• Keine Ergebnisse gefunden

3 Preliminary Clarifications from Visualistics

3.4 Image and Language

con-ceptual place of »resemblanceα« (now in opposition to »identity«), thus opening the possibility of perceptoid signs. Though at its core is still the older primary re-semblance only modified by the new possibility of individual ob-jects.

For the complete object consti-tution necessary to understand how resemblanceβ is conceived, another transition to an even more complicated field of concepts of behavior must be considered: we have to look at creatures that can establish relations between pre-objects in arbitrary contexts, and thus originally invent identity in the proper sense (Fig. 20). The anticipations of pre-object creatures essen-tially integrate contexts directly linked by means of distance stimuli. That is, those con-texts are organized along the course of coherent activity – they are, in an extended sense, all present in the course of the ongoing action.Hence such anticipations are not sufficient to establish identity between completely disparate situations. A re-present-ation of any contexts apart from those mediated by continuous action can only be medi-ated by means of signs (cf. also [PLESSNER 1928, Sect. 6.4]).

common “discourse universe” .16 If the actual situation of communication and the mate-rial objects “within” are meant, deictic expressions together with pointing gestures an-swer the purpose of nomination particularly well. But the function of nomination is not limited to those objects: we may of course utter expressions like ‘the last unicorn’, ‘the Platonic Form of beauty’ or ‘Otto von Guericke’, as part of an assertion, too, in order to point out an object of a corresponding discourse universe. They work as long as the dis-course partners are able to distinguish the object meant from the other disdis-course objects in question. In contrast to the nomination, predication has no immediate representational aspect: it is the sign act used by the sender to inform or propose to the others that a cer-tain custom of distinction – a concept – is relevant and applicable to the objects named.

In the example above, ‘being blurred’ carries the predication. In every assertion, one predication (that may indicate a complex combination of concepts) and one or more nominations concur systematically.

Assertions are context-relative: if the corresponding discourse universe is unknown, an assertive sentence remains essentially incomprehensible. The nomination can only be performed effectively if it is clear which set of objects is at stake at all.17 Objects are never given in isolation. We always speak of objects as something appearing as a figure in front of a background: they are part of a context. The expression ‘context’ is used in the following for indicating a finite structured set of intentional individualized entities, i.e., objects (as something known by somebody) with relations between them. More precisely, the relation between propositions and contexts is one of figure-ground to me-dium. A proposition offers a unique figure-ground distinction with the predication as figure on the ground of the objects known already and identified by the nominations. A medium offers the potential of figure-ground dichotomies, i.e., for many possible dis-tinctions. Objects, while forming the background for predications, are thus seen as fig-ure against other objects, as well.

Based on this introduction, many different types of contexts may be considered: for example, discourse universes are contexts shared by several creatures that communicate with each other. The situational context corresponds to what a single creature perceives as (individual) objects from its present environment. Other contexts are analogous to the situational one, but entail the “objective” environment of other times and places or even of fictitious and hypothetical situations.

Nomination can primarily be anchored in the overlapping parts of the interlocutors’

situational contexts. The physical environment of the sign act (and all simultaneous be-haviors of the interlocutors) provides then, it seems, the discourse universe of the ob-jects being commonly perceived.18 In contrast to that, the objects in the contexts evoked by previous assertions – or co-texts as they are called – do not have to be physically ac-cessible. An earlier characterization (the concept used in predication) can be applied as part of a consecutive nomination in the form of a definite description: ‘The blurred photo was taken by Hermione.’

16 Proper names (‘Harry Potter’), deictic particles (‘this’, ‘she’, ‘you know who’), definite descriptions (‘the picture of the fat lady’), and deictic descriptions (‘this blurred photo’) are the forms of nominations traditionally considered.

17 Though the power of words for spontaneous context-evocation should never be underestimated, see be-low.

18 Note that the different individual perspectives (as of pre-objects) must have been integrated conceptu-ally in order to allow us of speaking from anything “being commonly perceived”.

Objects (as we usually understand the expression) are members of many contexts.

What we call the identity of objects is basically the question of connecting an object in different contexts. Take for example a court of law trying to establish the identity of, for example, the dagger now presented (1st context), the pointed object that was used to stab the victim one year ago on the other side of the city (2nd context), and the knife bought by the accused in the neighbor city 13 months ago (3rd context). Note that it is impossi-ble to actually perceive simultaneously all the contexts in order to directly establish truth about identity. An important distinction of contexts – in FREGE’s terminology(cf.

below): of the “ways of being given” of objects – is the one between referential con-texts and intra-lexical contexts. Objects are said to be referentially given if they are elements of the current situational context. In this case, the legitimacy of an assertion can be tested directly at the object by, coarsely speaking, including it in corresponding sensory-motor behaviors. The sensory-motor anchoring in the referential context is ob-viously the foundation of any empirical research. If, however, an object has only been introduced verbally in the discourse universe, there remains nothing but to apply con-ceptual rules and draw conclusions from the predication about the objects that are not explicitly mentioned, and to check whether the assertion is logically compatible with the context [SCHIRRA 1995].

Assertions allow us apparently to make any context whatever a discourse universe, to share or harmonize it with the others, that is. Harmonizing perceptions between interlocutors by means of the sign acts has the obvious purpose of combining the diverse perspectives of an environment. Creatures thus endowed can perceive not only with their own senses, but also with the other’s senses; they can manipulate not only with their own hands but with the other’s hands, as well. Still, this would be a very restricted employment for assertions compared to what we usually do with them:

humans mostly talk about objects that none of the interlocutors can actually perceive in the situation – or that may even be not perceivable at all.

That is, assertions allow us to relate an arbitrary context with the current situational one. The use of proper names given in a christening situation long ago depends on that ability. As was noted above, speaking of a deception – viewed as an explicit lack of identity – also means to relate two different contexts of behavior with each other; so does considering resemblanceβ (in particular with something being absent). Thus, being able to use resemblance as a crucial component of a certain type of signs (iconic/per-ceptoid signs) depends on a faculty that appears to be essentially mediated by asser-tions, disclosing a strong conceptual dependency between assertions and perceptoid signs.

In summary, assertions are context-relative on the one hand; but on the other hand, they are context-independent, since we can, at least in principle, perform an assertion relative to some context in any situational context whatever. The two characterizations of assertions depend on each other because it is only possible to speak independently of the actual situational context if another context can be explicitly referred to.

3.4.2 Communication Among Pre-Object Creatures

In a Gedankenexperiment, TUGENDHAT [1982, Sec. 12] mentions a simpler class of communication games that are not independent of the situation of utterance: the “quasi-predicates”. A spontaneously sounded warning cry – ‘FIRE!!’ – may serve as an appro-ximation of that type; but also an infant’s utterance (‘bow-wow!’) in the one-word phase (ca. 20th month, [LOCK 1993]) is closely connected with quasi-predicates. Their

Figure 21: Ascribing a Quasi-Predication

rules of application must be strictly related to the correspon-ding context of utterance: a particular quasi-predicate is utter-ed only if certain conditions are being perceived (Fig. 21). The partner in communication also re-acts on such an utterance (as a quasi-predicate) only directly, e.g., by taking flight or by calming the excited child. That is: judging the correct use of a quasi-predicate (or explaining the correct use19) is to be conceived of as being bound strictly to the situation of use.

There is, so to speak, a fixed association between the (correct) use of a quasi-predicate and certain sensory-motoric routines (for testing or as a reaction).

Therefore, quasi-predicates articulate habits of distinguishing similar to a predication.

The difference becomes prominent if we compare the situations of utterance of the warning cry ‘Fire!’ and of an assertion like ‘The house is burning’: the rules of use of the predication can be discussed in absence of a concrete example; situation of usage and situation of explanation can be separated. Correspondingly, the spontaneous reac-tion can be dispensed with in the case of a predicareac-tion. That is impossible per defini-tionem when using the expression ‘fire’ (or also the sentence ‘the house is burning’) as a quasi-predicate. Quasi-predicates belong to the class of signals, a basic form of commu-nication widespread among animals, and also part of the biological endowing of the an-thropines (e.g., primary affective utterances; cf. [EIBL-EIBLESFELD 1984, Sect. 4.3]).

Of course the understanding of a child’s utterance ‘bow-wow!’ as the child telling us about a perceived individualized object suggests itself. But it is the field of concept of pre-object creatures that already allows us to construct a concept of communication based on quasi-predicates.20 Let us assume for the moment that the repertoire of behav-ior of the child is (still) quite simple, so that we cannot yet speak of him/her perceiving objects in our usual pretentious sense; but we may speak without problems about detec-tors and the perception of pre-objects. We can “explain” to that child that he/she used

‘bow-wow!’ in the wrong manner or we may “confirm” the regular use; but only as long as that pre-object is perceived – i.e., is part of the situational pre-context.21 Furthermore, we can provoke “accepting” reactions if we use ‘bow-wow!’ in the appropriate way: if the child perceives the corresponding pre-object; or the child begins to search for it. If there appears no corresponding perception the child reacts quite disconcerted. It would be rather odd if we react in a similar way on the nomination ‘our neighbor’s dog’ in case it is not simultaneously part of our situational context. Thus, assuming the child is

19 Note that such an explanation (in a wide sense) becomes necessary if the habits of distinguishing are no longer fixed genetically but by means of training, habituation, learning, etc.

20 Evidently, the simpler creatures particular to the field of concepts of »reflexes« cannot be ascribed of having any sort of communication in a proper sense: there are only reactions on stimuli that we under-stand as being causally linked to other creatures.

21 The ‘pre’ is necessary here since contexts have only been introduced on “full grown” objects.

a pre-object creature leads into interpreting its characterizing utterances as quasi-predicates.

Quasi-predicates are saturated: their communicative function cannot be set equal to neither the function of nominations nor that of predications. At best, the whole assertion may serve as an equivalent. But quasi-predicates depend on the situational context while assertions have become – despite their context-relativity – independent of the context of their utterance.

3.4.3 Context Builders and Referential Anchoring

In comparison to the elementary quasi-predicate, creatures gain with the complex sign act ‘assertion’ and its clearly separated partial acts ‘nomination’ and ‘predication’ the following essential advantage: it allows them originally to communicate about objects that are not present – not perceivable for them – in their actual environment. Using assertions is the only way at all to make “reachable” other contexts of action.

Indeed, we continuously use more or less consciously many verbal indications of con-texts / discourse universes. Applying tempus is a typical example, since assertions about past or future affairs do precisely not refer to the present situational context as their proper discourse universe; the latter must be derived from the former. The grammatical modifications of the verb are quite an implicit indicator. Explicit specifications of loca-tion and time may also serve to reconstruct the context used as the discourse universe to be considered further on.

In order to adequately fathom this crucial aspect of assertions we assume another necessary partial sign act beside nomination and predication. GILLES FAUCONNIER, who is particularly interested in the linguistic potentials and consequences of such a proposal from a cognitive science perspective, uses the expression ‘mental spaces’ for contexts;

correspondingly he speaks of ‘space builders’ – the verbal constructs that open up ex-plicitly or imex-plicitly contexts as the relevant discourse universes [FAUCONNIER 1985, 17]. In the following, the expression ‘context builder’ is used analogously for character-izing the partial sign act that in the frame of an assertion allows the interlocutors to re-construct the underlying context.

A special form of context building is the sequence of previous assertions, the co-text:

the (intentional) objects introduced or modified there may easily be referred to again by means of definite descriptions that employ the distinction mentioned before. Thus, each continuous propositional text can also be conceived of as the complex context builder for subsequent assertions: “In Tolstoy’s ›War and Peace‹, Platon is shot dead by a French soldier” (context builder in italics)

The distinction between the referential and the intra-lexical “way of being given” of objects has already been mentioned: only in the first cases, assertions about objects can be empirically checked – or as we shall say: can the assertion be referentially anchored.

Context builders pointing out locations give us at hand a method of how to transform the context meant by an utterance into the situative context in which the referential an-choring could actually be performed. Spatio-temporal coordinates play an important role [TUGENDHAT 1982, Sect. 26II]. Referentially anchoring an assertion then involves two steps: first, one has to know / recognize how the sensory-motor test routines linked to the nominations descriptive part (i.e., the one using distinctions formerly mentioned) are “prepared” – positioned, orientated, etc. (by transforming the context pointed out by the context builder into the situational context); second, one has to know how to actu-ally perform the sensory-motor test routines for the newly communicated habit of

dis-tinction associated with the predication (e.g., that I have to look in order to recognize whether something – take the photo mentioned in the example above – is really

“blurred”). Therewith, the “local” habits of distinguishing already associated with the elementary, i.e., strictly context-bound sign acts (quasi-predicates) may be employed.

However, they are modified by extra conditions that are necessary for the individuation of objects, i.e., the integration of their absent aspects. Such conditions are essentially stabilized socially.

Assertions may be conceived of as derived from quasi-predicates (gray arrows in Fig.

22) since they fulfill a very similar overall function – to harmonize situations of behav-ior. The additional differentiation into the three clearly distinguishable partial acts “con-text building”, “nomination”, and “predication” is the precondition for redeeming communication from the strict binding to the actual situation. So far, all contexts but the actual situation of communication can apparently be constituted only by means of being verbally evoked. That is, we are in the interesting situation of considering, on the one hand, creatures that are able to communicate in an elementary manner but are in a way completely restricted to the “here and now”.22 On the other hand, we think of creatures with a more complicated behavior; they master a kind of communication that is inde-pendent of the actual situation. However, this art of a relative independence from situa-tion depends circularly on their ability to communicate in such a complicated manner.

The tool for overcoming that horizon is given only in communication. We hardly know yet how to understand this sharp transition (cf. [ROS 2005]).

The problem we have reached here is indeed the question of the origin of (the field of concepts of) geometrical space (and measured time) per se: the medium needed for con-taining objects in the full-blown sense. A strange abstraction is necessary here for pre-object creatures: to learn to differentiate the places in space (and time) from the events and (pre-)objects “there”. Perceptoid signs may play a crucial role for this step, though this is not the place to continue investigating this thread of thoughts.

22 This includes more precisely all the locations in space and time that are directly connected with the pre-sent activity, i.e., not just a single (ideal) point of time or space.

Figure 22: »Contexts«, »Context Builders«, »Nomination«, »Predication«, and »Quasi-Predicates«

The human fact par excellence is perhaps not so much the creation of the tool but the domestication of time and space, i.e., the creation of a human time and a human space.

[LEROI-GOURHAN 1984, 387]

3.4.4 Secondary Object Constitution: Sortal Concepts & Geometry

While pre-objects are always referentially anchored, but cannot be accessed from another context, objects in the sense we usually associate with the expression ‘object’

are the constituting parts of many contexts. They are essentially viewed as instances of sortal concepts: perceptible, countable entities that are persistent over time even if they are not perceived, and that may even change their appearance dramatically during their lifetime (e.g., catapillar to butterfly). These are apparently also the kind of objects depicted in the most central cases of the concept »image« – from the animals of prehistorical cave paintings to ZEUXIS’ apples and grapes, from the author’s passport photograph to JUAN GRIS’ cubistic portrait of PICASSO (Fig. 6). In his Elements of Arithmetics [FREGE 1884, §54], FREGE distinguished this kind of concepts that “separate clearly and do not allow arbitrary divisions.” A chair, by means of being a chair, clearly can be separated as an individual from any other chair; and the parts of that chair are not also chairs again. Objects falling under concepts like »water« or »red« do not have these attributes: two red objects are not distinguishable by means of their being red alone.

And every part of a red surface is also red. Furthermore, sortal concepts allow us for pursuing an individual object in its singular temporal development across the contexts.

How can we be sure that something we saw this morning, e.g., a very bright star near the rising sun (let’s call it ‘the morning star’), and something we see right now in the evening, e.g., another bright star near the west horizon (correspondingly called ‘the evening star’), are the same object? Or: what is actually the communicative function of an assertion stating the identity of the morning star and the evening star? It is in fact the

Figure 23: How Can the Morning Star Be Identical to the Evening Star?

attribute of »planet« to be a sortal concept that renders an identification phrase like ‘the morning star is the same as the evening star’ to a meaningful utterance although the perceptual contexts of the two nominations are incompatible: the assertion has to be understood as ‘they are both the same planet’ (cf. Fig.s 23 & 24). That the referent of

‘the morning star’, which can merely be perceived in the morning, and the referent of

‘the evening star’, which correspondingly can be perceived only in the evening, are in fact identical, this proposition cannot be verified but by means of the criterion of spatio-temporal individuation given with the concept »planet«. In contrast to the planet Venus, i.e., the whole spatio-temporal extension of that object from its birth to its present existence (and beyond), which is only abstractly given and forms in FREGE’s terms the common “Bedeutung” (reference) of the nominators ‘the evening star’ and ‘the morning star’, the immediate sensations of the Venus at either the early morning or the late evening, are the “Gegebenheitsweisen” – ways in which the Venus is presented to us, and also the only concrete way for it to be given [FREGE 1892]. Obviously, these ‘ways of being given’ are closely related to pre-objects, their associated sensory-motor routines, and the referential anchoring basing any empirical observation.

Similarly, the referent of ‘this house’ while uttered from one particular point of view, and of ‘this house’ while uttered from a very different point of view may be the same house. In this case, the two nominators, which are in fact linguistically the same with the exception of their perceptual contexts, refer to two different manners of presentation of the same individual house meant – or, from the prespective of pre-object creatures, to two unrelated pre-objects.

In analytic philosophy, sortal concepts are conceived of as a systematic co-ordination between (a) configurational ‘Gestalt’ entities (of a ‘geometrical field of concepts’), and (b) objects involved in part-whole relations that allow us to assign functions to those ob-jects (of a ‘functional field’) (cf. Fig. 25; [VIEU 1991]). The field of objects with the functional part-whole relations, abstract as it is, does not describe or restrict in any manner the geometrical relations between an object and its parts. It only allows us to state that there are such parts, and that without this or that part, the whole object would be something different. The schema of sortal objects leads to entities that have not only parts, but also a geometrical shape and a location; and additionally, all the parts also have shapes and locations – the whole object is a configuration of the shapes of its parts.

Figure 24: Identifying the Appearances of a Sortal Object (Planet) in Several Contexts

Note that the pre-objects or views at different

“time slices” form another kind of parts of the whole sortal object.

The combination of the two fields of concepts has an interesting effect on the ability to identify corresponding instances: similar to two red ob-jects, which are not distinguishable by their being red alone, the functional parts of a car, for exam-ple, do not distinguish one car clearly from an-other one of the same type, since they both have the same functional structure, and are therefore functionally indistinguishable. Only the different geometrical components of two instances of

»car«, their different histories, allow us to distin-guish both. It is, on the other hand, not the mere geometric Gestalt that makes something a car, but the functional restrictions between its parts. Fur-thermore, it is not possible to distinguish purely by geometrical features an object from its mate-rial, e.g., a ring and the gold making it up: in many contexts, these two different objects, which stand in a particular functional relation, have the same Gestalt properties and are in-volved in the very same geometric relations.

We remember: a pre-object is usually perceived by means of just one of the sensors of the reflex arcs combined in the corresponding detector. In analogy, a sortal object is perceived by means of just one detector: of all the pre-objects covered by the concept the one pre-object that is possible in the actual situation. We cannot perceive an instance of a sortal object in its whole spatio-temporal extension but only what is given in the one, present situational context. But whereas the concept »pre-object« does not include an option of accessing the constituting reflexes as opposed to the whole pre-object, it is a central feature of sortal concepts that the corresponding “manners of being given” can be made explicit: the current perception (as of a pre-object) in opposition to the individ-ual with its complete history.23 The ability to separate the different views is indeed equivalent with the ability to access other contexts (i.e., holding the non-current views of a sortal object). The integration of a multitude of contextual views also subsumes the co-relation of the distinct perspectives of the different interlocutors in one situation.24

The field of concepts of geometric Gestalts is of particular interest for us. The in-stances in this field correspond approximately to visual pre-objects. They are immedi-ately observable. But they do not have the persistent identity of sortal objects and disap-pear if the beholder stops keeping them in his/her focus of attention. In contrast to mere pre-objects, they form however an incompatibility area of locations – a Euclidean co-ordinate system of potential, that is, not actually realized, situational contexts: space per se. Because that is after all what empty space is to be conceived of: as an infinite

23 – in opposition also to the abstract functional part-whole constituents, of course.

24 For MEAD, the anticipation of the perspective of another one forms the crucial step for pre-object crea-tures to reach “significant gescrea-tures”, i.e., context-independent communication by using signs with a common meaning, cf. Section 3.5.3.

Figure 25: Sketch on Sortal Object Constitution