) use.             ---> flexibility, unsharpness of rules
We discuss this second next.

What does it exaclty mean to have rules (for word use etc)?
Wittgenstein speaks about langage games with strict (social) rules.
In fact the parallel:     mechanism - rite - rule - language    suggests that languages games are (social) mechanisms.
This means the exsitence of a rigid, repeatable, tightly organized structure.

Cearly, any deviation from a rule means the breaking down of the rule, or mechanism -
    the introduction of an "idiosynchratic" or foreign speaker who is not part of the "game".
This is in exact parallel to the games (board games, ball games, etc) people play - within a given
    social context, i.e. in a group of players, only one rule set is applicable.

Yes, but there exist different groups of people who can (and do) play differently; also there exist various languages.
So we see a crude example here for existence of variations in rules of word use and grammar.
How refined is this picture?
If going from the individual to the whole sets of native speakers: at what level does the variation occur?
In other words, how does the fact of difference in rules affect our entire picture of what is a language?

The answer is: fundamentally.
Consider this picture.





The figure as a crude picture of language 
The picture reveals a deep structuring of language at several levels.
Languages (natural human languages) are not distinct entitites - differences are structured, according to the structuring of
human groups. As a consequence, several forms of difference exist - regional, social, cultural; and this at various scales
(country - county - city - village - street [!]).

Microsociology reveals differences among families and within the family.

Ideally, within each group, language is homogeneous (it exists as a well-defined social "game" or mechanism) - relative
to the given level of resolution.

Example. "all Japanese speak Japanese", "Kyoto people speak Kyoto dialect" - yes, but Japanese
or Kyoto dialect cannot be defined without some violent simplifications pertaining to the given level of difference considered
(where Japanese is distinguished from, say, Hungarian, and Kyoto is distinguised from Kanto). Note, however, that no individual
speaks exactly "Japanese" or "Kyoto dialect" - as language is not a prototype (top-down) but an aggregation (bottom-up).

At the bottom, there is the actual language use of individuals, as members of an elementary social group;
within each group language MUST obey more or less exactly the same rules (in the broad sense, as a shared set of rules), or the "game"
(the language mechanism) is impossible to play. (For instance, the intrudcution to foreing speakers of people with other
dialects/idiolects leads to a fast rearrangement of the "game" to the extent that it becomes "playable" (but this is never perfect
as the goal is social life and joint actitiy and not some puristic ideal of linguistics.

Yet functional social groups are just temporary coalitions (this is not shown on the simple illustration above!) - and the same individual
is a member of several groups, simultaneously or in succession. As a result, the same individual speaks several variants depending on
group situation. Language as a whole opens up into several overlapping, mutually penetrating, difficult-to-analyze subsets of variants.

The figure as a crude picture of species
Originally, this picture was constructed as a didactic illustration to the concept of species. (Of course this is a crude and exaggerated picture.)
On phase one we see two well-defined species from a birds-eye view. From this perspective they apper to constitute two homogeneous types
or categories. Stepping closer, it becomes visible that  they possess internal heterogeneity. Still later we see sub-species and then well distinguishable
varietes (which may deserve an own name - Darwin studied them extensively). Finally, we come to see the smallest groups and the individuals.
Transition is continuous: intermediate forms exist everywhere. The concept of category and type becomes useless: all forms dissolve, boundaries
exist nowhere.

Knots appear where the closely similar variants are very frequent (and therefore, from a distance [i.e. from an intermediate resolution level] or,
what is the same, on the basis of limited information they coincide).

As the telescope "zooms in", this illustrates not just the necessity of a conceptual change, the switching from a categorical and essentialistic to a
non-categorical and anti-essentalistic thinking, but it also illustrates the successive stages of the knowledge process in science.

Species, understood as types, were initially imagined from a few, sometimes only one, specimen. [A scholastic approach.]
Naturalist experience, victorian journeys, and the observation of a vast plentitude of actually existing, different, varying individuals was
the key to dissolve the rigid, imaginary unity of species brought forth by such scholastic contemplation.

Likewise in language - textbooks and dictionaries reflect the language use of one speaker, assuming (wrongly) that all others are identical.
Spoken language shows an even greater variability.

The picture as a metaphor: "languages are species"
The suggestion is this:
species - sub-species - variant - individual  =  language - dialect - group language - idiolect

Skolasticism   =  Saussurian (Chomskyan) linguistics
naturalism       =  sociolinguistics



Language as an Evolutionary System

Darwinian evolution is best understood as a relationship between variation and variability.
Species vary because they have variants, generated by the variability of inheritance.
The two mechanisms could not exist in separation -
        in fact these are two sides (the generatove and the eliminative) of the same process.
Another subject, which deals with a same kind of link between variation and variability, is the study of human language.
There exist old parallels:


".. in fact, a breed, like a dialect of a language, can hardly be said to have a distinct origin. A man preserves and breeds from an individual with some slight deviation of structure, or takes more care than usual in matching his best animals, and thus improves them, and the improved animals slowly spread in the immediate neighbourhood. But they will as yet hardly have a distinct name, and from being only slightly valued, their history will have been disregarded. When further improved by the same slow and gradual process, they will spread more widely, and will be recognised as something distinct and valuable, and will then probably first receive a provincial name. [..] As soon as the points of value are once acknowledged, the principle, as I have called it, of unconscious selection will always tend--perhaps more at one period than at another, as the breed rises or falls in fashion--perhaps more in one district than in another, according to the state of civilisation of the inhabitants--slowly to add to the characteristic features of the breed, whatever they may be. But the chance will be infinitely small of any record having been preserved of such slow, varying, and insensible changes."

Darwin: The Origin of Species, 6th edition, chapter 1.
http://www.literature.org/authors/darwin-charles/the-origin-of-species-6th-edition/chapter-01.html


"…egy fajtáról, akárcsak egy nyelv dialektusáról, aligha mondható, hogy külön eredete volna. Valaki megõriz és továbbtenyészt egy olyan egyedet, amely valami kis felépítésbeli eltéréssel rendelkezik, vagy egyszerûen csak a szokásosnál gondosabban párosítja a legjobb állatait, és ezzel tökéletesebbé teszi õket. Aztán e javított példányok lassan elterjednek a közvetlen szomszédságban. De ezeknek eleinte nincs még külön neve sem, és mivel az értékük még csekély, a történetükre sem figyelnek. Amikor egy ugyanilyen lassú és fokozatos eljárás révén még tovább javulnak, akkor már szélesebb körben fognak elterjedni; elismerik õket különlegesnek és értékesnek, s ekkor valószínûleg kapnak valamilyen helyi nevet. […] Amint azonban az értékes tulajdonságokat már egyszer elismerték, a szándéktalan kiválasztás elve (ahogy én nevezem), mindig arra fog törekedni […], hogy lassanként gyarapítsa a fajta jellegzetes vonásait, bármik legyenek is ezek. De annak az esélye, hogy egy ilyen lassú, változó, érzékelhetetlen folyamatról feljegyzések maradjanak ránk, végtelenül csekély."

Darwin: A fajok eredete, 1.fej., Kampis György ford. (Typotex, Budapest, 2000).


At this point I finish discussion of the topic, with tyhe remark that the evolutionary component of language needs an exploration
along the variation - variability line, and not some obscure mutation - selection - adaptation line, favored in many studies.
Obviously, it is at this pont where things start to get most interesting, and some of my own current research goes in this
direction.




(c) gk 2002