Sign Levels Synopsis
by
D. S. Clarke
Email: davidsclarke@mchsi.com
Foreword
1.1.
The Scope of Semiotic
1.2.
The Language Archetype and Analogical
Inference
1.3.
Primitiveness
2.
Natsigns
2.1. Alternative
Analogical Extensions
2.2. Causal Explanations and
Logic
2.3. Philosophical Understanding
3.
Communication
3.1. Varieties of Comsigns
3.2. Comsign
Signals
3.3. Other Comsigns
4.
Sentences
4.1. Types of Sentences
4.2. Pragmatic Specialization
4.3. Reference Extension
4.4 Meaning Transfer
4.5. Natural Languages and
Semantic Fields
5.
Natural Language Discourse
5.1. The Discourse Level
5.2. Quotation
5.3. Uses of Discourse
6.
Specialized Discourse and Philosophy
6.1. Writing and Specialization
6.2. Discourse Frameworks and
Category Terms
6.3. Philosophical Discourse
7.
Metaphysics
7.1. Metaphysics and Reductions
7.2. Scientific Realism
7.3. Vacuous Generality
7.4. The Role of Contemporary
Philosophy
Foreword
The following
is a brief summary in Tractatus style of my earlier Sign
Levels (paperback, 2004). It
is in the form of a sequence of claims – some very obvious, others not
so obvious but generally accepted within the philosophical community, still
others highly controversial. These have been stripped of all references to
sources in the book version and all but a few of supporting arguments. This online summary with its hyperlinks between
parts may prove useful as a compilation of some conclusions arrived at in
the past century and as a way of seeing connections between them. Its brevity and links are intended to help make
these connections more obvious.
Investigations
of language use by what is labeled “analytic philosophy” have transformed
our conception of philosophy in a way similar to an earlier transformation
by Descartes, Locke, and Kant. Unlike the earlier transformation, however,
recent philosophers accomplished theirs by a series of separate insightful
contributions, often without connecting threads. Philosophy has historically played the role
of enabling us to see how the diverse elements of our experience, in Rorty’s words, “hang together,” the role of
satisfying our urge for some comprehensive outlook.
If philosophy meeting past standards for rational criticism fails to
fulfill this role, allies of irrationalist religious
sentiment tend to fill the gap, and our culture becomes the poorer. I am under no illusions that the framework outlined
here provides the best means of fulfilling this role. There may be good reasons for rejecting some
of its central claims and the inferences made from them. I only ask that readers attempt to replace these
claims with others on more secure ground, and join in the task of providing
a more adequate comprehensive framework.
___________
1.1.1. Semiotic is the philosophical study of signs, including interpreted natural events (as when a deer interprets an odor as a sign of a predator), signals (warning calls by animal sentries), isolated sentences, and sentences combined to form blocks of discourse. Its origins can be traced back to Greek and Medieval philosophy. Its modern development is due principally to Charles Peirce and Charles Morris. Signs can be classified into different levels, depending on the stage of evolution in which they were introduced and the functions they perform when interpreted and used.
1.1.1.1. There are no nontrivial common features within the wide variety of signs. There is thus no informative definition of a sign that abstracts its “essence.”
1.1.1.2. We are instead restricted to contrasting and comparing features present at one level of signs with those at other levels. To specify the relevant similarities and differences is the special province of semiotic as the central branch of philosophy.
1.1.1.3. Semiotic is thus a purely descriptive and classificatory discipline, relying on judgments of what is relevantly similar or different. It provides an understanding, not an explanation, of relationships between what is described and classified.
1.1.2. The study of such comparative features within semiotic is distinguished from empirical investigations of animal learning by experimental psychology, ethology’s studies of animal communication systems, psycholinguistics, and linguistics.
1.1.2.1. These sciences investigate contingent features of signs, features that might have been otherwise had evolution taken a different course. In contrast, semiotic attempts to isolate necessary features, features that must be present at a given level of sign if it is to perform its characteristic functions. For semiotic, the relevantly similar and different at different levels is determined by what is necessary.
1.1.2.2. The empirical sciences, along with our intuitions as language users, provide the observational basis for conclusions reached within semiotic. These conclusions are thus of the form of a posteriori necessary propositions. Some of the conclusions that follow are based on behavioral and structural descriptions. Others are descriptions derived from our shared linguistic intuitions as users of language.
1.1.2.3. We cannot claim truth-guaranteeing certainty for these conclusions. The evidence provided by the sciences and our own intuitions may be mistaken, as may be our inferences from them.
1.1.2.4. Nor is there a linear relation between a selected number of propositions that constitute a starting point and others derived from them.
1.1.2.5. Conceding the fallibility of propositions about the necessary
and denying linear relations between them constitutes an important departure
from what was assumed in the systems of Descartes and Locke and those adhering
to the methodological ideals of these founders of modern philosophy.
1.2.
The Language Archetype and Analogical Inference
1.2.1. Nonlinearity follows from the fact that the starting point for semiotic’s investigations is the language we use in everyday discourse. This is what we are most familiar with as the common currency of human communication – “first in the order of knowledge,” in Aristotle’s words. This language constitutes an intermediary level between signs interpreted by infrahuman species and specialized forms of discourse used within social institutions.
1.2.1.1. Our status as participants in the communicative process provides us with linguistic intuitions of how words, sentences, and discourses function. It is relative to these intuitions that we test descriptions of language interpretation and use.
1.2.2. Each of us is a participant in a relatively small number of speech communities. We must attempt to isolate certain features of those languages with which we are familiar, and project them as logical features holding of any possible language. Among these are the distinction between the subject(s) and predicate of a sentence and the contrast between descriptive, prescriptive, and expressive sentences. Such projection is fallible.
1.2.3. Investigation of logical features is initially directed towards a natural language as our language archetype. This provides paradigm descriptive, prescriptive, and expressive forms of sentences in terms of which we understand the central logical concepts of meaning and reference. These sentences become the basis for comparisons and contrasts to both the relatively primitive signs interpreted by infrahuman species and advanced specialized forms of discourse.
1.2.3.1. These comparisons and contrasts are made through analogical extensions of sentences such as ‘The book is red’ and ‘Pick up the book’ with subject-predicate structure.
1.2.3.2. For signs interpreted by evolutionary predecessors, the basis for extension is provided by analogies derived from comparisons between animal and human physiological structures and behavior.
1.2.3.3. The materials for extension to specialized discourses are provided by the linguistic intuitions of those with backgrounds within social institutions employing a form of discourse and aptitudes for using it. Because of differences of background and aptitude, this extension leads to specialization within philosophy itself.
1.2.3.4.
These extensions from the language archetype provide the basis for distinguishing
necessary features of sign use and interpretation from contingent features
that might have been otherwise had evolution happened to have taken a different
course.
1.3.1. Signs are ordered into levels with respect to their primitiveness, as determined by their evolutionary priority, the degree to which they distinguish pragmatic functions, and their referential scope. These ordering criteria usually identify the same types of signs, but in some cases may not.
1.3.1.1. Signs at the most primitive level constitute a base from which an increasing number of features and functions are added at more advanced levels.
1.3.1.2. Signs at the most primitive level cannot be labeled as the “simples” of logical analysis. Simplicity must be judged relative to some respect by which it is attributed. In so far that they combine pragmatic functions (see 1.3.3 below) primitive signs are complex.
1.3.1.3. Descriptions of primitive signs are generated by comparisons and contrasts to sentence paradigms and by the analogical extension of logical terms such as ‘meaning’ and ‘reference’ and of psychological terms such as ‘belief’, ‘judgment’, and ‘decision’. These terms are applied to the persons with whom we communicate and the expressions they use.
1.3.2. Evolutionary Priority. By the standard of evolutionary priority, signs interpreted in associative learning are more primitive than communicative signals, signals more primitive than sentences, and sentences more primitive than discourses formed by combinations of sentences. Learning is exhibited in single-celled organisms such as amoeba, and these appeared earlier in evolution than organisms employing systems of communication.
1.3.2.1. Signs interpreted in associative learning are called natsigns. These are distinguished from the “natural signs” of Greek and Medieval philosophy – smoke as a sign of fire, yellow complexion as a sign of jaundice. Interpretation of such signs presupposes the use of a discursive inference from an observed effect to its hidden cause. This interpretation thus requires a capacity present only in relatively late stages of human evolution.
1.3.2.1.1. Events evoking reflex responses are not natsigns. That an event is described as a sign presupposes that there is some degree of spontaneity of response to it, however slight this may be. To qualify as a sign an event must be an object of interpretation.
1.3.2.2. Natsigns are distinguished from comsigns – signs used by species with the capacity for communication. Comsigns include animal signals, iconic representations with similarity to what they stand for (gestures, drawings, etc.), and sentences and discourses used in human communication. Comsigns are both interpreted and used to communicate.
1.3.2.2.1. There must also be some element of spontaneity of response by a communicator and of response by an interpreter for an event to be described as a comsign. Events that trigger sequences of coordinated reflex behavior do not constitute communication.
1.3.3.
Combination of Pragmatic Functions. At each sign level there are three basic
modes of interpretation: cognitive,
dynamic, and emotional. Sign levels can
be ordered with respect to the degree with which these three modes are combined,
with signs at more advanced levels exhibits exhibiting progressively greater
degrees of specialization of functions.
1.3.3.1.
The interpretation of natsigns combines all three
modes. An odor warning a deer of an approaching predator
is cognitively associated with an anticipated sight of this predator. It is also dynamically associated with the response
of fleeing, and evokes the emotional response of fear. It is thus classified as primitive with respect
to both evolutionary priority and function combination.
1.3.3.1.1.
Whether modes of interpretation can themselves be ordered with respect to
evolutionary priority is a topic for speculation. Panpsychists argue
that emotional interpretation in the form of what they describe as “feeling”
is prior to dynamic interpretation, which is in turn prior to cognitive interpretation.
The attribution of such pure feelings enables panpsychists
to attribute mentality to unified natural bodies incapable of the combined
cognitive and dynamic interpretation characteristic of associative learning.
1.3.3.2.
A comsign animal signal such as a warning cry also combines
the three principal modes. The cry
may be cognitively interpreted as a signal of a predator, but also be associated
with the response of fleeing, and arouse fear.
In this respect natsigns and this type of
signal are equally primitive. This
illustrates how evolutionary priority and combination of pragmatic functions
can produce inconsistent orderings.
1.3.3.3.
Though signals typically combine all three basic pragmatic functions, at the
comsign level some specialization of function does become
possible. Commands to dogs such as
‘Fetch’ and ‘Sit’ are purely prescriptive.
Single-word sentences such as ‘Tree’ or ‘Green’
can be used in a purely descriptive way, and ‘Ouch!’ is typically
expressive.
1.3.3.4. Greater specialization occurs at the more advanced level of sentences with subject-predicate structure, with specialization of function accomplished through the distinction between indicative, imperative, and optative moods. But combined functions can occur at this level, for a sentence such as ‘Your house is on fire’ informs, but it also implicitly prescribes actions such as fleeing or calling the fire department, and can express and evoke alarm.
1.3.3.5.
The greatest specialization occurs at the level of mathematical axioms and
theorems and in descriptions of measurable quantities.
1.3.4. Referential Scope. The standard for ordering with respect to primitiveness is provided by the degree to which signs at a certain level extend reference. The term ‘reference’ is a logical term derived from descriptive and prescriptive sentence paradigms. A descriptive paradigm might be a sentence such as ‘The book is red’ whose subject ‘the book’ refers to an object to be identified. The sentence’s predicate ‘red’ has meaning for a native speaker of English who hears the sentence on a given occasion, but not reference. Similarly, the imperative ‘Pick up the book’ has as its subject ‘the book’ referring to the book upon which the action of picking up is to be performed. This type of action is the meaning of its predicate ‘pick up’.
1.3.4.1. We distinguish between a sign type and a sign token as particular sensed events. Sign tokens have significance as instances of a type. Their reference is derived from the particular circumstances in which they occur.
1.3.4.2. The referent occasion of a sign token is that occasion at which its interpreter expects the type of event signified by the sign to occur.
1.3.4.3. The referent occasion of a natsign token is contiguous in time and place to this token. As an example, a rat is observed to associate the sound of a bell with an electric shock only if the shock follows the bell within a restricted interval of time. Beyond this interval, learning does not occur. This association is described in psychological terms as the expectation of the shock as the bell’s significance on hearing an instance of a type of bell sound. If the shock does follow, there is recognition of what was expected.
1.3.4.4. Contiguity of referent occasion marks natsigns as primitive relative to comsign signals and sentences.
1.3.4.4.1. Comsigns are more advanced than natsigns because of reference extension. This can be accomplished by gestures and the signaler’s spatial orientation. An animal sentry can utter a warning cry and indicate by head direction a general region at which a predator is to be expected. This has the effect of extending the referent occasion to what is proximate, though still restricted to the immediate environment of signaler and audience.
1.3.4.4.2.
Subject terms of sentences provide what is potentially an indefinite extension
of the referent occasion. The subject
of the sentence ‘The largest oil rig in
1.3.5. Iconic Representations. Iconic representations or icons are signs that signify by virtue of a similarity between a given sign and what it purports to represent.
1.3.5.1. Icons constitute a fundamentally different and often more economical alternative to signs that acquire significance through associations between sign and what is signified. (“A picture is worth a thousand words.”)
1.3.5.2. Icons occur at all levels of signs.
1.3.5.2.1. At the level of natsigns, icons seem to occur in the form of shadows, as when the shadow of a hawk alerts a squirrel to the hawk’s presence and warns it to flee. Also, the sight of an object may be a natural icon of its felt touch. Animals seem to have the capacity for forming internalized iconic representations in the form of mental maps orienting them to their environments.
1.3.5.2.2. At the comsign signal level, a tribesman’s undulating hand motion and sssing sound might be used to represent a snake for his audience (Jonathan Bennett’s example). The referent occasion would be some place indicated perhaps by the tribesman’s spatial orientation and pointing gesture at a time simultaneous with the hand motion and sound.
1.3.5.2.3. At the linguistic level of sentences, words can provide means of reference when occurring in such forms as city names on maps, labels of rooms on construction blueprints, and captions on paintings. Such combinations of the linguistic and iconic are hybrid icons. Pure icons such as the tribesman’s hand motion do not occur with linguistic accompaniments providing reference.
1.3.5.3.
Specialization of pragmatic function is less pronounced within advanced levels
of icons than for signs whose significance is derived from learned associations,
and in this respect at least both pure and hybrid icons are more primitive
than linguistic signs. A map functions
to both cognitively represent and prescriptively provide directions, though
a blueprint only has the prescriptive function of directing construction. Drawings may both represent and evoke strong
emotional associations.
2.1.
Alternative Analogical Extensions
2.1.1. Two Alternative Descriptions. Natural events with significance for their interpreters are described in two very different ways. First, internal descriptions describe them as signs by means of analogical extensions from the language archetype. These descriptions use logical and psychological terminology derived from descriptions of ourselves and other language users. External descriptions describe events in a second way as physical events causing changes in their interpreter’s physiological processes, nervous system structure, and behavior. Attempts are then made by means of reductive definitions of logical terms to analogically extend these external descriptions to language interpretation and use.
2.1.2.
Analogical Extensions from the Language Archetype.
The logical terms ‘meaning’ and ‘reference’
are derived from questions such as “What did he mean when he said he
would return soon?” (asking for a paraphrase
of what the person said) and “Who was he referring to when he used the
name ‘Alexander’?” (asking whether
it was Alexander the Great or the early 20th century British philosopher).
These questions indicate that the terms are originally applied to linguistic
expressions as used in certain contexts. Without reference to a person’s
intentions, we also ask “What does the word ‘libertarian’ mean?”
and “What do ‘Alexander the Great’ and ‘duckbill
platypus’ refer to?”
2.1.2.1. Because of the inseparability of ‘meaning’ and language use, the more general term ‘significance’ is more appropriate for extension to infrahuman sign use. The term ‘reference’ is retained for extension for the lack of a suitable replacement.
2.1.2.2. The basis for the extension of both ‘significance’ and ‘reference’ is their association with psychological attributions of expectation and recognition to members of infrahuman species. If we can say that a dog expects his master on hearing the sound of a car, then we can conclude that the sound has significance of him. If we can say that the dog recognizes what he has expected, then the sound refers to that time and place at which recognition occurs (cf. 1.3.4.2, 1.3.4.3). These associations create the possibility of extending these logical terms from the language archetype.
2.1.2.3. The psychological terms ‘belief’ and ‘judgment’ are linked to the use of linguistic expressions, though not necessarily so. ‘Belief’ is linked when we ask whether someone believes what he just said. ‘Judgment’ is linked when we assert that a person judged what she heard someone say to be true. But at other times we simply ascribe beliefs and judgments on the basis of behavior, as when we observe a person warily skating at the perimeter of a lake and conclude she believes or has judged the ice to be thin.
2.1.2.4. Because it can be separated from language use, ‘expectation’ (though not the language-specific ‘belief’) seems appropriate for extension to infrahuman species. On the basis of behavior observed to be similar to that used to attribute beliefs we conclude that a dog expects his master on hearing the car. We do this without sharing a language with the dog.
2.1.2.5. Similar considerations apply to ‘recognition’ (but not ‘judgment’), which is applied when we say the dog recognizes his master when she arrives.
2.1.2.6. Our describing members of infrahuman species as expecting and recognizing after perceiving certain natural events provides the basis for describing these events as signs and extending to them logical terminology.
2.1.2.6.1. An event recognized as an instance of what was expected is a significate occurrence at the sign’s referent occasion. The type of event of which this particular event is an instance is the sign’s significance. Recognition that what was expected did not occur is recognition of a significate non-occurrence.
2.1.2.7. Descriptive, prescriptive, and emotive sentence paradigms provide the basis for extension of logical and psychological terminology to natural events. The interpretation of natsigns combines cognitive, dynamic, and emotive modes of interpretation that are more clearly distinguished at the linguistic level by sentences of different moods (cf. 1.3.3.1).
2.1.3.
Extensions by Reductionist Strategies. A
second type of analogical extension is provided by reductionist strategies for defining logical terminology in
terms of external descriptions of natural events as elements of causal sequences
(cf. 2.1.1). The
reduction of a logical term is by means of a definition of the term in terms
of correlations between publicly observable events. The two principal reductionist
programs are those of the historically earlier behaviorism of experimental
psychology and of more recent functionalism using models derived from cognitive
science. Their aim has been to include
language interpretation and use within the scope of the natural sciences.
2.1.3.1. Experimental psychology
applied to animal learning defines signs and their significance and reference
in terms of correlations between stimuli and behavioral responses. The organism’s
physiological structure and the causal mechanisms mediating between stimuli
and responses were regarded as a “black box” about which we are
ignorant.
2.1.3.2. Cognitive science employs computer models that artificially simulate observed correlations between stimuli (inputs) and responses (outputs). Functional relationships between inputs and outputs, together with observations of physiological structure and processes, are then used to construct hypotheses about physiological structure and mediating causal mechanisms in natural organisms.
2.1.3.3. For both behaviorism and functionalism, sign, significance, and reference are defined as terms in causal sequences. For behaviorism, the sign is defined as either a causal stimulus evoking a response (classical conditioned reflex models of learning) or as a type of behavioral response (Skinner’s models of instrumental learning). For conditioned reflex models, the sign is typically defined in terms of the correlation between a conditioned stimulus (such as the sound of a bell) paired with another stimulus, the unconditioned stimulus, evoking a reflex response (such as an electric shock evoking a paw raising response by a dog). If the paw raising response is transferred from the shock to the bell tone in the absence of the shock, the bell tone as conditioned stimulus is defined as the sign; the shock as unconditioned stimulus is defined to be the sign’s significance. For functionalism, the sign is typically defined as a mental representation, an internal physiological state of the organism identified with a mental image. This representation is an effect of some environmental stimulus. This stimulus (though more often the type of object to which it is correlated) is defined as the significance of the representation. Its reference is defined as the particular object causing the mental representation. In some versions, there is a requirement that goal-promoting responses to the environmental stimulus be observed.
2.1.3.4. These definitions are in the form of identity sentences such as ‘A sign is an conditioned stimulus paired with an unconditioned stimulus’ or ‘A sign is a mental representation in the form of a physiological state’. The definitions are intended to reduce the second term of the identity to the first.
2.1.3.5.
These definitions of ‘sign’, ‘significance’, and ‘reference’
are then extended to the words and sentences of natural languages as used
and interpreted by humans.
2.1.3.6. These reductionist
strategies fail to solve two fundamental problems. The first is that of accounting
for negation and falsity, and is internal to the various definitions that
have been proposed. The second is that
of scope, the problem of extending definitions derived from causal analyses
of relatively simple forms of animal learning to the much more complex human
use and interpretation of language.
2.1.3.6.1. Animal learning occurs through stimulus generalization and discrimination. Generalization is illustrated by the dog generalizing from a particular bell tone (defined as the sign token) followed by a shock by transferring his paw withdrawal reflex response from the shock to any bell tone (the sign type). The behavioral test of discrimination is whether the dog learns to respond by paw withdrawal to certain bell tones but not to others. The capacity for discrimination is acquired, however, when the shock is absent, when the bell sounds and the shock does not follow. This absence of shock is itself not a stimulus evoking a response. Similar problems of defining sign and significance for absent events arise for models of instrumental learning where there is discrimination.
2.1.3.6.2. Sign discrimination is within quality dimensions of sounds, colors, tastes, smells, and tactile sensations. The informational content of a natsign is a function of the qualities excluded within a given dimension as a consequence of prior learning. One dog may have learned to discriminate high from low pitch bell tones. A second may have learned to discriminate three pitches corresponding to C, E, and G on a 7-tone scale. A given C pitch tone will then have higher informational content for the second dog than for the first, since it excludes more within its dimension.
2.1.3.6.3. One example frequently used for functionalist reductions is that of a type of stimulus correlated to the presence of a fly (the significance) that evokes a snapping response by a toad. Mediating between stimulus and response is a mental representation (the sign) as the image of a dark spot. If the snapping response is reflex (as it is for the toad), then the internal representation cannot be described as a sign if there is to be a comparison to the use and interpretation of language. Where such responses are learned through discimination, we encounter the same problem of absence that infects behavioral reductions. If the toad were to discriminate between a fly and a BB shot dangled in front of it, the toad would have to recognize that the BB shot is not what its mental image signifies. Such recognition would make possible the withholding of the snapping response.
2.1.3.6.4.
The scope problem for both behavioral and functionalist models arises because
of the indefinite variety of ways that humans respond to events in their natural
environments. When I see an apple,
I may eat it, but then again I may not. When
I see it, I may think of an apple and form an ‘Apple’ mental representation. But I may also think of redness or of the orchard
from which the apple came, and make a variety of other associations.
Even if there were to be observed invariant correlations for relatively
stereotyped animal behavior (and at this level the correlations are typically
described by statistical generalizations), reductive definitions based on
them seem to have no application to the use and interpretation of language
from which logical terminology is derived.
2.2.
Causal Explanations and Logic
2.2.1. Causal descriptions are of the form ‘X causes Y’, where X and Y can be either dateable events or persistent states of affairs. They must be either actually or potentially publicly observable events or states. A paradigm causal descriptions might be ‘Last night’s rain caused the streets to be wet’, ‘Clouds and atmospheric conditions cause rain’, ‘His allergy and pollen in the air caused his running nose’, and ‘The sounding of the bell caused the dog to raise its paw’.
2.2.1.1.
We typically attend to only those events and states of affairs of immediate
interest, ignoring more complex factors such as events or states of affairs
intervening between what we single out as cause and effect. Our descriptions of what is singled out vary
in degrees of specificity, again depending on interests, as when we describe
an allergy as causing a reaction and ignore environmental conditions as a
contributing cause.
2.2.1.2. The requirement of public observability of cause and effect has implications for descriptions of causal relations between publicly observable stimuli (e.g., the stimulus of a hot poker in contact with my hand) and what is described by first-person avowals such as ‘I feel a pain’.
2.2.1.2.1. Only in so far as the pain can be identified with some physical event or state of affairs can it be regarded as a term of a causal relation.
2.2.1.2.2. What is referred to as “agent causation” is described in ways very different from empirical causal descriptions. ‘Who murdered Smith?’ seems to be of the same form as ‘What caused the streets to be wet?’. But the purpose for which the question about the murder is raised is that of assigning responsibility for it, not that of providing a causal explanation, and it is only misleading to speak of both the murderer and rain as causes.
2.2.1.2.3. Similar confusions occur when we explain a person’s action in terms of the person’s beliefs and desires. To attribute a belief to someone, say the belief that it will rain (‘She brought her umbrella because she believed it would rain’) is to suggest that the belief may be mistaken. Correction of the belief may alter the behavior of both the person and others sharing the same belief. Similarly, to attribute a desire (as in ‘He ate chocolates because he wanted to gain weight’) suggests that maybe the person should not have desired to gain weight. Attributions of belief and desire occur in this way within social contexts that invite interactive criticism and correction. Such contexts are typically absent when we make empirical causal descriptions (though such descriptions may themselves be mistaken).
2.2.1.3. The requirement of public observability of cause and effect also has consequences for attempts to describe signs, significance, and reference as causes and effects.
2.2.1.3.1. For me to describe what I hear as a sentence is not to describe what I hear as either a type of auditory or visual stimulus or the neurological effects or such stimuli. To ask what a person said is to ask someone to be a relay of what was said. The question arises in a communicative context in which an attempt is being made to understand what is being relayed. The purpose of the question is very different from that of describing stimuli and neurological effects in order to provide a causal explanation.
2.2.1.3.2. To describe a natural event such as the sounding of a bell as a sign is to compare it to a sentence, since ‘sign’ is a generic term for whatever has significance and reference for interpreters. The purpose for describing a bell tone as a conditioned stimulus with a position in a causal sequence is thus very different from that of labeling what is heard as a sign.
2.2.1.3.3. To describe a shock as an unconditioned stimulus makes it impossible to describe it also as the significance of a sign. There are causal relations between paired conditioned and unconditioned stimuli (the bell and the shock) and between this pairing and a reflex response transferred from one to the other, but the logical term ‘significance’ as analogically extended from language has no application to what occurs in such causal sequences.
2.2.1.3.4. There are also causal relations between an object (a fly), a visual stimulus, a neurological event in a toad’s cortex, and a snapping response. But these elements in a causal sequence cannot be identified with what logic describes as sign, significance, and reference. They cannot be identified because the purposes of a causal explanation are different for those of a description of what has significance and reference. Identification can only occur within types of discourse used for a common purpose.
2.2.1.4.
Reductive models derived from animal learning cannot there be justifiably
extended to language use and interpretation.
This is not to deny, of course, that causal explanations of neurological
events and states and of behavior, to the extent they are possible for both
humans and infrahuman species, are of great value.
2.3.
Philosophical Understanding
2.3.1. Historically, philosophy has used analogical reasoning (often not made explicit) in order to describe relationships between between natural languages shared by all and specialized activities and communications within social institutions. This project has been humanistic, restricted to the human species. The goal has been an understanding of shared/specialized relationships, not an explanation. This understanding relies on results obtained within the empirical sciences.
2.3.2. The use of analogical extensions of logical conclusions reached about everyday language use to sign interpretation and use within infrahuman species is anticipated by Epicurus, Sextus Empiricus, Augustine, and the British empiricists. An understanding of relationships between the human and infrahuman has not been a central focus of philosophy since Descartes and Locke, however.
2.3.3.
The aim of reductionist projects is to provide a
basis for causal explanations of what is observed.
This goal of causally explaining is distinct from that of philosophical
understanding, but becomes confused with it when logical and psychological
terms are imported.
3.1.1. A comsign is any sign with significance and reference for an interpreter that is produced by a communicator with communicative intent.
3.1.2. For an event to be classified as a comsign requires it to fulfill three basic conditions.
3.1.2.1. First, an event such as a warning cry or verbal utterance must be produced with the intent to produce some effect in an audience. The cry or utterance may be a reflex response to some environmental stimulus, similar in this respect to the bristling of the hairs on the back of a dog (cf. 1.3.2.2.1). Such an event may be a sign for some interpreter, possibly a natsign, but in the absence of communicative intent is not a comsign. This condition is in itself sufficient to distinguish comsigns from natsigns.
3.1.2.1.1. The distinction between what is reflex (stereotyped) and what is intended (spontaneous) will often depend on how specific is the description of a certain type of behavior. A dog may invariably snarl and bare its teeth in a certain type of social situation, but the intensity of the snarl and the degree of teeth baring may vary in unpredictable ways. Such variation would make it possible to attribute communicative intentions in a way impossible for the bristling.
3.1.2.2. The second condition is that the communicator intend to produce an effect in an audience by virtue of the audience’s recognition of a communicative intent. This condition has the effect of distinguishing communication from manipulation. An experimenter might apply a shock to a dog’s paw in order to elicit a withdrawal response, but this fails to be communication.
3.1.2.3. Finally, the effect produced in the audience must occur as the result of an actual recognition by the audience of the communicative intent. A speaker may address someone in French, assuming that her audience understands the language. But if the audience does not, if there is no “uptake” to what she says, genuine communication is absent. This uptake usually takes the form of understanding what was said. Where commands are issued, it takes the form of both understanding and obeying. A degenerate comsign is one produced with communicative intent, is not an instance of manipulation, but fails to satisfy this last condition. The defect here is not with the intent, but instead with the effect actually produced.
3.1.2.4. Conventional signs. Uptake can be achieved in two different ways, and these ways distinguish conventional and nonconventional comsigns. For conventional signs, an event is discriminated as a token of a sign type, and this type has a significance that constitutes a norm or rule to be followed by both communicator and audience.
3.1.2.4.1. With conventionality of sign is introduced the distinction between truth and falsity, a distinction absent at the natsign level. At the natsign level, if the sign’s interpreter recognizes at the referent occasion that what the sign signifies does not occur (a bell is not followed by a shock), then this tone of bell is discriminated from those for which the shock follows. Recognition of a non-occurrence is followed by a change in the sign’s significance. For a conventional sign, in contrast, significance is relatively constant, and recognition of a non-occurrence is normally the occasion for a judgment of falsity.
3.1.2.4.2. Despite this conservativeness for conventional signs, changes of significance can occur with changes of community practice. If a prestigious member of a community starts applying ‘Red’ to what had been distinguished within the community by the adjective ‘Orange’ and others pick up this practice, the significance of ‘Red’ may be widened and a new convention established by the change of practice.
3.1.2.5.
Nonconventional comsigns. These have significance for an interpreter,
not by virtue of rules or norms applying to a sign type, but instead only
by recognition of the communicator’s intentions on a particular occasion.
They may also be icons recognized by the interpreter as similar to
what they are intended to represent (cf. 1.3.5).
3.2.1. Comsign signals are comsigns without the internal subject-predicate structure characteristic of sentences. Reference is supplied, not by a part of the sign, but by supplementing devices (cf. 1.3.4.4.1).
3.2.1.1. Comsign signals seem to be present in species lower down in the phylogenetic scale than primates. They may include bird alarm, territorial, and mating calls, and possibly insect sounds, tactile contacts, and pheromones. Because all such events can be described with varying degrees of specificity, it is arbitrary to use relatively gross descriptions of what is labeled stereotyped behavior to deny communicative intent (cf. 3.1.2.1.1).
3.2.1.2. All comsign signals, whether used at human or infrahuman levels, usually combine to some degree pragmatic functions (cf. 1.3.2.2). At this level, however, there is an important distinction between content elements with descriptive/prescriptive/expressive functions and communication establishing elements used to establish contact and perhaps some social hierarchy between communicator and audience. The greeting ‘Hello’ is used to open communication, not to convey information, prescribe, or express. Within animal communities there are contact calls (grunts for vervet monkeys) enabling the distinction between friend and foe. Degree of elevation of communicator (raising on hind legs, cowering) may be used to establish dominance or submission.
3.2.1.2.1. Some contact calls are signature calls, as illustrated by calls between mating bird pairs. These identify a particular individual, not simply a type of individual. The interpretation of such signs requires the capacity for recognizing quantitative identity, as contrasted to the capacity for recognizing a significate occurrence as qualitatively identical with what had been previously experienced. Whether recognition of quantitative identity is present also for other types of comsigns and at the natsign level is unclear.
3.2.2. Conventional signals. Comsign signals may be either conventional or nonconventional. Conventional signals include single-word sentences such as the adjective ‘Red’ and noun ‘Dog’. They seem to include also animal warning calls such as the distinct vervet monkey alarm calls warning of eagles, leopards, and pythons, since significate non-occurrences have been observed to bring about a change in an assessment of communicator reliability. There may also be rule-governed signals among other infrahuman species.
3.2.2.1. Reference for comsign signals is typically supplied by pointing gestures, spatial orientation of communicator, and environmental context.
3.2.3. Iconic signals. Iconic signals signify by virtue of some similarity between the signal and what it is intended to represent. Successful communication requires the interpreter to recognize this intended similarity.
3.2.3.1. A tribesman guide may indicate to a tourist the presence of a snake by means of an undulating gesture, and perhaps accompany this with a pointing gesture. Other examples of iconic signals include drawings, a parking attendant’s spacing between hands to indicate distance to curb, etc.
3.2.3.1.1. Such comsign signals are distinguished from hybrid icons combining conventionalized means of reference with an icon. These emerge at the sentence level of signs. Examples of hybrid icons would be drawings with name captions underneath indicating referents and maps with names of cities and towns.
3.2.3.1.2. Gestures, drawings, and other forms of comsign signals may be conventionalized as aids to recognizing communicator intentions, as illustrated by cartoon caricatures of the famous. There are degrees of such conventionalization.
3.2.3.2.
Iconic signals are used by infrahuman species, as for variation in intensity
of alarm calls to represent proximity of predator.
3.3.1. Besides signals, both conventional and nonconventional, comsigns include all signs at progressively more advanced levels. These include isolated sentences, sentences combined to form natural discourse blocks, and specialized forms of discourse used within social institutions. The special features of these other comsigns are the topics of later sections.
3.3.2. Shared Features.
Despite very basic differences, these other comsigns
share with signals three basic characteristics: (1) communicative intent and
uptake; (2) a combination of both invariance and change in significance through
recognition of significate non-occurrences; (3)
a distinction between communication establishing and descriptive/prescriptive/expressive
elements; and (4) inclusion of iconic representations.
3.3.3. Applying the Necessary/Contingent Distinction. Are all of these shared features necessary
features of comsigns? Is successful communication possible without
them? Or are only some of the features necessary for communication?
3.3.3.1. Communicative intent and uptake enable the distinctions between natsign and comsign levels and between dynamic interpretation and reflex response to a signal. They are thus necessary features of comsigns.
3.3.3.2. Both the possibility of stability and change of significance would seem to be necessary for communication, and thus not a contingent fact of evolution on this planet. It seems impossible to conceive of communication without learning, and a combination of stability and change is necessary for learning.
3.3.3.3. The distinction between establishing a channel of communication and conveying information through that channel does seem necessary to the comsign level. This necessity is derived from the impossibility of conceiving communication without some means of establishing it.
3.3.3.4. We can conceive a system of communication utilizing
only icons. We can conceive another
that utilizes only signs with conventional significance. The combination we actually observe is undoubtedly
more efficient than either system in isolation, but this observed combination
is a contingent feature of comsigns. However, comsign signals
with conventional significance would seem to be a necessary condition for
the extension of reference that characterizes the sentence level of signs.
4.1.1. Paradigm Sentences. A sentence is a conventional comsign combining a predicate term having significance with one or more subject terms that both signify and refer. The significance of a term is at this level called its meaning (cf. 2.1.2) . In this section and the next we consider only sentences used in everyday conversational speech. In Section 6 written sentences are introduced.
4.1.1.1. The sentence level introduces the categories of object, quality, relation, and quantity. The role of the predicate of a sentence is to either attribute a quality to what is referred to by its single subject or express a relation between the referents of two or more subjects. These referents are termed objects as what can be reidentified on different occasions. The reference of its subject or subjects is to either one or a plurality of objects. Reference of subjects thus presupposes the distinction between one and many and the category of quantity.
4.1.1.2. Examples of singular sentences are ‘John is tired’ (with proper noun ‘John’ as subject and ‘is tired’ the monadic predicate expressing an attribute), ‘John hit Bill’ (with ‘hit’ expressing a relation between the two named referents), and ‘The tallest building in town will be torn down’ (with the definite description ‘the tallest building in town’ as subject and ‘will be torn down’ as predicate).
4.1.1.3. Examples of general sentences with subjects used to refer to a plurality of objects include ‘All the books on the table are red’, ‘Some swans are white’, and ‘Three magazines were placed on top of the one table’, where ‘all’, ‘some’, ‘three’, and ‘one’ are the quantifiers functioning to indicate how many objects are being referred to by the subjects ‘the books on the table’, ‘swans’, ‘magazines’ and ‘table’.
4.1.1.4. Paradigm sentences have sortal terms as subjects. Sortals include general count nouns such as ‘table’, singular terms such as the name ‘John’ and the noun phrase ‘the book(s) on the table’. Sortals can also occur in the predicate position, as for ‘Igor is a dog’. In this position, however, they express complexes of qualities or attributes, and no longer occur with quantifiers to measure quantity.
4.1.1.5. Attributive terms – adjectives such as ‘red’ or verb phrases such as ‘placed on top of’ – occur only in predicate position to signify attributes (qualities) or relations. We cannot ask how many reds there are, and hence red as a quality cannot be quantified. Lacking quantification, attributives cannot occur as genuine subjects.
4.1.2. Degenerate Sentences. These are sentences with terms in the grammatical subject position of a sentence, but which do not indicate a quantity and lack the potential for extending reference.
4.1.2.1. One form of degenerate sentence is a feature-placing sentence. Examples include ‘It is raining’, ‘There is water here’, and ‘Snow is falling’ in which occur the mass nouns ‘rain’, ‘water’, and ‘snow’. In the first two sentences the words ‘it’ and ‘there’ are grammatical fillers for subjects, but without the function of referring to a reidentifiable object. Quantification also fails to apply to the mass noun ‘snow’. We can count the number lumps or pieces of snow in a given area, but the question ‘How many snows are there?’ is grammatically ill-formed, and admits of no answer.
4.1.2.1.1. Failure of quantification identifies feature-placing sentences as primitive antecedents of paradigm sentences sharing the basic features of conventional comsign signals.
4.1.2.2. A second type of degenerate sentence is constituted by sentences with indexicals such as ‘here’, ‘he’, ‘I’, ‘this’, and ‘that’ in the subject position. These subjects do not extend reference beyond the proximate environment, and hence have the limitations of comsign signals. The personal pronouns ‘I’ and ‘we’ also have distinctive addressing functions that will be discussed below.
4.1.2.3.
Finally, there are a variety of sentences capable of paraphrases converting
their grammatical subjects into the form of paradigm sentences.
These include sentences with abstract nouns in the subject position
such as ‘Virtue has its rewards’ and ‘Redness is a quality
of roses’. As for feature-placing sentences, quantity has
no application to such subjects, as we cannot count virtues and color qualities.
These are subject to paraphrases such as ‘Virtuous persons will
be rewarded’ and ‘Roses are red’ in which abstract nouns are
replaced by qualifying adjectives and count nouns are introduced as subjects.
4.2.1. Sentence Moods. The mood of a sentence is one means of indicating the purpose for which a sentence is used (cf. 1.3.3.4). By means of the indicative, imperative, and optative moods (‘Would that it stop raining today!) the English language enables us to distinguish between descriptive, prescriptive, and expressive uses of a sentence.
4.2.1.1. The use of a given mood, however, does not indicate that a sentence is being exclusively used in one of these three basic ways. ‘The door is closed’ may be intended as a request to open the door; ‘Close the door’ conveys the information that the door is open.
4.2.2. Illocutionary Force Indicators. A much more precise way of indicating purpose is provided by illocutionary force indicators. These are expressions such as ‘I promise that …’ and ‘I am certain that …’ that indicate how the content that follows is to be understood. These are distinguished from sentence radicals expressing this content. Thus, in ‘I promise that Bill will arrive on time’, ‘I promise that …’ is the illocutionary force indicator, while ‘Bill will arrive on time’ with the name ‘Bill’ as subject expresses the sentence’s content.
4.2.2.1. There is a wide variety of illocutionary force indicators. Among them are performative prefixes such as ‘I promise that …’ and ‘I order you to …’ whose use is the performance of the speech act expressed by the verb. Thus, to say ‘I promise that Bill will arrive’ is to perform the speech act of promising, assuming certain conditions are met; to say ‘I order you to close the door’ is to perform the speech act of ordering.
4.2.2.2. There are also psychological force indicators such as ‘I believe that …’ and ‘I know that …’ used to express the degree of hesitancy or conviction with which the sentence’s content is conveyed. These can be prefixed, as in ‘I believe that Bill will come’ that expresses some doubt about Bill’s arrival. But they can also be inserted parenthetically within a sentence, as for ‘Bill will come, I know’ and ‘Bill will be late, it seems, though his train is on time’.
4.2.2.3. In the English language there are also a variety of other devices that function to express psychological force. These include the adverbial force indicators in ‘Bill will apparently be late’ and ‘Bill will certainly arrive on time’.
4.2.2.4. Other natural languages have different means of conveying force. But some form or other of force conveyance is necessary for the increase of pragmatic specialization that characterizes the sentence level of comsigns.
4.3.1. Logical Subjects. A logical subject of a sentence is a sortal used to refer to one or more reidentifiable objects. The capacity for reidentification provides the basis for the extension of reference beyond the proximate environment that characterizes the sentence level of signs. Degenerate sentences (cf. 4.1.2) are those whose grammatical subjects are not logical subjects.
4.3.2. Addresses. An address is any expression used to establish a channel of communication at the sentence level. Addresses are analogous to contact signal calls (cf. 3.2.1.2). They indicate both the source and intended audience of what is conveyed by a sentence’s content. Because of their different function, addresses are not logical subjects.
4.3.2.1. In the sentences ‘Bill, the door is closed’ and ‘Bill, close the door’ the name ‘Bill’ is an audience address indicating the person to whom the descriptions and commands that follow are intended. The role of ‘Bill’ is very different from that of the logical subject ‘the door’ in both sentences. The noun phrase ‘the door’ refers to the object to which is predicated the attribute of being closed and on which the action of closing is to be performed. In contrast, ‘Bill’ indicates who is to receive the information conveyed and who is to perform the action.
4.3.2.2. In the phone conversation ‘Hello, Bill. This is Tom. John will arrive later today’, ‘Tom’ is a source address indicating who the information conveyed by the third sentence is from. The salutation and close of written letters combine in a similar way audience and source addresses.
4.3.2.3. The personal pronouns ‘I’, ‘we’, and ‘you’ function in face-to-face communication as indexical source and audience addresses. The first-person ‘I’ and ‘we’ may indicate the individual or group that is a source of conveyed content; ‘you’ may indicate the intended audience.
4.3.4.3.1. This addressing function is obvious when these pronouns occur in performative and psychological prefixes conveying illocutionary force, as in the sentences ‘I warn you that Bill will arrive soon’ or ‘We order you to meet Bill at the airport’.
4.3.3. Avowals. Avowals are first-person reports of psychological occurrences or states. Their grammatical subject is typically the personal pronoun ‘I’. Examples are ‘I feel a pain in my toe’ and ‘I feel tired’. Also classified as avowals are reports such as ‘My pain is acute’ and reports to optometrists or audiologists such as ‘I see a blurred reddish image’, ‘The letter looks blurred to me’, and ‘The tone sounds loud’.
4.3.3.1. The indexical ‘I’ in avowals is an address indicating the source of what is being conveyed. It does not function to refer to some object. Unlike the indexicals ‘this’ and ‘that’, ‘I’ is not accompanied by gestures that indicate a referent.
4.3.3.2. Nor is ‘my pain’
in the sentence ‘My pain is acute’ a logical subject.
It can be paraphrased by ‘I have an acute pain’ in which
there is a source address and a content conveyed
by ‘have an acute pain’. The
psychological events and states reported by avowals are not objects that are
the referents of logical subjects.
4.4.1. Enabling Identification of the Remote. Increasing information content of subjects enables reference to the remote. This subject content functions as either a set of directions for locating an individual object or means for picking out from what is similar.
4.4.1.1. For singular subjects, the more general the subject, the greater reliance on indexicals related to context of utterance. ‘The book is red’ will typically require a supplementing point gesture to indicate the intended referent. The subject of ‘The book on the table in the study is red’ enables the hearer to identify what cannot be pointed to. The extension of reference beyond the proximate is enabled by the increase of subject specification.
4.4.1.2. For general subjects, increase of information content of subject enables picking out types of objects from those similar. The subject ‘crow’ of ‘Crows are black’ may signify for a hearer only being a bird. If so, the hearer would have difficulty picking out crows from other birds within an area. But if it signifies also a large bird of a certain shape, identification may become possible. This picking out or identifying function of the subject is very different from the predicative function of ‘black’.
4.4.2. Acceptance and Meaning Transfer. Acceptance of an utterance of a sentence as true is accompanied by transfer of the meaning of its predicate to its subject. Acceptance thus results in increased subject content, which in turn enables by 4.4.1 reference to the remote.
4.4.2.1. For someone to accept as true ‘Bill is tall’ is for the attribute of tallness to become part of the meaning of the proper name ‘Bill’. Being tall then becomes a means of identifying an individual as the individual named by ‘Bill’.
4.4.2.2. Similarly, to accept ‘All crows are black’ as true is for the meaning of ‘black’ to be transferred to the subject ‘crow’ and a means in the future of identifying a type of individual as a crow. To refuse to employ blackness as a means of identification is to indicate that the sentence was not accepted as true.
4.4.2.3.
What is accepted as true may later prove to be false. ‘Bill’ may come to mean ‘person
who is tall’, but even if the name acquired this meaning, ‘Bill
is tall’ is not necessarily true. Similarly,
blackness may become criterial for a bird being identified as a crow through meaning
transfer. But because the original
assertion may prove false, ‘All crows are black’ could also be mistaken. Correction of errors thus typically results
in changes in the meaning of sortal subjects, while
the meaning of attributive terms remains constant.
4.5.
Natural Languages and Semantic Fields
4.5.1. Natural Languages. A language is a set of lexical items included in its vocabulary and rules for both combining these items into noun and verb phrases and combining these phrases into sentences. A natural language is a language shared by all members of a speech community.
4.5.1.1. All sentences, no matter how primitive, are formed within a language. An utterance of ‘Apple, red’ accompanied by a pointing gesture might be formed in accordance with an implicit ordering rule indicating that the first part is the referring subject, the second the attributing predicate.
4.5.2. Fields, Exclusion, and Information. A semantic field is a set of mutually exclusive lexical items. The number of items within a given field determines the information conveyed by any given item, for information is a function of the exclusion of possibilities. The more possibilities excluded, the more information.
4.5.2.1. Attributive terms typically occupy positions within exhaustive fields. Examples include the adjective ‘red’ within a color quality field that includes ‘orange’, ‘yellow’, ‘blue’, ‘green’, etc. and ‘runs’ within an action field including ‘walks’, ‘stands’, and ‘sits’. Because such fields are exhaustive, a given object must be either red or orange or yellow or … , with the number of alternatives depending on the specificity of terms within the field.
4.5.2.1.1. Adding a lexical item to a given attributive field changes the meaning of other neighboring terms. Thus, adding ‘magenta’ to the color quality field changes the meaning of ‘red’ and ‘orange’.
4.5.2.2. Sortal terms typically occupy positions within nonexhaustive fields. The household furniture field may include items such as ‘bed’, ‘table’, ‘desk’, ‘chair’, etc., but it is not necessary that a given piece of furniture is either a bed, or a table or a desk or a chair or … .
4.5.2.2.1. Adding an item to a sortal field does not therefore typically change the meaning of neighboring terms. Adding ‘futon’ to the furniture field does not change the meaning of ‘bed’.
4.5.2.3. The number of items within a given semantic field will depend on the purposes of communication. The adjective ‘white’ as contrasted to ‘grey’ and ‘black’ might be used to convey information about neighboring houses. For purposes of painting interior walls the choice may be between shades of white described by ‘oyster white’, ‘eggshell white’, etc. Whether the term ‘oyster white’ is chosen to convey more information than ‘white’ is a function of the purposes for which its more detailed field is used.
4.5.2.4. The way in which informational
content is determined at the sentence level is fundamentally different from
the way that discrimination within a quality dimension at the natsign level determines information content (cf. 2.1.3.6.2).
At this level informational content is determined by prior learning
that changes the significance of a sign. At
the sentence level informational content is determined by choice of semantic
fields whose lexical items have relatively fixed conventional meaning.
5.1.1. Discourse Blocks. A natural language discourse is any combination of sentences formulated from the lexicon of a natural language. Such a discourse is distinguished from a specialized discourse whose vocabulary is used and understood by those with special educational backgrounds, and is therefore not the common coin of a speech community. The combined sentences may include embedded sentences in the form of relative clauses.
5.1.1.1.
A discourse block is a combination of sentences
with one or more common topics, e.g., a conversation about the weather, an
extended description of a certain couple’s activities during the week,
or instructions on how to build a model airplane.
In conversation, a discourse is often a series of questions and answers. Sentences with complex subjects (‘Either
John or Tom will arrive’) or predicates (‘John will arrive either
today or tomorrow’) and relative clauses (‘The
man who works at the bank married the woman who lives next door’) are
considered discourse blocks, since they are equivalent in meaning to combinations
of sentences.
5.1.1.2.
On our planet the capacity for formulating and interpreting discourse blocks
is unique to the human species. Sentence
formation and use may exist in infrahuman primates, and is the evolutionary
prerequisite for the discourse capacity of humans.
5.1.2. Linking Devices. These are means for combining sentences to form blocks of discourse about common topics. These devices include proforms (pronouns, proverbs, and prosentences), logical connectives, and juxtaposition.
5.1.2.1.
Proforms are words that stand in place of others, accomplishing
both economy of expression and an indication of common topics within a discourse.
They include pronouns of laziness (the ‘he’ in ‘John
ran to the store. He returned later’) and relative pronouns
such as the ‘who’ in ‘The man who ran to the store returned
later’). Also included are proverbs (‘this’
in ‘John ran.
5.1.2.2. Logical connectives include sentence connectives and inference connectives. Sentence connectives such as ‘and’, ‘or’, and ‘if … then’ combine sentences to form complex wholes. Inference connectives are expressions such as ‘Assume that … . It follows that …’, ‘hence’, and ‘therefore’ that relate premisses to conclusions inferred from them.
5.1.2.3.
Often the relation between sentences of a given discourse
block is indicated only by their juxtaposition, as in ‘Apples are red.
Lemons are yellow’ where juxtaposition of the two sentences indicates
that they are about the common topic of fruit.
There are many women named ‘
5.1.3.
There has been a historical progression in extensions
of units of logical analysis. Until
the middle of the 19th Century it was the word. With Frege, Peirce, and Russell, the unit of analysis became the sentence.
One of the major contributions of the last half of the 20th
Century through the writings of Quine and Davidson was to extend this unit to blocks of discourse.
5.2.1. Relaying. Discourse makes possible the relaying of what one person said to another. This relaying extends the range of verbal communication indefinitely beyond those within hearing distance.
5.2.2. This relaying is accomplished through quotation by means of a complex sentence that includes what is being relayed. There are two types of quotation. Direct quotation, as in ‘John said, “I will be late”, relays the words by the person or persons being quoted. Indirect quotation employs a ‘that’ clause to paraphrase what was said, as in ‘John said that he would be late’.
5.2.2.1. In spoken communication, it is difficult to distinguish direct from indirect quotation, though hand gestures parasitic on written quotation are sometimes employed. All quotation by speech is typically indirect. Direct quotation and the use of quotation marks is made possible by the advent of writing.
5.2.3. Displaying. To quote what someone said is to display what was said. This displaying must be distinguished from reference.
5.2.3.1. Consider the direct quotation ‘John said, “I will arrive tonight”’. The expression ‘I will arrive tonight’ does not refer to the sentence uttered by John. It displays the sentence in order to relay it to an audience.
5.2.3.2. This displaying function is assumed by prosentences standing for what is directly or indirectly quoted. A and B might have the following conversation. A: ‘John said that he would arrive tonight’; B: ‘Did he really say that?’. The ‘that’ in B’s question does not refer to what is indirectly quoted in A’s assertion. The reference of logical subjects is to extend the referent occasion beyond the proximate, and this is clearly not being accomplished. Instead, the prosentence ‘that’ displays what is quoted for possible evaluation.
5.2.3.3.
One kind of evaluation of what is displayed is accomplished
by predicating ‘true’ and ‘false’.
A might say to B:
‘John said that it rained last night’. B’s response might be: ‘That is true’. Though ‘that’ occupies the position
of subject, it is not a logical subject; it has no referring role. Again, its role is to display what is being
evaluated.
5.3.1. Discourses have as many uses as sentences. Their principal uses are to describe, prescribe, and express feelings and emotions. But with the addition of illocutionary force indicators, they can be also used in an indefinite variety of ways to warn, request, ask, command, etc.
5.3.2. Narratives. A narrative is a discourse block whose constituent sentences are all in the indicative mood. A narrative may be either descriptive (factual) or fictional.
5.3.2.1. Whether descriptive or fictional, a narrative provides meaning transfer from predicate to subject. In descriptive narratives, acceptance of an earlier sentence within a narrative provides the basis for a change of meaning of its sortal subject when it reoccurs in a later sentence. As an example, consider the simple narrative ‘John is tall and wears a hat. He walks with a limp. John is …’. By predicate meaning transfer (cf. 4.4.2), acceptance of the constituent sentences as true has the effect of transferring the meanings of ‘tall’, ‘wears a hat’, ‘walks with a limp’ to the second occurrence of ‘John’. The name now is understood as referring to the same individual as the definite description ‘the tall man who wears a hat and walks with a limp’.
5.3.2.2. Such transference makes possible fictional narratives. Consider a children’s story that begins ‘Lucius was a squirrel who lived in a tree. He liked to eat acorns. Lucius …’. The second occurrence of the name acquires from the preceding sentences and relative clause the meaning of being a squirrel, living in a tree, and liking acorns.
5.3.2.2.1. This transfer of meaning in fiction imposes constraints on later accounts of the discourse topic. The sentence ‘Lucius was five feet tall and weighed 120 pounds’ must be rejected as inconsistent with the meanings acquired by its subject in the narrative just described. For fiction, falsity is inconsistency, truth coherence.
5.3.3. Prescriptive Discourse. Prescriptive discourse is either pure or hybrid. The pure variety has all of its constituent sentences in the imperative mood, as for the conjunction ‘Pick up the book, and bring it to me’. A hybrid discourse combines sentences in both indicative and imperative moods, as for the conditional imperative ‘If it is raining, then don’t go outside’.
5.3.3.1. One important function of indicative
descriptions within hybrid prescriptive discourse is to specify through meaning
transfer an imperative subject’s reference. This function is performed in the following
combination of sentences: ‘John is tall, and wears a brown coat. Pick him up at the train station’. The transference of being tall and wearing a
coat enables the hearer to identify the individual referred to by the pronoun
‘him’ and perform the prescribed action.
5.4.1. An inference is a combination of sentences one of which, the conclusion, is claimed to “follow from” the others, the premisses. Inference connectives such as ‘therefore’, ‘hence’, and ‘it follows that’ link premisses to conclusions, and distinguish inferences from other types of discourse.
5.4.2. Deductive Inferences and the Principle of Identity. One form of inference is a deductive inference whose conclusion follows necessarily from its premisses by virtue of the meaning of the inference’s constituent terms. A simple example of a deductive inference is the inference ‘It is raining and cloudy. Therefore, it is raining’ whose conclusion follows from the premiss by virtue of the meaning of sentence connective ‘and’.
5.4.2.1. The necessary relation between premisses and conclusion of a deductive inference requires the assumption of the principle of identity, the principle that the same proposition expressed by sentences within an inference must retain the same truth value within the context of the inference. Without this assumption, no inference would be deductively valid. This includes the inference ‘John is sitting; therefore, he is sitting’, for without the assumption it would be possible for John to change for sitting position to standing during the time taken to utter the premiss and utter the conclusion, and thus for the premiss to be true, the conclusion false.
5.4.2.2. Within the context of an inference we also assume constancy of reference for subject terms. It is possible for the name ‘John’ to be used to refer within different contexts to two different individuals, say John Smith and John Jones. If reference were to vary within an inferential context, then ‘John is sitting; therefore John is sitting’ would be invalid. This constancy requires the distinction between the denotation of a term and the reference of a term. ‘Denotation’ is a logical term whose use is restricted to inferential contexts.
5.4.3. Other Inferences. Besides deductive inferences, there are a variety of others. These are distinguished from deductive inferences by the defeasibility of their conclusions, the fact that it is possible for their premisses to be true and their conclusions false. The task is then to propose for these nondeductive inferences procedures that reduce (though never eliminate) this possibility.
5.4.3.1. Included among nondeductive inferences are inductive inferences and analogical inferences. Inductive inferences typically have general conclusions that follow from statements about particular individuals. Users of analogical inferences typically infer from the fact that some individual or class of individuals has some property to the conclusion that another individual or class like them in most respects also has this property.
5.4.3.2. Also included among inferences with defeasible conclusions are practical inferences. These typically have an ‘ought’ or ‘should’ conclusion requiring a certain action and a mixture of premisses expressing preferences and valuations and descriptive premisses. An example would be ‘I want to keep warm this winter. In order to do so I must buy a coat if the winter is a cold one. It will be a cold winter. Therefore, I should buy a coat’. The first premiss is expressive, the second and third are descriptive, and the conclusion is prescriptive. As for inductive and analogical inferences, there are procedures for reducing the risk of a mistaken conclusion and of performing an action that should not have been performed.
5.4.3.3. There are two basic kinds of practical inferences, which are distinguished by the nature of their expressive premisses and the relation between premisses and conclusion.
5.4.3.3.1. Prudential inferences have an expressive singular or plural first-person premiss of the form ‘I (we) want (or desire) to attain E’, where E is some end or goal.
5.4.3.3.2. Moral inferences have a plural premiss of the form ‘We want to attain E’ expressing a shared want within a community. This is combined with a generalization premiss that has the effect of preventing members of the community making themselves exceptions to a practice that brings about what all want. An example of a moral inference is ‘We want conditions of mutual trust. Only if almost all keep their promises is mutual trust possible. If almost all should keep their promises, then everyone should. Therefore, I (he or she, they) should keep my (his or her, their) promises.’ The last premiss is the generalization premiss that rules out an individual making herself an exception to a rule having general benefit.
5.4.3.3.3.
The constituent sentences of practical inferences have specialized functions. Their combination within these inferences retains
a feature of sign use and interpretation characteristic of more primitive
sign levels (cf. 1.3.3).
This confers on practical inferences special importance in relation
to the social role that philosophy fulfills.
6.
Specialized Discourse and Philosophy
6.1.
Writing and Specialization
6.1.1. Writing and Civilization. With the introduction of writing, discourse blocks became texts. This introduction makes possible an advanced civilization, an enlarged community with social institutions using specialized forms of language. This development is in turn made possible by some advantages of writing.
6.1.1.1. By means of quotation and relaying, voice communication can be extended to an audience distant in time and space by means of intermediaries (cf. 5.2.1). Writing introduces direct quotation and the possibility of eliminating the distortion that is inevitable with verbal relaying. This is especially important for exercising social control through written edicts reinforcing the authority of the source and for preserving past innovations.
6.1.1.2. Writing is an aid to the linking characteristic of all discourse. The sequential linking in time of heard speech is replaced by simultaneous viewing. Sequential linking imposes demands on memory that can be overcome by writing. This is especially important in facilitating the use of deductive inferences, and makes possible the development of mathematics.
6.1.1.3. Writing allows verbal counting to be replaced by numerals as special symbols. This in turn makes possible more precise measurement with the introduction of ratios and rational numbers.
6.1.1.4. Writing also makes possible the development of the special vocabularies used within different social institutions. These include the vocabularies of science, law, literature, and religion.
6.1.2. Pragmatic Specialization. At the level of natural discourse, pragmatic specialization is indicated by sentence mood, illocutionary force indicators, and context. At the level of specialized discourse, these devices continue to be employed. But we increasingly rely on context in the form of a discourse’s institutional source to determine the purposes for which a text is used. Publication in a scientific journal marks an article as descriptive; publication as a novel marks a book as fiction. The forms of address so important to face-to-face communication (cf. 4.3.2) begin to play a less prominent role.
6.1.2.1. Hybrid prescriptive discourse constitutes one form of
natural language discourse (cf. 5.3.3).
A written law is typically hybrid in form.
Here descriptive sentences may describe the scope and duration of what the law prescribes,
the conditions for its application, and the circumstances under which it arose.
Circumstances may be described within the “whereas” of a
law’s preamble.
6.2.
Discourse Frameworks and Category Terms
6.2.1.
Discourse Frameworks. The general
types of discourse used within daily life and specialized social institutions
are discourse frameworks, each with characteristic
topics, whether the weather, a neighbor, fluids, electrons within solids,
real numbers, fictional warriors, or citizens to which a law applies.
6.2.1.1.
The most basic distinction is between the frameworks
of everyday conversation formulated within a natural language used and understood
by all (ordinary language) and frameworks used for more specialized purposes.
The topics of natural language discourses include macroscopic
material things – the tables, apples, and people we directly observe
– events, and processes. These are commonly
distinguished from the theoretical entities
that constitute the topics of some of the natural sciences. Theoretical entities such as genes, atoms, nuclear
particles, and force fields are typically observed only with the aid of special
instruments. In the absence of any
kind of observation they have been postulated for explanatory purposes.
6.2.1.2. The topic of a given discourse framework can sometimes be revealed with the aid of logical paraphrases in which original subjects of sentences are replaced by a single subject having wider referential scope. These paraphrases are used by logic to insure a common subject for the purposes of evaluating as valid or invalid a deductive inference.
6.2.1.2.1. Consider, for example, the standard Aristotelian syllogism ‘All humans are mortal. All Greeks are humans. Therefore, all Greeks are mortal’. Deductive logic first paraphrases the first premiss by ‘All things are such that if they are men, then they are mortal’, thus introducing the term ‘thing’ as the logical subject, and converting the original subject ‘men’ into a predicate. Similar paraphrases of the second premiss and conclusion will have the effect of instating ‘thing’ as the common subject of the inference.
6.2.1.2.2. After paraphrase, the syllogism is represented symbolically by "x(Px→Mx), "x(Sx→Mx), \"x(Sx→Px), where the variable x represents the common subject ‘things’, or more accurately, ‘macroscopic material things’. This common subject denotes what is called the domain of the variable.
6.2.1.3. Each discourse framework, whether those of ordinary language (as for our syllogism example), the natural sciences, mathematics, or fiction, will have its common generalized logical subject revealed by such paraphrase and symbolic representation. For ordinary language this will include the subjects ‘event’ and ‘process’. For the natural sciences this may be ‘theoretical entity’, while for mathematics it may be ‘number’ and for fiction ‘fictional object’.
6.2.1.4.
Logical subjects function different within different discourse frameworks. Within the material thing frameworks, subjects
function to both extend the referent occasion and acquire meaning through
transference from predicates (cf. 4.4.2). The subjects of formulas and sentences of mathematical
and fictional frameworks, in contrast, lack reference, and have only transference
functions.
6.2.2.
Category Terms. A generalized logical subject common to a given
discourse framework is a category term.
Category terms are terms with the widest
extension of sortal terms within a discourse framework.
Within the framework for ordinary language, for example, we might specify
a hierarchy constituted by ‘Socrates’, ‘Athenian’, ‘Greek’,
‘human’, ‘organism’, and ‘macroscopic material thing’.
The last of these is the category term for this framework, and is used
in logical paraphrase to insure a common subject for any inference formulated
within it.
6.2.3. It is tempting to generalize further and introduce
as the subject of paraphrase the sortal ‘object’
to be applied to all discourse frameworks as their common logical subject.
6.2.4. Such a generalization would be a serious mistake. There are different standards for judging the
truth or falsity of sentences relative to specific frameworks and means of
identifying and counting the various types of objects delimited by category
terms. But we have no means of identifying
and counting objects in general. There is no answer to one who asks how many objects
are in a certain room, though it makes sense to ask for a count of macroscopic
objects or microscopic objects such as atoms.
6.2.5. Objects. The term ‘object’ should instead be regarded as a place marker for a logical subject, not as itself the subject of a sentence. In order to have content sufficient for identification, the term ‘object’ must be relativized to a specific discourse framework.
6.2.5.1. There are procedures for identifying a number within mathematical discourse and distinguishing it from other numbers. These are very different from the empirical procedures used in identifying macroscopic objects and theoretical entities, and from those used in identifying the fictional characters of a novel.
6.2.5.2. There are also different standards employed in evaluating the truth or falsity of propositions formulated within different discourse frameworks. We use empirical standards for the macroscopic material thing and theoretical entities framework, while coherence is the standard for fiction and provability from axioms for mathematical propositions.
6.2.6. Type Crossings. Discourse relativity extends from subjects to predicates. This is shown by type crossings, combinations of subject and predicates within descriptive sentences for which there are no standards for evaluating a proposition as true or false. In the absence of such standards, the sentences are cognitively meaningless, though may be used for metaphorical and poetic purposes.
6.2.6.1.
Examples of type crossings include ‘
6.2.6.2.
There are also pragmatic errors occurring when standards
of exactness appropriate to one discourse framework are applied to another
with different standards. Examples
would include ‘Hamlet was 6 feet and .075 inches tall’ and ‘This table top is 2.0000067 meters in diameter’. Normal standards of measurement cannot be applied
to fictional characters; the measurement of table tops by
a carpenter are not done with the same degree of accuracy as an engineer
designing a critical part for a machine.
6.3.1.
The Origins of Philosophy. The
origins of philosophy as a distinctive discipline can be traced to the development
of civilizations with writing and institutional specialization in
6.3.1.1. Some attribute philosophy to cultures lacking systems of writing, but having oral mythological traditions that serve to organize and make sense of the world around them and provide expressions of shared values and hopes. They thus regard the early Greek mythological tradition preserved in Homer’s Illiad and Odyssey as an early form of philosophy. But this is inconsistent with our understanding of philosophy as a distinctive discipline discussing issues arising from competing perspectives. We should instead describe these unifying oral traditions as belonging to a pre-philosophical period.
6.3.1.2. Early Greek philosophy reflects conflicts between the perspectives developed within established and emerging social institutions. The established institution was that of religion with its priestly caste believed to mediate between citizens and the divine and confer legitimacy of political power. This institution gave expression to traditional values shared by most Greeks. The emerging institutions were those of mathematics and the natural sciences.
6.3.1.3. We can easily identify some early philosophers with institutions, and can see how this identification influences their philosophical view. The early cosmologists, for example, can be identified with the natural sciences that later emerged, the Sophists with the legal profession, Pythagoras and Plato with mathematics, and Aristotle later with biology and the social sciences. Their perspectives were often used to provide rational support for the views of the dominant religious institution, though sometimes (as in the case of the Sophists’ questioning of universal values) to criticize them.
6.3.1.4.
This identification with institutions continues in later philosophy, most
obviously with Descartes’s attempt to apply
the deductive model of reasoning derived from mathematics to philosophy and
Nietzsche’s later identification with the arts and literature.
7.1.
Metaphysics and Reductions
7.1.1. Metaphysics is the branch of philosophy that attempts to satisfy philosophy’s urge for comprehensiveness. It has been used to justify the priority of forms of discourse in some social institutions over others. This justification can take the form of reducing the topics of the less privileged form of discourse to those of the privileged. Such reduction is by means of metaphysical assertions of identity.
7.1.2. The Relativity of Identity. Metaphysical assertions of identity assume an absolute conception of identity that differs from the identities used within specific discourse frameworks. We normally evaluate assertions of identity relative to types of sortals and different standards. For example, it may be true that an individual John Smith is the same person he was ten years ago in the legal sense of being held responsible for past actions or in the psychological sense of retaining the same memories. But it is false that he has the same body in the sense of being the same collection of cells, though it may be true that the body of today is the same as that of ten years ago in the sense of being continuous with it. Whether we choose to apply the standard of sameness of parts or that of continuity between his past and present body will determine different evaluations.
7.1.3. Metaphysical identities between terms used within discourse frameworks pose problems because of the lack of similar standards for identification and evaluation.
7.1.3.1. The term ‘water’ is a mass noun occurring in the discourse framework of macroscopic material stuff, while the term ‘H2O’ occurs in the framework of theoretical entities. We accept as true the identity ‘Water is H2O’, though the identity sentence relates terms from different frameworks. This acceptance assumes standards for determining the sentence’s truth or falsity. The sentence is not itself metaphysical.
7.1.3.2. A reduction is a special form of an assertion of identity used within philosophy. An assertion of one relates a topic of a less privileged discourse framework to that of a more privileged. The identity ‘Water is H2O’ may be used simply for educational purposes to describe a part/whole relation. Asserted as a reduction of water to H2O molecules, however, it accords the ordinary discourse framework of macroscopic stuff a less privileged status than the specialized discourse framework of theoretical entities used by the natural sciences. Used for reductive purposes, the identity is metaphysical.
7.1.3.2.1. The reduction relation expressed by ‘A is reduced to B’ is asymmetric, while standard identities are symmetric. A reductive sentence is an exceptional form of identity used to bestow privileged status.
7.1.3.3. The term ‘pain’ is a psychological term used in first-person avowals, while the term ‘brain process’ is used to refer to a publicly observable process. Philosophers have formulated identities such as ‘The pain I am presently experiencing is identical with a particular brain process’, a combination of terms also from different discourse frameworks. Such a sentence is invariably used as a reduction of the mental or psychological to the physical topics of natural science discourse frameworks.
7.1.3.3.1. Parallel to these reductions are assertions of sentences such as ‘A pain is not a brain process’. These nonidentity sentences are used in formulations of dualism, while identities are used by advocates of monistic physicalism.
7.1.3.4. The identities and nonidentities used by monism and dualism should be recognized as special forms of illicit type crossing.
7.1.3.4.1. The noun ‘pain’ has associated
with it no criteria for identifying one experienced pain as the same as another,
for such criteria become associated through through
public applications of a term. But
without such criteria for identification, it would not seem that ‘pain’
can function as a term of either a meaningful identity or nonidentity sentence.
7.1.3.4.2. Both the terms ‘water’ and ‘H2O’ can be used to refer to portions of stuff and molecules, and we use coincidence of reference as a basis for judging the identity ‘Water is H2O’ to be true. But ‘pain’ does not have the referring role of extending the referent occasion, nor is it like mathematical and fictional terms in having a purely meaning transfer function (cf. 6.2.1.4).
7.1.3.4.3. The term ‘pain’ instead has an essentially predicative use as occurring in an avowal such as ‘I feel (or have) a pain’. The sentence nominalization ‘my feeling pain’ can occur as grammatical subject in the sentence ‘My feeling (of) pain is acute’, or in shortened form as simply ‘My pain is acute’. But the proper logical form of this sentence is ‘I have an acute pain’, since ‘my pain’ is not a logical subject with referential or meaning transfer functions.
7.1.3.4.4. Lacking identifying criteria and referential and meaning transference functions, the mental term ‘pain’ does not qualify as a term of an identity or nonidentity. Similar considerations apply to other mental terms such as ‘my sensation of redness’, ‘my belief that it will rain’, etc. When paired with physical terms in identity and nonidentity sentences, all become components of illicit type crossings.
7.1.3.5. Instead of educating us on the relation
between the macro world of daily life and the micro world of the physical
sciences, mental/physical identities such as ‘My present pain is a brain
process’ are reductive identities whose effect is to bestow privileged
status on the discourse frameworks of the natural sciences. In asserting such identities, philosophers assume
the role of representatives of the natural sciences, and philosophy becomes
the handmaid of the experimental sciences. The desire for comprehensiveness is now regarded
as being satisfied by an instrument extending the domain of the experimental
sciences over what had been considered topics of distinct types of discourse.
7.2.1. The bestowal of privileged status is often made within philosophy by applying variants of the abstract noun ‘reality’. Thus a philosopher asserting that water is a collection of H2O molecules or that gold is the chemical element with atomic number 79 may say that water is really the molecules and that the observed properties of water are only appearances, that gold is really the chemical element and only apparently yellow and malleable.
7.2.1.1. Scientific realism can take the form of a thesis about the “real” reference of types of macroscopic substances such as water, gold, ants, or tigers, the so-called “natural kinds” referred to by the nouns of natural languages. The meaning of such nouns is acquired through meaning transference from predicates when occurring as subjects of accepted sentences. But each such sentence could be mistaken; the transferred predicate may hold of the subject’s referent. It seems to follow that the subject’s referent is not what satisfies the combination of transferred predicates that expresses its meaning. What then is this referent? Scientific realism’s answer is that it is whatever satisfies the theoretical description given by the natural sciences of what was referred to when the subject term was first introduced. As an example, the term ‘gold’ refers to that type of stuff with the chemical properties we can ascribe to whatever was initially referred to when the term was first introduced into natural languages, a reference that is historically transmitted to the term’s present use.
7.2.1.2. There are several difficulties with this view. In many cases we are entirely ignorant of a term’s initial application. Does this mean that we don’t know what it refers to? Secondly, the view seems to give intuitive results. The term ‘gold’ when first introduced may have been used to refer to pyrite or “fool’s gold.” Is the reference of the term we presently use then to an iron sulfide mineral? And finally, subject terms are neither true or false. Their role is to enable coincidence of identification, to enable audiences to identify the same objects in order to determine whether predicates are true or false of them. The theoretical descriptions of the sciences enable more exact discriminations of objects, and thus a greater guarantee of coincidence of identification among those able to apply such descriptions. But this in itself does not answer the question of what the “real” reference of a term is, for this question does not admit of an answer.
7.2.2. Like the ‘is’ of identity, ‘real’ is discourse relative. It has no application in adjudicating between forms of discourse used for different purposes.
7.2.3.
Metaphysics is sometimes conceived as the branch of philosophy that distinguishes
the real from the apparent. This conception
should be rejected.
7.3.1. ‘Reality’ is only one of three central abstract nouns used in metaphysical discussions with vacuous generality, and hence without meaningful application. The others are ‘being’ and ‘existence’. Like ‘reality’, they are used in relation to the topics of different discourse frameworks.
7.3.1.1. By far the most versatile of the three is ‘being’, the nominalization of the verb ‘to be’. This verb occurs as both the ‘is’ of identity (‘Mark Twain is Samuel Clemens’, ‘Water is H2O’), of predication (‘Mark Twain is famous’, ‘Water is a fluid’), and of existence ‘There is a table in the next room’ and ‘I am’. Given these very different uses of the verb, it should be obvious there is no intelligible application of ‘being’ as its nominalization. Nor do the nouns ‘being-in-itself’ or ‘modes of being’ used in metaphysical discussions have meaning for us, though they may be used to evoke religious feelings through their association with traditional religious language.
7.3.1.2. The noun ‘existence’ is synonymous with ‘is’ in its existential sense when we say ‘There exists (is) a table in the next room’ or ‘I exist’. For such a sentence there are standards for evaluating it relative to the discourse framework of macroscopic material things. There are different standards for evaluating the mathematical sentence ‘There exists a prime number greater than 101’, the fictional sentence ‘There exists a clown in Shakespeare’s Hamlet’, and the philosophical ‘I exist’. These standards for evaluation are specific to the different frameworks.
7.3.1.3.
Metaphysics formulates sentences such as ‘Only
material things exist’ (materialism), ‘Only I and my ideas exist’
(solipsistic idealism), ‘Material things don’t exist, but ideas
of them do’ (idealism), and ‘Numbers exist’ (mathematical realism).
As for metaphysical applications of ‘real’, these uses of
‘exist’ purport to adjudicate between discourse frameworks, but
have the effect of bestowing privileged status.
In the absence of evaluative standards for them, the sentences are
cognitively meaningless.
7.4.
The Role of Contemporary Philosophy
7.4.1. The subject titled “metaphysics” has neither had a clearly defined purpose nor subject matter, having been understood in at least three different ways. First, it has been allied with what is also called “ontology,” the science of being, and in this role usually performs the reductive role described above in 7.2. and 7.3 above. Secondly, it has been understood since Kant as the discipline that deals with questions related to the justification of religious beliefs, questions such as whether God exists, whether we have free will, and whether we have a mind or soul that outlives the destruction of the body. And finally, the subject can be conceived in the Aristotelian sense as the discipline that analogically extends certain select central features to all natural bodies.
7.4.2. Metaphysics and Religious Beliefs. For many, the basis of religious belief is the authority of Scriptures. For others, however, no religious belief should be held unless it is also reasonable, that is, consistent with what we justifiably accept. Some of the latter claim that some or all religious beliefs fail to pass the test of reason, and therefore should be rejected as false. Others claim that these beliefs are expressed within a special discourse framework expressive of hopes and values to which cognitive standards cannot be applied. Within this framework, belief is called faith.
7.4.2.1. The total separation of faith allied to hope from cognitive belief seems to ignore the distinction between a wish and a hope. We can wish for anything, no matter how unlikely it will come about, but hope seems to imply some reasonable belief in fulfillment. This imposes a requirement on religions to adjust beliefs in order to comply with cognitive standards.
7.4.2.2. Rational argument has been historically used to support the reasonableness of beliefs in a supernatural controlling power, the survival of a person after death, and the capacity for making choices between alternatives. The belief in the capacity for choice seems incompatible with the principle of causality, the principle that every event is an effect of causes that could not have resulted in any other outcome. In effect, the principle states that every statistical generalization stating probabilities can be converted to a uniform generalization through further specification of the cause. What is commonly overlooked is that every causal generalization can also be converted to a statistical generalization through more precise specification of the effect (cf. 3.1.2.1.1). Observed uniform regularities are as much the product of ignorance and imprecise measurement as probabilities. The principle of causality is not a methodological principle of the natural sciences, which are neutral on the question whether there is spontaneity of behavior at all levels of organisms. At least on the question of free will, religious belief is consistent with the assumptions of the discourse frameworks of the sciences.
7.4.3.
Analogical Extensions. The more fruitful conception of metaphysics
is derived from Aristotle. Metaphysics
is understood as a discipline that begins by describing some subject matter
regarded as “first in the order of knowledge” and then analogically
extending these descriptions. For Aristotle,
this subject matter was macroscopic objects described as substances with form
and matter. These descriptions were
then generalized to all natural bodies and a Prime Mover (pure form) and substratum
(pure matter). For semiotic, the initial
subject matter is sentences as used and interpreted in daily life with meaning
and reference, with extensions made to more primitive forms of signs and specialized
discourse. The historical period in
philosophy from Descartes to Peirce and Morris provides
the background for this shift of subject matter.
7.4.4. Guided by this conception, philosophy becomes a classificatory study of similarities and differences between signs at different levels. Quasi-political battles between advocates of “scientism” with its reductionist projects and those advocating a dominant role within philosophy for literature and the arts are replaced by a common effort to understand relationships between the specialized discourses used within different social institutions and between these discourses and the language of everyday conversation shared by all.