K O D I K A S / C O D E Volume 40 (2017) · No. 3 - 4 Gunter Narr Verlag Tübingen Why Observation Matters A Characterization of the Sciences as Contrasted with Fiction and Religion on Semantical Grounds Claus Schlaberg (Celle) Observation is described as that which is informationally linked to the observed with the help of its being characterized both internally and externally. The external characterization refers to what perception really is (exemplified by seeing) in the manner semantic externalism treats natural kinds. Observable predicates are treated as reducible to appearance behaviour thus characterized. Referring to this way of semantical reduction distinguishes cultures of knowledge from cultures which acknowledge linguistic utterances as truthmakers. Introductory Remarks Theories of the sciences have mainly dealt with how observation confirms or disconfirms a hypothesis. The question why observation matters at all has hardly been dealt with. Instead of giving empirical evidence one could consider looking for evidence in books. Indeed, we use books in order to gain knowledge but in the sciences this is commonly justified for the reason that scientific books refer to formerly published empirical data utilized in current theories — because of their reference to observation. Imagine a text contains only true statements. Why not consider referring to it in order to decide upon the truth of a hypothesis? It does not matter whether I know for certain that its statements are true. Our concern is what I am justified to infer from the text, presupposing that its statements are true. A striking peculiarity is that in the sciences I am not justified to infer anything from it, even if the text leads me to true conclusions only. Obviously, the evidence I am finally allowed to rely on in the sciences is not a linguistic utterance be it a spoken or a written one. It seems that linguistic utterances do not count as reasonable truthmakers in the sciences as some utterances do in religion. (Quotes from the Bible or the Koran are examples.) It has traditionally been presupposed rather than justified that observation is basic in the sciences. To stress the role of observation seems trivial. How is observation related to scientific statements? In order to find an answer to this question this work will mainly refer to Carnap ’ s and Hempel ’ s ideas on confirmation by means of reduction instead of definition and then follow Barwise ’ s account of meaning and how situations support the information, including his reference to the Gricean distinction between natural and non-natural meaning as has been further developed by Searle, Strawson, Schiffer and others. One main conclusion will be that the sciences are not to count as one of the narrative practices insofar as they — though communicative — communicate indications. 1 What Remains from Empiricism? Jon Barwise characterized science as a “ search for meaning ” (Barwise 1984: 16) which seemingly suggests a conception of the world as readable or textual. In contrast to this Barwise ’ s idea is rooted in the empiristic tradition. Barwise refers to Grice ’ s distinction between natural and non-natural meaning not to misconceive the natural as readable but to somehow integrate the readable in the natural: “ [. . .] the driving force behind situation semantics is a commitment to a form of realism, [. . .] the claim that meaning does not reside in the head or in some mysterious realm but in the interaction of real, living things and their actual environment ” (Barwise 1984: 12). What does this have to do with the observable as the basis of scientific reasoning? To what extent is the observable — although theory laden — basic? The question is nowadays very important with regard to appreciating a tradition which teachers in chemistry express in a very clear sequence the steps to follow in documenting an experiment and its consequences or — as you may say — its interpretation: 1) experiment 2) observation 3) interpretation Why should we follow that order? Remember that in certain cultures what is written in a book is still regarded as true because it is written in that book. There is no doubt that reading is in many cases helpful in knowledge acquisition. However, as moreor lesseducated former attendants of secondary schools, we know that scientific books as well as school books refer to formerly published empirical data utilized in current theories. But the relation between the observable and statements which are acknowledged as scientific is closer than is often believed to be the case. Empirical data not only serve to somehow confirm, disconfirm or falsify something that is claimed to be true in the sciences. Moreover, something observable belongs to what the thesis means. That is why modern empiricists were semanticists as opposed to former empiricists who rather claimed psychological ideas concerning what is or is not innate. In addition, it will be argued here that the semiotic gap between the natural and the arbitrary is intended to lessen in the sciences. This is especially suggested by Barwise when he considers the way a clock means what time it is (Barwise 1984: 7 - 9). Accordingly, a semiotic characterization of the sciences as well as how empiricism is still at the core of what we rationally believe to make a statement true is intended. Carnap mitigated the requirements that empiricism in a very strict sense forces upon the sciences by distinguishing four levels of connection between the meaning of a linguistic expression and observation, according to whether confirmability or, additionally, testability Why Observation Matters 333 is required and whether the confirmability and the testability are or are not complete (Carnap 1953: 85). In general, descriptive predicates are expected to have some connection with possible observation (Carnap 1953: 85). The lowest degree of connection is confirmability. Carnap offers his analysis of disposition terms in order to explain his idea of confirmation. It is useful to mention that Carnap remarks that reduction determines the meaning of a term “ for some cases only, leaving its further determination for other cases to decisions which we expect to obtain in the future ” (Carnap 1953: 59), as opposed to definitions which fix the meaning of a term once for all. Concerning magnetism as an example the claim that a piece a of iron is magnetic is supported by different kinds of behaviour of a: its attracting other pieces of iron as well as its having a positive and a negative pole with the well-known behaviour of different poles attracting each other and poles of the same kind repulsing each other. Therefore, Carnap distinguishes between reduction sentences of disposition terms Q 3 some of which support that Q 3 obtains (a is magnetic) and some of which support that Q 3 does not obtain (a is non-magnetic). (R1) Q 1 ⊃ (Q 2 ⊃ Q 3 ) (R2) Q 4 ⊃ (Q 5 ⊃ -Q 3 ) In special cases Q 4 coincides with Q 1 and Q 5 with -Q 2 which results in a bilateral reduction sentence: Q 1 ⊃ (Q 3 ≡ Q 2 ). For example, being soluble in water (Q 3 ) is reducible to Q 1 and Q 2 in the following manner: a being put in water (under normal conditions) ⊃ (a dissolves ≡ a is soluble in water) The decisive point is that the reduction finally only contains observable predicates. Certainly theoretical terms have to be considered thoroughly (Carnap 1956, Hempel 1965: 173 - 222). But first, more simple cases as we are familiar with from school need to be taken into account. Being put in water and dissolving in water count as observable. Being soluble is thus reduced to observable predicates. (What an observable predicate characterizes as such will perhaps be a little clearer in the end of this paper.) Why are we interested in the observable? One answer once was given by a theorist of knowledge, Alvin Goldman, and afterwards rejected by himself: The state of observation is somehow caused by the observed and thus a state of knowledge; and knowledge is what we aim at in the sciences (Goldman1967). Goldman rejected his causal analysis of knowledge afterwards referring to fake barn situations in which fake barns would cause the same attitude towards what is seen in them as real barn situations and which thus would lack a kind of discrimination which he regards as required for knowledge (Goldman 1976). Nevertheless, there seems to be a causal connection between the observed and a state of knowledge about it. Barwise referred to the role of the ‘ flow of information ’ in knowledge. From that point in time the treatment of the topic became more semiotic than before. Barwise correctly points out that “ the ‘ truth conditions ’ on watches are that they point to t only at time t. However, a watch may point to 4 o ’ clock without that meaning that it really is 4 o ’ clock, for a variety of reasons, even if it is, in fact, 4 p. m. [. . .] ” (Barwise 1984: 8). He draws the reader ’ s attention to the flow of information which an utterance is part of. Due to the type to which it belongs, an utterance of “ The earth is flat ” has a meaning (meaning t ), but no utterance of it can ever carry the information (mean s ) that the earth is flat (meaning s ). It can rather, relative to some 334 Claus Schlaberg (Celle) constraint, mean s that the speaker believes that the earth is flat. Barwise emphasizes how conventions and natural constraints are interrelated in meanings like the meanings of time telling situations including conditions concerning the speed of sound: “ Also involved are things like natural laws about quartz, my knowing how to tell time, and the fact that the speed of sound is great enough that we can ignore the time it takes for you to hear my utterance ” (Barwise 1984: 8). We will find such a complexity in the role of observation in sciences. In a typical natural meaning, what a situation means is somehow causally related to the situation which carries that meaning. In contrast to this in the sciences observations play a role as reasons rather than as causes or effects. Observations may be good reasons for a statement taken to be true in two related ways: It is not only that fire causes smoke and thus smoke informs me about something being on fire. Additionally, a fulfilling relation obtains between the description of something as being on fire on the one hand and experiences like seeing flames, light, feeling heat, and often smoke, on the other hand. Flame-experiences are not only good reasons for the assumption that something is on fire. They, additionally, contribute to its truth in the sense of truthmaking which distinguishes it from the way a bible text tells me something or even informs me about a fire (see Smith & Simon 2007 on truthmaking). 2 The Three-Dimensional Model Let ’ s focus on one assumption which is basic in the sciences and also in everyday philosophy supported by education in school. According to this assumption we are inclined to regard reality mainly as an arrangement of entities in time and space. Even if one adheres the idea of scientific essentialism that causal powers, capacities and propensities of things do not depend on categorial properties such as number, shape, size and configuration of parts of things (and the laws of nature) but are genuine properties (Ellis 2001: 49), arrangements in space and time remain essential as long as physical force as well as speed and acceleration are introduced with the help of how such arrangements in space and time change. Things move from one location at time t 1 to another at time t 2 . And if such a movement becomes faster there is acceleration caused by force. What is usually left undone is to ask what we mean by “ y 1 is at time t 1 in location o 1 ” . It is common to refer to three spatial dimensions — one of which corresponds to <right — left>, a second corresponds to <above — below>, and a third one to <in-front-of — behind>. Accordingly, locations o i are considered as triples of values the three dimensions take respectively. o 1 may be <1; 4; 2>. o 2 , another location, be <2; 5; 3>. o 1 may be the location of a bicycle y 1 , o 2 the location of another bicycle y 2 . If you, as a spectator, look at both bicycles from <0; 0; 0> you call the bicycle y 2 right from y 1 , higher than y 1 , and behind y 1 which is mirrored in the three values of o 2 being larger than the values of o 1 . You do not see locations. What you really see are the bicycles y 1 and y 2 . You may consider seeing that y 1 is at o 1 . After some deliberation you prefer to say: You see that the bicycle y 1 is next to the bicycle y 2 or that y 2 is above y 1 or that y 2 is behind y 1 . Using “ above ” , “ below ” , “ under ” , “ behind ” or “ in front of ” generally: relations instead of triple ascriptions, is closer to how we usually describe spatial arrangements rather than referring to three spatial dimensions by means of triples. This corresponds to the concept that locations themselves Why Observation Matters 335 are not seen or perceived in any mode. They seem to be results of an abstraction gained from how we experience things as related to one another. There is a plausible explanation for this which conceives of the dimensions as useful tools in a causal explanation of experiences. Locations in the sense of triples <k 1 ; k 2 ; k 3 > exist insofar as they play a role in explanations (Quine 1961). I see the bicycle y 2 stand behind y 1 . How can I see this? One possibility consists in having one bicycle-image above another, like this (Fig. 1): Fig. 1 There is at least one further possibility, of course. It depends on from where I see the situation. If I see it from above it will appear to me as is shown above. If I see it from below, perhaps through a glass plate, it will appear like this (Fig. 2): Fig. 2 Both views support the claim that y 2 is behind y 1 . Views like these play the role which traditionally is connected with appearances or even with sense data in the visual mode. Such entities have been discredited as peculiar (Sellars 1991, Dennett 2003, Schellenberg 2017). Yet it is not at all implausible to refer to them. Some reasons are given more consideration for taking them into account. The most important reason is to see how far it will take us to proceed in the manner of using illustrations like the ones just given. They obviously are not completely incomprehensible. (It is natural to use them in books on perception, see Goldstein 2002: 112, 193, 237.) My view of the whole two-bicycle-situation includes two parts which both are bicycle views. Certain kinds of experience concerning the way the 336 Claus Schlaberg (Celle) views change during my movements enable me to identify those parts as well as the bicycles they are views of during a period of time. I more or less remain aware of which part is a y 1 appearance and which part is a y 2 -appearance. That means, my view is characterized as - a two-bicycle-appearance - a y 1 -appearance - a y 2 -appearance - a y 1 -and-y 2 -appearance And parts of it are characterized as - a bicycle-appearance - a y 1 -appearance - a y 2 -appearance respectively such that - the y 1 -appearance is below the y 2 -appearance or - the y 1 -appearance is above the y 2 -appearance. In addition to that, experiences with how my view changed during my movements (see Schellenberg 2007) justify my characterization of the bicycle-situations as seen from below or from above. That is why appearances can be described as dependant on observation conditions such as seeing the situation from above or seeing it from below. We refer to such observation conditions as B. It has to be stressed that the values <from below> or <from above> B takes are different from what “ below ” and “ above ” mean with regard to how the y 1 -appearance and the y 2 -appearance are related to each other. Spatial relations obtained between the bicycles y 1 and y 2 include being above and being below. y 2 may be in a higher position than y 1 , for example, if y 2 is on top of a hill and y 1 is not. The y 1 -appearance being below the y 2 -appearance means something else which is not to confuse as being below with regard to the appearing, the bicycles. (We may call those terms regarding appearances “ intrinsic ” , referring to Boghossian & Velleman 1989 with regard to colours. That spatial relations are not to be ascribed to appearances was remarked by Husserl and later by McGinn: Husserl stresses that it does not make sense to say that a distance of one meter obtains between a point of my view and a point of a table (Husserl 1980: 30, § 1); also McGinn 1991: 12; 2005). Subsequently we refer to spatial relations between y 1 and y 2 as F and to relations between appearances as F ′ . Whether in my view of the whole situation y 2 appears above y 1 or y 2 appears below y 1 depends upon the location from where I look at it. Accordingly, whether the y 2 -appearance in my y 1 -and-y 2 -appearance is situated above the y 1 -appearance in my y 1 -and-y 2 -appearance depends on the observation conditions B. If B includes looking at it from above: B ⊃ Above (y 2 -appearance, y 1 -appearance) In other words, looking at the y 1 -and-y 2 -situation from a higher position, from above, (B) the y 2 -appearance is located above the y 1 -appearance. Looking at it from a lower position, from below, the y 2 -appearance is located below the y 1 -appearance. Generally: Why Observation Matters 337 B ⊃ F ′ (y 2 -appearance, y 1 -appearance) B takes looking-from-below or looking-from-above as values. F ′ takes being above or below in the intrinsic sense as values. The common experiences described so far are explained causally in the following manner (Fig. 3): Y1-image Y2-image y 2 y 1 Y1-image Y2-image k 2 k 1 Fig. 3 Light rays meet the retina differently depending on whether the observer is located above or below (B) the situation. As a result the y 1 -image on the retina is situated above the y 2 -image on the retina or below the y 2 -image on the retina respectively (F ′ ). The y 1 -appearance and the y 2 -appearance we have mirror the further processing of the retinal images in the brain about which we probably will never have complete knowledge. What matters here concerning the explanation of how the y 1 -appearance and the y 2 appearance are situated within the y 1 -and-y 2 -appearance, is that the ascription of triples <k 1 ; k 2 ; k 3 > to everything that plays a causal role is decisive. The explanation assumes that - to each (relevant) part of the eye and each (relevant) part of what is between y 1 and y 2 on the one hand and the eye on the other hand a location <k 1 ; k 2 ; k 3 > is ascribed. - the light moves straight. - y 1 and y 2 emit (mostly reflect) light. - the anatomy of the eye is such that the light rays emitted by y 1 and y 2 cross within the eye. 338 Claus Schlaberg (Celle) The spatial arrangement of y 1 and y 2 as well as the straight movement of the light make plausible how the light meets the retina. The explanation supposes an ascription of triples <k 1 ; k 2 ; k 3 > to each relevant part of the observation situation. If, for example, the first value of the location o 2 of y 2 is smaller than the first value of the location o 1 of y 1 , then the second value ascribed to the y 2 -image is smaller than the second value ascribed to the y 1 -image if the eye is located above y 1 and y 2 .If it is located below the situation, things are reversed: If the first value of the location o 2 of y 2 is smaller than the first value of the location o 1 of y 1 , then the second value ascribed to the y 2 -image is larger than the second value ascribed to the y 1 -image. From the spectator ’ s point of view, y 2 being behind y 1 is indicated to the spectator by an appearance behaviour which is such that 1) the y 2 -appearance is below the y 1 appearance if the spectator is below y 1 and y 2 and 2) the y 2 -appearance is above the y 1 appearance if the spectator is above y 1 and y 2 . The explanation offered is almost trivial because it belongs to knowledge acquired in school days. The more trivial it seems the more important it is to stress that it presupposes (or, if explicitly said, supposes) that to each part of the observation situation a triple <k 1 ; k 2 ; k 3 > is ascribed as its location. What is regarded as actually seen in that two-bicycleobserving-situation are the bicycles although a lot more things contribute to the explanation. That y 2 is behind y 1 may also count as what is seen. But the location triples as well as the light which moves from y 1 and y 2 to the retina are surely not accepted as actually seen. In accordance with how educated the person is, someone who explains how the y 1 -and-y 2 -appearance changes during the observer ’ s movements will refer to locations as triples <k 1 ; k 2 ; k 3 > and light moving straight from y 1 and y 2 to the retina. Referring to these unseen parts of the observing situation one will admit that he neither sees triples nor light moving and crossing on its way to the retina. (As Wittgenstein remarks, the observer anyway does not see his own eyes, Wittgenstein 1963: 91; 5.633.) The distinction made between what is seen on the one hand and parts of the seeing situation which contribute to the seeing and nevertheless are not seen themselves on the other hand is stressed in order to prepare for a well-known phenomenological distinction: There is an object seen and, additionally, there is the process of seeing including a directedness towards the object. Whatever you see, you don ’ t see triples as where the things are located (Table 1). Table 1 visible non-visible y 1, y 2 o 1 , o 2 , k 1 , k 2 , k 3 , <k 1 ; k 2 ; k 3 >, light The non-visible and the visible serve to explain that we see y 1 and y 2 and how their appearance changes likewise. They are likewise natural insofar as they are taken into account in physical or physiological explanations. Besides, there are the things we just ‘ have ’ . These are the things involved in seeing that we are at least as familiar with as with the actually seen things and which, nevertheless, are not assumed to exist in space: the appearances or views. One of the reasons why they are to be admitted is that two of the most common pictures — say: two photographs of the same thing, say: a bicycle — may differ only with regard to the camera ’ s point of view or with regard to the luminous intensity or from Why Observation Matters 339 where the light comes. Hence they are pictures of the same object which differ in how the visible y appears, that means: they differ regarding the y-appearance they correspond to. Its resemblance to a y-appearance enables you to see it as a y-picture (see Posner 2010 who reintroduces similarity in the theory of pictures after it was abandoned by Goodman 1968: 3 - 43.) Concerning the distinction between the visible entities involved in seeing on the one hand and those entities involved in seeing which are taken into account in a causal explanation on the other hand, it is remarkable that y 1 and y 2 belong to both. y 1 as well as y 2 are both what you see and what reflects light that meets the retina. After further consideration terms like “ the one bicycle ” and “ the other bicycle ” will be replaced by more theoretical expressions. Before doing so you nevertheless refer to what you see — the bicycles — even in theory. What has been said so far amounts to the following structuring (Table 2): Table 2 just had visible non-visible y 1 -appearance y 2 -appearance y 1 y 2 o 1 , o 2 , … k 1 , k 2 , k 3 , … <k 1 ; k 2 ; k 3 > … light theoretical and non-visible (theoretical in the narrower sense) theoretical familiar Remember how triple ascriptions are meant to explain the retinal y 2 -image being below the retinal y 1 -image. The explanation includes y 1 , y 2 , the light and the retinal images equally. Of course, how exactly appearances — the ‘ just had ’ entities — and retinal images are connected remains more or less mysterious. But the way retinal images are expected to behave in accordance with the explanation fits well with the way appearances behave. As a consequence, we take what is suitable to explain how retinal images behave as suitable to explain how appearances behave. Although the appearances do not belong to the model which assumes a spatial arrangement of real things, it seems inevitable to take into account the visible, the non-visible and the ‘ just had ’ entities equally in order to explain what is intended to explain. The appearance behaviour is what everything else in the explanation 340 Claus Schlaberg (Celle) depends on. We assume y 1 , y 2 , the light, the way the light moves and the observer to be the way we do because we suppose the retinal images to behave in accordance with the appearances. 3 Truthmakers for Statements on Reality and Fiction What is the meaning of “ behind ” ? Before we draw our attention directly to the meaning of “ behind ” , the relation between what the three-dimensional model explains on the one hand and the behaviour of appearances on the other hand needs to be described more vividly. The behaviour of y-experiences is explained with the help of the three-dimensional model including the ascription of triples <k 1 ; k 2 ; k 3 > to y 1 , y 2 , the light and the retina as their locations respectively. The appearances make a contribution in the opposite direction. Their behaviour is what the existence of y 1 and y 2 makes true. Unfamiliar behaviour of appearances makes you think of y 1 and y 2 as being ghostly, somehow unreal, for example when a y-appearance suddenly vanishes or comes back (see Schlaberg 2017: 113 - 117). Therefore, we regard appearance behaviour as a truth maker. Experiences in any further mode be also taken into account. Imagine you have a stretching-your-hand-straight-to-yappearance accompanied by a touching-y-appearance, but no sensation of touching y. To some degree this makes you doubt the existence of y if there is no reason to believe that your senses do not work properly. The relationship between appearance behaviour — more generally: experience — on the one hand and the real including its spatial relations to other real things as modelled three dimensionally on the other hand is characterized by two oppositional directions (see Schlaberg 2017: 147 - 160): Appearance behaviour makes real existence and obtaining spatial relations between really existent things true as the assumption of really existent things serves to explain appearance behaviour (and experience more generally) causally (Fig. 4): y-appearance behaviour y exists in <k 1 ; k 2 ; k 3 > makes true explains causally Fig. 4 Why Observation Matters 341 Both, y-appearance behaviour as well as spatial arrangements of entities whose assumption serves to explain how y-appearances behave within the three dimensional model which ascribes to each of the things a triple <k 1 ; k 2 ; k 3 > as its location, are involved in the meanings of several terms, one of which is “ behind ” . Their semantic role becomes clearer when we consider the idea of semantic externalism. As will finally become clearer, the observable is of interest because of the information it carries about the observed we do research on. Information may be gathered from several sources. There are some peculiarities of how y and y-appearances are related to each other because of which the latter are of special interest. 3.1. Internal and External Characterization of Appearances First of all, the meaning of “ y-appearance ” has to be explained. Goodman suggests to classify pictures irrespectively of what they denote: “ What tends to mislead us is that such locutions as ‘ picture of ’ and ‘ represents ’ have the appearance of mannerly two-place predicates and can sometimes be so interpreted. But ‘ picture of Pickwick ’ and ‘ represents a unicorn ’ are better considered unbreakable one place predicates, or class terms, like ‘ desk ’ and ‘ table ’ ” (Goodman 1968: 21). There is a remarkable parallel between how Chisholm characterizes appearances and how Goodman describes pictures. Chisholm compares “ The cabin on the hill is blue ” to “ I am now appeared to blue ” (Chisholm 1957: 62; with regard to certainty which is not our concern here). Such ‘ adverbial ’ treatment characterizes perception internally rather than as related to the perceived. Similarly Goodman describes pictures internally rather than as related to what the picture represents (if it represents anything at all). Now we are justified to trace a picture being a y-picture, for example a Pegasus-picture, back to references we make as a spectator, as a painter or whoever deals with it. This amounts to Chisholm ’ s treatment of perception: Being a y-picture, for example a Holmes-picture, is based on references which can be described correctly as y-references without having to assume that any y exists. An illustration in a book by Conan Doyle may be rightly regarded as Holmes-picture notwithstanding that there is no Holmes to whom it could be related causally like a photograph. The same applies to references which may be classified as y-references respectively independently of whether there is any y. Likewise the term “ bicycle-appearance ” contains two parts, the first of which specifies a subset of what the second denotes. ֎ -appearances are a special kind of appearances. ֎ does not have to denote anything but makes the whole expression more specific. We call the way the term for y characterizes y-appearances an internal characterization. Another example may be helpful: I have many thoughts about whether the monster of Loch Ness will someday kill Sherlock Holmes in front of my house. Even if there is neither a monster of Loch Ness nor Sherlock Holmes somewhere in the universe, my thoughts nevertheless refer to the monster of Loch Ness and to Sherlock Holmes. In order to avoid any terms for non-existing individuals we use terms which — although not denoting individuals — serve to specify a subset of what “ thought ” denotes. The way these seemingly individual denoting expressions modify the meaning of “ thought ” is adequately conceived of as specifying a subset of the set of thoughts in a regular manner whose result is clearly not arbitrary. In addition to their internal characterization, y-appearances are characterized externally. In short: A y-appearance is what it is like to see y, and seeing y is just what it really is. Using 342 Claus Schlaberg (Celle) “ to see ” correctly does not require the speaker to know exactly how seeing works physiologically, but it requires him to refer to what seeing actually is as is investigated by researchers and probably will never be resolved in every detail. Putnam ’ s idea of linguistic labour division be applied here as is usual with regard to the term “ water ” (Putnam 1975). What do we mean by “ water ” ? Roughly said, “ water ” means what water really is. Referring to Putnam, Chalmers made the distinction between primary and secondary intension: “ Given that the primary intension ( ‘ watery stuff ’ ) picks out H 2 O in the actual world, it follows from rigidification that the secondary intension picks out H 2 O in all possible worlds ” (Chalmers 1996: 59). Comparing “ water ” with “ bachelor ” a difference is striking: What bachelors are completely depends on a convention according to which someone is a bachelor if and only if he is an unmarried man (see Haslanger 2000, 2010 concerning so called “ social terms ” as opposed to natural kinds). Regarding “ water ” the case is different. What counts as water does not solely depend on a convention. On the one hand there are criteria as to categorize something as water concerning how it is experienced. On the other hand, what at first sight looks like, smells like or tastes like . . . water may not be water. Hence further criteria which refer to a microstructure are to consider — criteria which are to discover by means of empirical investigation. The experts who do the respective research participate in the same “ linguistic labor division ” as does the common member of the linguistic community who uses the term “ water ” . They all know that the meaning of “ water ” resides in the water itself — hence the term “ semantic externalism ” . What does this tell us about the meaning of “ behind ” ? y 2 being behind y 1 is what it really is. In order to know whether y 2 is behind y 1 we consult certain experiences, for example appearances. Most prominently among them, y 1 -appearances and y 2 -appearances are to be considered. We are acquainted with kinds of appearances and appearance behaviour that we consider in order to decide on whether one thing is situated behind another. It suffices to give examples as has already been done so far: the way y 1 -appearances and y 2 -appearances are located below or above one another in dependence of from where I seem to look at the y 1 -and-y 2 -situation; additionally, the way y-appearances reduce or enlarge as dependant on how far I am from y. The crucial point is that the decision of whether y 2 is behind y 1 is made upon the behaviour of y 1 -appearances, y 2 -appearances, and y 1 -and-y 2 -appearances. Nevertheless, generally we are not certain about what it means that the one thing is located behind the other as long as we are not certain about what space generally consists in. We experienced y 1 -and-y 2 -appearances to change systematically during our movements (see Noë 2004, Schellenberg 2007). The usual ascription of triples <k 1 ; k 2 ; k 3 > to y 1 and y 2 as their locations respectively serves as part of an explanation of the way these appearances behave. y 2 being behind y 1 is indicated by the one being covered by the other, the y 2 -appearance being above or below the y 1 -appearance, and both enlarging during my moving-forwardexperience including that sooner or later a passing-by-y 1 -and-after-that-passing-by-y 2 experience occurs. These experiences belong to what under the heading “ y-appearancebehaviour ” has been juxtaposed (Fig. 4) with the theoretic (though not necessarily nonvisible) entity y as is ascribed to a location triple <k 1 ; k 2 ; k 3 > which itself belongs to the theoretic (and itself is non-visible). The appearance behaviour makes true that y 2 is behind y 1 whereas to y 2 and y 1 themselves theoretic location triples are ascribed in such a way that the y 1 -and-y 2 -appearance behaviour is explained with the help of these ascriptions. In other Why Observation Matters 343 words: The theory is meant to explain which relations F ‘ between y 1 -appearances and y 2 appearances obtain. F ′ (y 1 -appearance, y 2 -appearance) justifies the claim F(y 2 , y 1 ); “ behind ” is an example of F here. But it has to be stressed that the meaning of “ behind ” includes F ′ (y 1 appearance, y 2 -appearance) as well as a more or less detailed account of how y 1 and y 2 are located within a theoretically proposed model. This is similar to how the H 2 O model and how water is experienced both are part of what “ water ” means as Chalmers points out by distinguishing a secondary from a primary intension. Things are different as fictional individuals and fictional kinds are concerned. Whereas statements about one thing y 2 being behind another y 1 are to be verified or confirmed by y 1 and-y 2 -experiences, statements about Sherlock Holmes are to be made true by written or spoken words as they testify how people refer to Holmes. Kripke points out that things like Holmes or unicorns could not have been real (Kripke 1980). Real things are called as they have been baptized in real baptism situations. What deserves to be called “ water ” is of the same kind as what has been given this name in a baptism situation. As opposed to water, in no possible world Holmes or unicorns could have been given their names. If something y belongs to a category K, this is due to a historical — hence real — baptism situation in which examples of K have been baptized K. If such examples could have been given their names only in situations of the real world there is no possible world which includes examples of K as Holmes or unicorns and the like concerns. Consequently there is neither Holmes nor a unicorn in any possible world. Nevertheless, statements about fictional entities like Holmes or unicorns can be more or less true. Hence the question arises as to what the truth of these statements is decided upon. Insofar as it is true that Holmes is a detective but not a horse there must be something that truth and falsity concerning Holmes depends on. Indeed baptism situations concerning fictional entities have been considered (see Zalta 1987). It is peculiar to entities like Holmes that they have been invented rather than discovered. The baptism situation in which Holmes has been baptized Holmes, the situation in which Holmes was given his name, was a situation of invention, thus a situation of a mental reference characterized internally as Holmes-reference. As are y-appearances and more generally y-experiences, y-references in the case of fictional entities y are internally characterized. Yet statements on Holmes-references are not to be considered as made true by experiences, as occur in understanding the plot of novels or films. The reason is that as fictional entities y which are known as such concerns, y-references are not characterized externally the way water-experiences are with regard to a more or less known theoretical structure. Although in understanding the plot of fiction y-references occur there is no theory about the way y-references themselves contribute causally to how they are experienced in events of understanding a plot. y-references emerge from such mental processes rather than making causal contributions to them. Hence one difficulty with regard to fictional entities y arises from the peculiarity that y-references are not such that claims about y are made true by y-experience characterized externally. We do not ever explain y-reference behaviour referring to y. Hence experiences as occur in reading novels or watching films are not characterized externally as y-experiences regarding to fictional things y. 344 Claus Schlaberg (Celle) 4 Indication and Communication To give an example of how one and the same message may be conveyed in two essentially distinct manners compare: a) flames, light, heat and smoke as produced when something burns b) an utterance of “ fire! ” as expected from someone who intends to warn people about fire a) and b) are essentially different although they are suitable to convey the same message. a) is typical of the case when an indication is given whereas b) is typical of when someone intends to convey a message in such a way that we are justified to say that he means something by making his utterance. This has been the starting point of the Gricean account of meaning (Grice 1957). Schiffer (1972) summarizes a discussion to which Grice (1957, 1968, 1969), Strawson (1964), Searle (1975) and Schiffer (1972) made contributions (more recently Moore 2017). Barwise (1984) refers to it as a distinction between natural and conventional meanings. Posner (1993) presents an approach meant to include both a) and b) as sign processes. This fits well into Barwise ’ s idea of constraints (Barwise 1984: 12 - 14). Referring to Barwise ’ s idea, constraints are conceived of as being natural in some cases and being conventional in other cases. A constraint S => S ‘ links situation types S and S ′ . If a situation s is of type S according to the constraint it is also of type S ‘ . Barwise gives an example which is apt to show all the difficulties: If a watch says that it is 4 p. m. according to a constraint it is 4 p. m. Which kind does the constraint belong to? Is it natural or conventional? “ For a theory of chronometric reading, the state of the watch does not fall cleanly into either category of sign or symbol. Normal laws about quartz, for example, are exploited in the design of the watch and are responsible for its pointing to 4 p. m. at 4 p. m. and not at noon. On the other hand, the system of time measurement that the watch is measuring is conventional. [. . .] Timepieces, like people, can tell us the wrong time, if they do not fit the conventions within which they are placed ” (Barwise 1984: 7). Barwise points out that although the truth conditions of “ It is 4 p. m. ” are that a statement made with it is true if and only if it is 4 p. m. at the time and place of utterance, my statement of “ It is 4 p. m. ” does not necessarily mean that it is 4 p. m., even if it is true. The reason is that the flow of information does not fulfil the conditions necessary for a watch really meaning what time it is, for example when it is out of order. Two kinds of meaning are to be considered here, the one which concerns situation types and the other which concerns a particular situation. The little hand of the watch pointing to four and the big hand pointing to twelve belongs to a type which means t that it is four o ’ clock whereas only in particular situations this means s that it is really four o ’ clock. Only a particular situation carries information. It depends on a constraint that links situation types because of which a situation of the one type carries the information that it belongs to the other type. Whether a situation of the two hands pointing to numbers four and twelve respectively carries the information — means s — that it is four o ’ clock, depends on how information flows. The sentence “ The earth is flat ” “ has a meaning t as projected by the rules of English, but no utterance of it can ever mean s that the earth is flat. It can, of course, relative to certain other constraints, mean s that the speaker believes that the earth is flat ” (Barwise 1984: 15). This example illustrates well how details of the information flow determine what a situation means s : “ To understand inference [. . .] we need to recognize that Why Observation Matters 345 there are, in general, two parallel sets of constraints, one on some activityA and the other on cognitive activity about A, and we need to understand the relation that enables cognitive activity to adequately ’ track ’ the activity it is about ” (Barwise 1984: 15). This paper, including its application of the concepts of internal characterization and external characterization of experiences, especially appearances, and references is intended to show how the flow of information from observation to a truth bearer meets the requirements under which the observation situation is meaningful s in the relevant manner. We are dealing with cases in which observation obviously matters in order to find out why it matters. It is decisive to insist on a semantical relation between the truth bearer and observation which has been focussed in the early days of neopositivism. The claim that y 2 is behind y 1 seemingly abbreviates things similarly to how the claim that a piece of iron a is magnetic seems to abbreviate things. In order to give some further explanations we have to turn to Carnap ’ s disposition analysis. Hempel shows that reduction sentences combine “ the stipulative assignment of meaning, and the assertion or description of empirical fact ” (Hempel 1965: 207): “ If, for example, the term ‘ Q ’ is introduced by the two reduction sentences (8.1) C 1 x ⊃ (Qx ≡ E 1 x) (8.2) C 2 x ⊃ (Qx ≡ E 2 x) then the following empirical law is asserted by implication: (8.3) (x) [(C 1 x • E 1 x) ⊃ (C 2 x ⊃ E 2 x)] i. e., roughly speaking: any object that shows a positive response under the first test condition will, when put into the second test condition, show a positive response as well. Thus, a chain of reduction sentences for a given term normally combines two functions of language that are often considered as sharply distinct: the stipulative assignment of meaning, and the assertion or description of empirical fact ” (Hempel 1965: 207). Carnap emphasizes that “ in most cases a predicate will be introduced by either several reduction pairs or several bilateral reduction sentences ” . As an example: “ The intensity of an electric current can be measured for instance by measuring the heat produced in the conductor, or the deviation of a magnetic needle, or the quantity of silver separated out of a solution, or the quantity of hydrogen separated out of water etc. We may state a set of bilateral reduction sentences, one corresponding to each of these methods. ” Now the decisive point: “ The factual content of this set is not null because it comprehends such sentences as e. g. ‘ If the deviation of a magnetic needle is such and such then the quantity of silver separated in one minute is such and such, and vice versa ’ which do not contain the term ‘ intensity of electric current ’ , and which obviously are synthetic ” (Carnap 1953: 56). The conclusion is that introductions of predicates like “ magnetic ” are not mere abbreviations. Although reduction sentences may contain different observable predicates, these observable predicates are not linked by mere convention. They are linked somehow naturally as can only be discovered empirically. The same applies to such seemingly basic predicates like “ behind ” . One reduction sentence be: Behind (y 2 , y 1 ) ⊃ (B 1 ⊃ Above (y 2 -appearance, y 1 -appearance)) 346 Claus Schlaberg (Celle) (B 1 denotes looking from above at the y 1 -and-y 2 -situation.) The whole sentence says that an y 2 -appearance is located above (in the intrinsic sense) the y 1 -appearance if you look at y 1 and y 2 from above. Certainly it has to be mentioned that whether you look from above or from below is supported by going-upwards-experiences and going-downwards-experiences (internally characterized as such) during one ’ s own movements, regarding to which we are not striving for detailed descriptions. What matters is characterized roughly as follows. As only two relevant reduction sentences be given: R1) Behind (y 2 , y 1 ) ⊃ (B 1 ⊃ Above (y 2 -appearance, y 1 -appearance)) R2) Behind (y 2 , y 1 ) ⊃ (B 2 ⊃ Below (y 2 -appearance, y 1 -appearance)) (B 2 is looking from below, perhaps with the help of a glass plate.) Both sentences state different appearance behaviours respectively linked to the same predicate “ behind ” which belongs to what Carnap calls “ thing language ” (Carnap 1953: 69), whereas “ above ” , “ below ” , “ y 1 appearance ” , and “ y 2 -appearance ” refer to relations between or to properties of parts of the visual experience and thus rather belong to how we refer to what it is like to see y 1 and y 2 than to the ‘ thing language ’ . The foregoing considerations were not only meant to clarify that “ behind ” is not completely reducible to a bilateral reduction sentence. The observation behaviour stated in (B 1 ⊃ Above (y 2 -appearance, y 1 -appearance)) is linked to the observation behaviour stated in (B 2 ⊃ Below (y 2 -appearance, y 1 -appearance)) insofar as both reduce being behind to observation behaviours which remain different however. The reason why they are linked is empirical. As opposed to empirically supported connections mere abbreviations are purely conventional. Hence “ behind ” or “ magnetic ” are not mere abbreviations. y 2 being behind y 1 does not abbreviate that y 2 -appearances and y 1 appearances behave so-and-so. y 1 -and-y 2 -appearance behaviour rather indicates that y 2 is behind y 1 . The indicative character becomes obvious when you consider that you may be erroneous in inferring from (B 1 ⊃ Above (y 2 -appearance, y 1 -appearance)) that y 2 is behind y 1 . The reason why (B 1 ⊃ Above (y 2 -appearance, y 1 -appearance)) obtains may be that you look at an anamorphous picture. If the picture is made precise enough you will not realize the illusion. In contrast to the relation between an indicator and what it indicates the relation between what is abbreviated on the one hand and the abbreviation on the other hand cannot be erroneous. If “ Behind (y 2 , y 1 ) ” was just an abbreviation of detailed descriptions of how y 1 -and-y 2 -appearances behave in dependence on observation conditions B, one could not be mistaken in inferring that y 2 is behind y 1 from such a detailed description. (Besides, the appearance behaviour which is to be considered responsible for y 2 being behind y 1 includes infinitely many B-F ‘ -combinations matching infinitely many periods of time. y 2 being behind y 1 implies that y 2 -appearances are situated above y 1 appearances at time t i if you look at y 1 and y 2 from above at time t i and likewise for any other periods of time t k . If the relation between Behind (y 2 , y 1 ) and (B i ⊃ F ′ (y 2 -appearance, y 1 appearance)) was stipulative as abbreviations are taken to be it was not a good candidate for being based on inductive inference. Hence we should rather conceive of it as what Barwise regards as constrained naturally rather than conventionally.) Abbreviations are essentially communicative rather than solely indicative insofar as the hearer is expected to realize the speaker ’ s intentions in order to understand the respective abbreviation. This will be explained subsequently. From now on, supposing that the relation between (B 1 ⊃ Above (y 2 - Why Observation Matters 347 appearance, y 1 -appearance)) or (B 2 ⊃ Below (y 2 -appearance, y 1 -appearance)) on the one hand and Behind (y 2 , y 1 ) on the other hand is not abbreviative we are free to consider it indicative. The task of explaining what indication consists of remains to be performed. Yet it is more useful to start with what communication as opposed to pure indication consists of. 4.1. The Gricean Account As one of its most elaborated results the Gricean account of meaning was developed further by Schiffer who gives the following example (Schiffer1972: 17 - 30): - S wants A to believe that the house is rat-infested. - S decides to bring about this belief in A by letting loose a sewer rat. - S knows that A believes that S is unaware that A is watching him. - It is S ’ s intention that A should infer from the fact that S let the rat loose that S did so with the intention that A should infer that the house is rat-infested. - S knows that A will believe that S would not so contrive to get A to believe that the house is rat-infested unless S had very good reasons for thinking it was. - S expects A to infer that the house is rat-infested from the fact that S is letting the rat loose with the intention of getting A to believe that the house is rat-infested. This is surely not a case of communication. Strawson (1964: 446 - 447) suggests adding the further condition that - S has the intention that A should recognize S ’ s intention that A infer (at least in part) from the fact that x is f that S uttered x intending S ’ s utterance of x to produce a certain response r in A. In the case of letting the rat loose: S has the intention that A should recognize S ’ s intention that A infer (from x being so and so . . .) that S let the rat loose intending that this gets A to believe that the house is rat-infested. A way to exclude a further counter-example is to add the condition that S have the intention - that A should recognize S ’ s intention that A ’ s recognition of S ’ s intention to get A to produce a certain response r shall function as (at least part of ) A ’ s reason for producing the response r. Consider the “ Moon over Miami ” -example and subsequently the “ Tipperary ” -example (Schiffer 1972: 18 - 22): - S intends that A will believe that S plans to get rid of A by means of S ’ s repulsive singing of “ Moon over Miami ” , but S expects and intends that A ’ s reason for leaving will in fact be A ’ s recognition of S ’ s intention to get him to leave. - By singing “ Tipperary ” S intends A to believe what S intends by singing in the “ Moon over Miami ” example. Although here S intends to show what he really intends this example also is not one in which S means something by singing the song (Fig. 5). 348 Claus Schlaberg (Celle) S intends A to believe that S intends: Singing --> Leaving S intends: reason for leaving “Tipperary”: S intends A to believe: Fig. 5 At least for x to count as an act of communication both x being public among S and A as well as the reasons S intends A to have for producing r being public and commonly known among S and H are necessary (Schiffer 1972: 30 - 36). 4.2. Communicating Indications in the Sciences As the respective contributions of communicative and non-communicative acts concerns there is something peculiar to the sciences: No doubt, as a scientist Stephen Schiffer meant something in the communicative sense by writing the following: “ [. . .] S meant something by uttering x just in case S uttered x intending A to recognize that S uttered x intending to produce a response r in A by means of A ’ s recognition of S ’ s intention to produce r in A ” (Schiffer 1972: 19). Scientific statements like Schiffer ’ s are communicative in the sense that something is meant by them like any other statement. However what contributes to their scientific character is that they are intended to give reasons which are convincing, that means that they produce a conviction in the audience not by means of the audience ’ s recognition of the author ’ s intention to produce that conviction (see Schiffer 1972: 42 - 48). The reasons the scientist gives in order to convince his audience play the role of indicators rather than the role of communicative intentions as are expected to be the crucial reasons in requests like “ May you pass the salt? ” . Referring back to the statement that y 2 is behind y 1 , one actually has to give reasons for this statement. And the reasons he gives contain that truth-making requirements are satisfied. As such requirements statements of the form (B ⊃ F ′ (y 2 -appearance, y 1 -appearance)) have been suggested. Hence the most peculiar characteristic of scientific utterances is mainly that they communicate indications! By the way, this contradicts how assertions as a kind of speech acts have often been characterized. Meggle (1993: 491) states that x is an act of communication ( “ Kommunikationsversuch ” ) of S directed to H with regard to the propositional content p iff S intends to bring about the belief that p by uttering x and S believes that H believes p if and only if H knows that S intends* H to believe that p ( “ * ” indicates a further restriction called “ absolute Offenheit ” by Meggle). This is mistaken because the more scientific the situation of my utterance, the more the reason why I expect my audience to believe that p is true is that truthmaking conditions of p are satisfied. Only in very non-scientific situations, my Why Observation Matters 349 intention to make H believe that p can more or less be expected to be a reason for H to believe that p (see Schiffer 1972: 42 - 48). The way the indication — meaning naturally — is intended by meaning something nonnaturally, thus by communication, is illustrated with the help of the following example. Let ’ s subsume discoverers of the 15th or 16th century in the category of geographers. Certainly their discoveries are taken for granted nowadays. Nevertheless, they are to count as scientific insofar as they contributed to what belongs to common geographical knowledge nowadays. The discoverer who provides his audience with information to understand that there is a country behind the ocean which has not been entered by members of his civilization previously means that there is a country behind the ocean, more precisely: He intends his audience to respond by recognizing what he wants to tell them by means of recognition of his intention. Nevertheless, he surely does not rely on his audience ’ s credulity, but rather aims at convincing them. The way he tries to evoke the intended response is offering empirical evidence. It belongs among such evidence that after some time of travelling at sea in a westerly direction land will be seen (Fig. 6): Fig. 6: http: / / www.gudrunv.com/ uploads/ blog/ 2009/ 12/ land-in-sight.jpg (accessed 06. 06. 18) The scientific community is the audience which the scientist addresses. On the one hand he means what he intends them to believe. The response he intends to cause in them is mainly the belief that he wants to convince them of . . . On the other hand he gives reasons which are to serve as plausible whether they are empirical evidence in a theory or a logical inference. The reasons indicate what his thesis claims rather than cause a belief in the thesis by means 350 Claus Schlaberg (Celle) of recognizing the intention to cause it. Again, the empirical evidence as given in the form (B ⊃ F ′ (y 2 -appearance, y 1 -appearance)) will be supplied the theoretical explanation which consists in ascriptions of locations <k 1 ; k 2 ; k 3 > to each part of the observing situation. Having a land-appearance above the water-appearance is easy to explain by referring to locations of the observer, the land, the water and the relevant light rays. These location ascriptions as well as the appearance behaviour constitute what we mean by the coast being behind the sea. The appearance behaviour contributes to making true the theoretically assumed arrangement of entities in space which serves to explain the appearance behaviour causally. A difference in the respective valuation of communication and indication may be justified. It concerns how we are involved in the flow of information from what the information means to its carrier: Communicating on a subject m as well as understanding communication on a subject m does not require any interest in m at all. We are able to understand a discourse on m and we are able to work out treatises on m without being affected by m in any way. Knowledge acquisition in school and in universities would not be possible if it depended upon interest. This is partially due to its communicative character. In contrast to the non-requirement of interest on the topic for communicating on it, pure indication depends on interest: Surroundings are more or less suitable for what I am trying to achieve. The aims I try to achieve depend on my interests. These surely influence the way I perceive my surroundings generally (see Gibson 1979 on affordance). An example will help to clarify how one ’ s interests are related to indication and communication respectively: S distributes rat poison in the house. This may serve to poison rats as well as giving H to understand that the house is rat-infested. It depends on further states obtaining in the situation. S putting a package of rat poison on the table so that S and H commonly know that S puts the package on the table differs from simply poisoning the rats, at least as follows: The act of putting the package on the table as such is useless with regard to killing the rats. It rather serves to show the internal state of S ’ s mind. And it is rational from S to intend H to realize what S tries to show — if, for example, S and H commonly know that the house being rat-infested has been considered by both occasionally. Now imagine how you react to S ’ s putting the package on the table if you are used to S ’ s having been afraid of rats for many years in which no rats ever could be found in the house. S ’ s act is then likely not to be taken seriously. Compare this to a situation where you see a rat running down the stairs in your cellar. Seeing the rat is likely to affect you immediately. (Or think of an alligator instead.) Considering the example may help to attain more general results: Seeing a rat running down the cellar stairs indicates that there is a rat in the cellar. There is a seeing situation s 1 which links the observer a to a situation seen, s 2 , which supports an information unit, an infon << ∃ x (Rat (x) ∧ Runs down the stairs (x)) ; 1>>: s 1 ╞ <<Sees (a, s 2 ) ; 1>> s 2 ╞ << ∃ x (Rat (x) ∧ Runs down the stairs (x)) ; 1>> The indication is a relation that obtains between a (the observer) and the proposition s 2 ╞ << ∃ x (Rat (x) ∧ Runs down the stairs (x)) ; 1>> (on indication see Posner 1993; propositions are modelled as Austinian propositions in situation semantics, see Cooper Why Observation Matters 351 1997). No constraint in the original sense is involved thus far. Rather, the support relation is crucial as is the way it is connected with situations both internally and externally characterized. It is hardly possible to make a distinction between a perceived situation and the way it is characterized internally. The infon supported by s 2 accords to how s 2 is characterized internally as opposed to further inferences from seeing s 2 you may be allowed to make. This does not contradict any assumptions that what a situation supports may depend on cultural impacts. It rather links perception closer to behaviour dispositions than to considerations concerning a sender ’ s intentions. On the background of common considerations concerning whether the house is ratinfested the situations s 3 and s 4 be constrained as follows: s 3 ╞ <<Puts a package of rat poison on the table (b); 1>> s 4 ╞ <<Intends a to believe that the house is rat-infested (b); 1>> s 3 and s 4 are of situation types S 3 and S 4 respectively, linked by a constraint S 3 => S 4 . Being attuned to this constraint (see Barwise 1984: 12) admits a to infer from seeing s 3 that there is a situation of the type S 4 in which b intends a to believe that the house is rat-infested. This inference is surely not justified to the same degree to which the support relation between s 2 and the information that a rat is running down the stairs is justified. In addition, the inference from s 3 to s 2 is justified to an even less degree. b ’ s putting a package of rat poison on the table can have meanings other than or in addition to the meaning that the house is rat-infested. Consequently, the communicative act b performs by putting rat poison on the table is not as likely to arouse your interest as the situation to which a rat itself seems to belong. What links a seeing situation s 1 to the situation seen s 2 differs significantly from the constraint that links a communication situation to the sender ’ s intention and even more to the focal situation the sender ’ s utterance is about (see Devlin 2001: 6). Although they supplement your everyday information, communicative acts are generally not as important to your life as indicators and, most prominently among them, perceived situations (s 2 ) are with regard to what is perceived in them (which depends on how s 2 is characterized with regard to an infon). In the case of situations of linguistic utterances of “ The house is ratinfested ” the constraint that links the utterance situation type to what is meant, the type of focal situation in which the house is rat-infested, is to a high degree conventional, including conventions that concern what is meant by “ the house ”— the house certainly is not any house, but one the speaker expects the hearer to know to be the one the speaker refers to. Corresponding to the support relation between the perceived situation and the infon that enables one to categorize the perceived situation immediately, gathering information by perceiving is obviously more relevant to the observer than gathering information by understanding. One has to realize which type a situation belongs to immediately in order to act adequately. One cannot merely remain indifferent to a lion running down the stairs unless he is to be considered pathological. In contrast to the case of perception, in the case of meaning that . . . there are a considerable number of reasons why not to pay too much attention to it. Undoubtedly hypotheses and theories meant non-naturally in scientific situations have contributed to improvements of our lives. Nevertheless, one thing we had to learn in school is to understand a lot of boring things. Reading as one means to gather information can be extremely boring. As opposed to understanding situations (for example 352 Claus Schlaberg (Celle) reading situations) being in perception situations is of interest with some degree of necessity. We have to walk around and look at our surroundings. But we don ’ t have to deal with theories. People who prefer to deal with theories and neglect what is going on around them need a lot of help from people who don ’ t neglect what is going on around them. (A further point is: Indication requires no confidence in a sender ’ s trustworthiness. Although there is unreliable flow of information already among non-human sign users (see Lee 2011: 118 - 119), lying does not occur until communication, non-natural meaning, occurs. As opposed to deception on the level of natural meaning, someone who puts rat poison on the table intending to make you believe that the house is rat-infested although he himself does not believe that the house is rat-infested fools you at least.) The sciences (are expected to) base their assumptions on indication. This is how they reintegrate natural meaning in non-natural meaning. The arrangements of entities in space and time they basically claim are reducible to appearance (more general: to experience) behaviour (reducible not in the sense of abbreviation). Appearance behaviour of certain kinds reversely indicates arrangements in space and time. As was exemplified by the behind-relation: Behind (y 2 , y 1 ) ⊃ [(B 1 ⊃ Above (y 2 -appearance, y 1 -appearance)) ∧ (B 2 ⊃ Below (y 2 -appearance, y 1 -appearance))]. Although the sciences certainly include communicative practices, they are not narrative in the sense in which telling stories is narrative. The subject of scientific communication are indications. A further remark concerning the distinction between the observable and the theoretical is in order here. Referring to the distinction between empirical generalization on the one hand and theory formation on the other hand, Hempel states: “ The early stages in the development of a scientific discipline usually belong to the former level, which is characterized by the search for laws (of universal or statistical form) which establish connections among the directly observable aspects of the subject matter under study. The more advanced stages belong to the second level, where research is aimed at comprehensive laws, in terms of hypothetical entities, which will account for the uniformity on the first level ” (Hempel 1965: 178). Even “ behind ” is a theoretical term, as opposed to “ above ” or “ below ” with regard to appearances. Triple ascriptions to one thing as being behind another, are surely even more theoretic. Nevertheless common expressions as used for descriptions of spatial arrangements are more theoretic than those used for appearance descriptions — which often have been neglected as “ subjective ” and still are to consider as intrinsic, for example, with regard to colours (Boghossian & Velleman 1989: 88 - 91). That y 2 is behind y 1 is indicated by and — if requested — would have to be proved by experiences like y 2 -appearances and y 1 -appearances being below or above one another in order to meet demands which admittedly are usually too natural to be mentioned explicitly and yet constitute the demand of being observable. Observable predicates in the wider sense (F) are connected to observable predicates in a narrower sense (F ‘ ) by reduction sentences such that the program of the sciences may be summarized as follows: They have to find out how F is justified to be ascribed to y itself: F ′ (y-experience) → (F y). During the process of enclosing the predicates in the parentheses they become increasingly theoretic. Finally, the sciences, fiction and religion as three major cultural practices are compared by referring to their respective truth makers. Why Observation Matters 353 Although y 2 being behind y 1 means more than that y 2 -appearances and y 1 -appearances (and relevant experiences in other than the visual mode) behave in a certain regular manner and thus a “ surplus meaning ” (see Reichenbach 1961: 51) of “ behind ” in contrast to purely summarizing a description of y 1 -and-y 2 -experiences is to assume, y 1 -and-y 2 -experiences, especially appearances, are the only that remains to make statements of y 2 being behind y 1 true. On the one hand Behind (y 2 , y 1 ) — more generally referring to predicates F that stand for spatial relations: F (y i , y j , . . .) — is indicated by B ⊃ F ‘ (y i -appearance, y j -appearance , . . . ). On the other hand B ⊃ F ′ (y i -appearance, y j -appearance , . . . ) is the only empirical evidence of F (y i , y j , . . .). And it is peculiar to the sciences that they have to account for B ⊃ F ′ (y i -appearance, y j appearance , . . . ). Although scientific statements are about y i , y j, . . . as statements from daily life are, the truth of scientific statements is decided upon the behaviour of y i -experiences, y j experiences, . . . This was meant by the brief characterization of scientific statements as communicating (meaning non-naturally) indications. How do the humanities reintegrate assertions, be they fictional or realistic, in the natural? For example, a work in the humanities may be about a fictional thing like Sherlock Holmes. It nevertheless has to base its statements about Sherlock Holmes on previous Holmesreferences. Statements about these Holmes-statements confirming Holmes-references themselves are to confirm by statements about experiences that we may have had in reading or in watching films (which, however, are not characterized as Holmes-referenceexperiences externally). The sciences and fiction have in common that they are directed to an ‘ intentional object ’ , be it real or fictitious. In order to confirm statements the sciences have to explore yexperience. In the case of humanities dealing with fictitious objects y, that which confirms statements on y are y-references. Religious statements neither have to consider y-experiences nor y-references. They simply refer to y whether they confirm their statements considering y-references from holy books or not. As a conclusion this may be illustrated by the following structuring which is intended to show what the respective practices refer to, be it as the object, be it as a truthmaker (Table 3): Table 3 object y y-experience y-references sciences + + humanities + + + fiction + + religion + The sciences deal with their respective topic y. In doing this they have to refer to yexperiences as truth makers of their statements on y. In the humanities statements on y — if fictitious — are to be made true by y-references (writings, films, . . .). In contrast to this, fictitious practices also refer to subject matters y, but in order to make their statements on y solely accomplish y-references. In religion no truth-making is at work. Certainly there are 354 Claus Schlaberg (Celle) books important to what religious people believe. But generally their references to subject matters y like God do without any truth makers. By no means y or y-references are considerable truth makers of statements on real things in the sciences.This is how one kind of cultures or family of cultures which may be called “ cultures of knowledge ” differs from other cultures. The essential role which y-experiences play in cultures of knowledge remains an important component from Russell ’ s claim: “ Every proposition which we can understand must be composed wholly of constituents with which we are acquainted ” (Russell 1989: 32). Nowadays we don ’ t expect propositions to be composed wholly of constituents with which we are acquainted anymore and prefer to acknowledge what semantic externalism has taught concerning how to treat the not-yet-discovered matters of fact. 5 Prospects One of the further steps needed to characterize ‘ cultures of knowledge ’ is to delineate how publicity, which is one main characteristic of the reasons for believing in the communicated indications in the sciences, continues with regard to what the expected reasons are in democratic decisions. On several occasions Rawls remarks that publicity belongs to the formal constraints of the social order chosen behind the veil of ignorance as a basis of being just (Rawls 1999: 15, 48, 112 - 118; the idea has occasionally been considered by Nida- Rümelin (see for example 2010) with reference to Kant ’ s “ Zum ewigen Frieden ” ). The idea that democracy is somehow connected to the sciences may be justified by the role which publicity is expected to play in both: in the sciences as a characteristic of truthmakers and in democracy as a characteristic of political reasoning which again is expected to refer to the sciences. This belongs to the well-known epistemic democracy project. Another line of thought may lead to a kind of Hegelian treatment of the Gettier problem: Since Gettier ’ s article “ Is Justified True Belief Knowledge? ” (Gettier 1963) it has been widely accepted that it is not sufficient to explicate “ knowledge that p ” as ‘ justified true belief that p ’ . A so far hardly known addition X seems to be missing. This remark is certainly not apt to contribute to the long debate on this topic but is meant to draw the reader ’ s attention a little closer to how Hegel distinguished knowledge from states of individual minds. Whereas ‘ belief that p ’ may be conceived of as an internal state or even a physiological state (which surely has been disputed by several accounts such as those concerning a Swampman ’ s mind (Davidson 1987)), ‘ knowledge that p ’ rather belongs to an ongoing process of completion which takes truthmakers of p into account as situation semantics has done ever since. Hegel claims that knowledge is real only within the sciences: “ daß das Wissen nur als Wissenschaft [. . .] wirklich ist ” (Hegel 1986: 27). Bibliography Barwise, Jon 1984: “ The Situation in Logic I ” , Center for the Study of Language and Information, Leland Stanford Junior University Black, Max (ed.) 1965: Philosophy in America, London: George Allen & Unwin LTD Boghossian, Paul A. & J. David Velleman 1989: “ Colour as a Secondary Quality ” , in: Mind, New Series, Vol. 98, No 389 ( Jan., 1989): 81 - 103 Why Observation Matters 355 Carnap, Rudolf 1936, 1937: “ Testability and Meaning. Philosophy of Science ” , vol. 3 (1936), pp. 419 - 471, and vol. 4 (1937), pp. 1 - 40. Reprinted in: Herbert Feigl & Max Brodbeck (eds.) 1953: 47 - 93 Carnap, Rudolf 1956: “ The Methodological Character of Theoretical Concepts ” , in: Minnesota Studies in the Philosophy of Science 1(1): 28 - 76. Chalmers, David 1996: The Conscious Mind. In Search of a Fundamental Theory, New York and Oxford: Oxford University Press Chisholm, Roderick M. 1957: Perceiving, Ithaca and New York: Cornell Cooper, Robin 1997: “ Austinian Propositions, Davidsonian Events and Perception Complements ” , in: Ginzburg, Khasidashvili, Vogel, Levi & Vallduvi (eds.) Proceedings of the Tbilisi Symposium on Language, Logic and Computation: selected papers, CSLI, 1998, https: / / www.researchgate.net/ publication/ 2658594_Austinian_propositions_Davidsonian_events_and_perception_complements (accessed 26. 05. 18). Davidson, Donald 1987: “ Knowing One ’ s Own Mind ” , in: Proceedings and Addresses of the American Philosophical Association 61: 441 - 58 Dennett, Daniel C. 2003: “ Who ’ s On First? Heterophenomenology Explained ” , in: Journal of Consciousness Studies, Special Issue: Trusting the Subject? (Part 1),10, No.9 - 10, October 2003: 19 - 30 Ellis, Brian 2001: Scientific Essentialism, Cambridge: Cambridge University Press Devlin, Keith 2001: “ Lecture 3: Introduction to Situation Theory ” , http: / / www.helsinki.fi/ esslli/ courses/ readers/ K1/ K1 - 3.pdf (accessed 26. 05. 18). Feigl, Herbert & Max Brodbeck (eds.) 1953: Readings in the Philosophy of Science. New York: Appleton- Century-Crofts Gettier, Edmund L. 1963: “ Is Justified True Belief Knowledge? ” , in: Analysis, Vol. 23, Issue 6, 1 June 1963: 121 - 123 Gibson, James J. 1979: The Ecological Approach to Visual Perception, Boston: Mifflin Goldman, Alvin I. 1967: “ A Causal Theory of Knowledge ” , in: The Journal of Philosophy, Vol. 64 ( June 22, 1967): 357 - 372. Goldman, Alvin I. 1976: “ Discrimination and Perceptual Knowledge ” , in: The Journal of Philosophy, Vol. 23 (1976): 271 - 291 Goldstein, E. Bruce 2002: Wahrnehmungspsychologie. Heidelberg and Berlin: Spektrum. Second edition: Transl. Gabriele Herbst and Manfred Ritter. Original: Sensation and Perception. Australia: Wadsworth-Thomson Learning. 2002. Goodman, Nelson 1968: Languages of Art. An Approach to a Theory of Symbols, Indianapolis: Hackett Publishing Company Grice, Herbert Paul 1957: “ Meaning ” , in: Philosophical Review 66: 377 - 388 Grice, Herbert Paul 1968: “ Utterer ’ s Meaning, Sentence Meaning and Word-Meaning ” , in: Foundations of Language 5: 377 - 388 Grice, Herbert Paul 1969: “ Utterer ’ s Meaning and Intentions ” , in: The Philosophical Review 78: 147 - 177 Haslanger, Sally 2000: “ Gender and Race: (What) AreThey? (What) DoWeWant Them To Be? ” , in: Noûs 34: 1 (2000): 31 - 55 Haslanger, Sally 2010: “ Language, Politics and ‘ the Folk ’ : Looking for ‘ The Meaning ’ of ‘ Race ’” , in: The Monist Vol. 93, no. 2: 169 - 187. Hegel, Georg Wilhelm Friedrich 1986: Phänomenologie des Geistes, Frankfurt am Main: Suhrkamp. New edition on the basis of the Works of 1832 - 1845 Hempel, Carl Gustav 1965: Aspects of Scientific Explanation and other Essays in the Philosophy of Science, New York: The Free Press Husserl, Edmund 1980: Vorlesungen zur Phänomenologie des inneren Zeitbewußtseins, Tübingen: Niemeyer. First edition 1928 356 Claus Schlaberg (Celle) Jorna, René, Barend van Heusden & Roland Posner (eds.) 1993: Signs, Search and Communication: Semiotic Aspects of Artificial Intelligence. Berlin and New York: de Gruyter Kripke, Saul 1980: Naming and Necessity, Oxford: Basil Blackwell (revised edition) Lee, Jacob Ian 2011: Situation Theory: A Survey. https: / / jacoblee.net/ documents/ ThesisJacobLee.pdf (accessed 26. 05. 18). Maurer, Dieter & Claudia Riboni (eds.) 2010: Bild und Bildgenese. Frankfurt a. M.: Lang 2010 McGinn, Colin 1991: “ Can We Solve the Mind-Body-Problem? ” , in: Colin McGinn 1991 Colin McGinn 1991: The Problem of Consciousness: Essays Towards a Resolution, Oxford (UK) and Cambridge (USA): Blackwell McGinn, Colin 2005: “ Bewusstsein und Raum ” , in: Thomas Metzinger (ed.) 2005. Fifth edition: 183 - 201 Meggle, Georg 1993: “ Kommunikation, Bedeutung, Implikatur - Eine Skizze ” , in: Georg Meggle (ed.) 1993: 483 - 508 Georg Meggle (ed.) 1993: Handlung, Kommunikation, Bedeutung, Frankfurt am Main: Suhrkamp Metzinger, Thomas (ed.) 2005, Bewußtsein. Beiträge aus der Gegenwartsphilosophie, Paderborn: Mentis. Fifth edition Monnoyer, Jean-Maurice (ed.) 2007, Metaphysics and Truthmakers, Frankfurt am Main: Ontos-Verlag Moore, Richard 2017: “ Gricean Communication and Cognitive Development ” , in: The Philosophical Quarterly, Vol. 67, Issue 267, 1 April 2017: 303 - 326 Nida-Rümelin, Julian 2010: “ Demokratie will Öffentlichkeit ” , in: DIE ZEIT Nr. 51/ 2010 Noë, Alva 2004: Action in Perception, Cambridge: MIT Press Posner, Roland 1993: “ Believing, Causing, Intending: The Basis for a Hierarchy of Sign Concepts in the Reconstruction of Communication ” , in: René Jorna, Barend van Heusden & Roland Posner (eds.) 1993: 215 - 270 Posner, Roland 2010: “ Die Wahrnehmung von Bildern als Zeichenprozess ” , in: Dieter Maurer & Claudia Riboni (eds.) 2010: 139 - 183 Putnam, Hilary 1975: “ The Meaning of ‘ Meaning ’” , in: Hilary Putnam 1975: 215 - 271. Putnam, Hilary 1975: Philosophical Papers, Band II, Cambridge: Harvard University Press Quine, Willard Van Orman 1961: From a Logical Point of View. Logico-Philosophical Essays, New York: Harper Torchbooks. Second edition Rawls, John 1999: A Theory of Justice, Cambridge: Harvard University Press. Second edition Reichenbach, Hans 1938: Experience and Prediction, Chicago: Phoenix Books. The University of Chicago Press. First Phoenix Edition 1961 Russell, Bertrand 1989: The Problems of Philosophy, Oxford: Oxford University Press. First edition 1912 Schellenberg, Susanna 2007: “ Action and Self-Location in Perception ” , in: Mind, Vol. 116. 463. July 2007: 603 - 631 Schellenberg, Susanna 2017: “ Perceptual Consciousness as a Mental Activity ” , in: Noûs 2017: 1 - 20. 10. 1111/ nous. 12209 Schiffer, Stephen 1972: Meaning, Oxford. Second edition 1988 Schlaberg, Claus 2017: Das Vorhandene als Gegenstand der Wissenschaften, Bonn: Bouvier Searle, John 1965: “ What Is a Speech Act? ” , in: Max Black (ed.) 1965: 221 - 240 Sellars, Wilfrid 1991: “ Empiricism and the Philosophy of Mind ” , in: Sellars 1991: 127 - 196 Sellars, Wilfrid 1991: Science, Perception and Reality, Atascadero: Riedgeview Publishing Company Smith, Barry & Jonathan Simon 2007: “ Truthmaker Explanations ” , in: Jean-Maurice Monnoyer (ed.) 2007: 79 - 98 Strawson, Peter F. 1964: “ Intention and Convention in Speech Acts ” , in: Philosophical Review 73 (1964): 439 - 460 Wittgenstein, Ludwig 1963: Tractatus Logico-philosophicus, Frankfurt am Main: Suhrkamp Zalta, Edward N. 1987: “ Erzählung als Taufe des Helden ” , in: Zeitschrift für Semiotik 9: 85 - 97 Why Observation Matters 357