Quine, Willard Van Orman;
From Stimulus to Science
Harvard University Press, 1995, 114 pages [gbook]
ISBN 0674326350, 9780674326354
topics: | philosophy | science | history |
We and other animals notice what goes on around us. This helps us by suggesting what we might expect and even how to prevent it, and thus fosters survival. However, the expedient works only imperfectly. There are surprises, and they are unsettling. How can we tell when we are right? We are faced with the problem of error. These are worries about our knowledge of the external world. To deal with them we have had to turn inward and seek knowledge of knowledge. "Know thyself": the injunction is attributed to Socrates and even to Thales, purportedly the father of philosophy. Thales and his successors were concerned not only with man and his errors; they speculated on the cosmos. But anxiety over the problem of error continued through Greek antiquity. The paradoxes of Zeno and Eubulides were calculated to show the limitations of our judgment, as were the sophistries of the Sophists. The Skeptics took the melancholy conclusions to heart. 1
How do we know things? Plato held that we do so by apprehending ideas-really forms, as we would say-that are the essences of things. He thought we were born knowing these forms and their interrelations in a blurred way, and that they could be brought into focus by Socratic dialogue. It seems from one of the dialogues, the Meno, that Plato arrived at this theory by thinking about mathematical argumentation from self-evident truths. Somehow, though, he accommodated observation too. "Save the appearances," he wrote. Aristotle tried to shore up the ways of knowing. He formalized the syllogism. Knowledge itself, however, outpaced knowledge about knowledge. Natural history throve in Aristotle's hands, and mathematics in those of Eudoxus and Euclid. Eratosthenes even calculated the size of the earth, nearly enough. But darkness descended, and mists of myth and mysticism settled in for a thousand years. Ways of knowing dwindled to one: higher authority. Remnants from Aristotle were deemed authoritative, but now had he known? Infinite regress loomed. The problem was shelved by positing supernatural revelation. This position was a stubborn one, for we cannot then question divine revelation without begging the question. [NOTE: posit ==> position - etymology L. positus, pp. of ponere, "put, place", noun 1703 (dance steps), 1883 (sexual intercourse), back to verb in 1817 "to put in a given position"]
Roger Bacon did beg it, along about 1290, by espousing observation and experiment. Two and a half centuries later, Copernicus made the breakthrough that put science unmistakably on the upward track. For fourteen centuries straggling astronomers has struggled with Ptolemaic epicycles to systematize an astronomy centered on the earth; but at last Copernicus put our planet in its place and set it moving. Thus inspired, Francis Bacon took up the old question of the ways of knowing. The spirit of Roger Bacon was reawakened, but now with more substance and sophistication, the wisdom of hindsight. Science had broken through, though traditionalists tried to restrain it. A full century after Copernicus, the clergy prosecuted Galileo for embracing the Copernican heresy. One thinks of the creationists today, one hundred thirty years after Darwin's Origin of Species. [p.3] Despite obstructions, science attained full flower fifty years after Galileo's work. It flowered in Newton's Philosophiae Naturalis Principia Mathematica (1687).
In the more broadly philosophical arena, meanwhile, there was Thomas Hobbes. He was twenty-four years younger than Galileo, and was inspired by Galileo's strides. Hobbes professed utter materialism, indeed mechanism, like Democritus two thousand years before: there is nothing but matter in motion. Thought is motion in the brain. Hobbes's view of knowledge was strikingly modern. Our sensations are the effects upon us of the otherwise unknowable material world. It is on these that we base our ideas about the world, and we have nothing further to go on but the meshing of the ideas. Hobbes uses the word 'idea' in its modern sense, to mean something like a thought or a concept. He was the first to do so in English, according to the Oxford English Dictionary. It was an odd reversal of Plato's usage. For Plato ideas or forms had been reality par excellence; things of the material world were their imperfect counterfeits. For Hobbes and us, ideas are rather man's faltering attempt to encompass material reality. [ETYMOLOGY "Idea":: 1430, "figure, image, symbol," from L. idea "idea," and in Platonic philosophy "archetype," from Gk. idea "ideal prototype," lit. "look, form," from idein "to see." Sense of "result of thinking" first recorded 1645. Archetype = "original model", original, that which is real outside. Link to "idein", to see. ]
Rene Descartes was Hobbes's junior by eight years. He was a dualist: he recognized both mind and matter. Descartes came to grips, more vigorously than Hobbes or Bacon, with the question of how we know. In his famous thought experiment he proceeded from scratch. He tried doubting everything, but found that he could not doubt that he was doubting. He concluded that he existed, as a mind. Then he proceeded to the existence of God. Sensing more treacherous ground there, he offered four proofs. The existence of matter then came easily: we have a clear and distinct idea of matter, and since God is by definition good, he would not give us a clear and distinct but false idea. Such, roughly stated, was Descartes's theory of how we know. Clear and distinct ideas are knowledge, and God given; confused ideas are not. There are echoes here of Plato's doctrine on innate knowledge and reminiscence, and of the Hebrew and Christian doctrine of divine revelation. But the Platonic bit may have been a case of parallel inspiration rather than heritage. Plato's view was evidently inspired by mathematics, and Descartes was a mathematician.
In the five-year period 1646-1650 five neoclassical philosophers were alive: Hobbes, Descartes, Spinoza, Leibniz, and Locke. In the history of epistemology the next significant figure after Descartes is John Locke. Locke shunned Descartes's theological epistemology. Like Hobbes, he saw knowledge in the coherence of ideas. He accepted the material world as real, but as known only tentatively by conjecture from ideas. Sense impressions, caused by the material world, implant our simple ideas; we build or abstract all other ideas from these. Contrary to Descartes, Locke repudiated innate ideas. He subscribed to the empiricist manifesto: nihil in mente quod non prius in sensu. [There is nothing in the mind that isn't there (earlier??) in the senses] Locke did not explain how to form complex ideas of material objects, real or fictitious, on the basis of simple sense impressions. He wrote of the association of ideas by contiguity, succession, and resemblance, but this is the barest beginning of what goes into the most primitive report on the material world around us. What of our identification of an intermittently observed body as the same body? An identical body can look different over time, and different bodies can look alike. Much remained to be explained.
[p.5] Bishop George Berkeley, fifty years younger than Locke, saw no cogency in Hobbes's or Locke's conjecture of a material world. Nothing exists, Berkeley held, but what is directly perceived. Nothing, therefore, but sensory patterns, or occurrences of sensory patterns? No, he was more generous than that. He admitted souls; we perceive ours. And somehow he admitted God. This done, Berkeley provides for the persistence of things irrespective of whether or when they are perceived by man or beast; for they remain faithfully perceived by God. Berkeley's disavowal of matter, then, would seem to be a matter of words.
David Hume, twenty-six years Berkeley's junior, acquiesced rather in the conclusion that there simply is no evidence for the continued existence of an object between one occasion and another of our perceiving it. The very identification of it as the same object, on the one occasion and the other, is in his view a confusion of identity with similarity. Locke, Berkeley, and Hume were the classic British empiricists, and their pertinent writings appeared in the years 1690-1757. All three agreed that our lore about the world is a fabric of ideas based on sense impressions. Regarding the structured details of the fabric and its fabrication, all three were at a loss for the rudiments of an account. The idea is a frail reed. As Wittgenstein observed, even a simple sense quality is elusive unless braced by public language. An individual might reckon many sensory events as recurrences of one quality on the strength of resemblance of each to the next, despite a substantial accumulation of slight differences. Public naming and monitoring are what arrest such drift. Random deviations of individual speakers are held within bounds by the speakers' communicating with one another. Public words anchor ideas. Irresponsible appeal to the idea is still our popular usage. The purpose of language is said to be the communication of ideas. Irresponsible appeal to the idea is still our popular usage. The purpose of language is said to be the communication of ideas. We learn a word from our elders by associating it with the same idea, and we use it in the communication of ideas. How do we know that the words we use to express our ideas are conjuring up the same ideas in the minds of our listeners? Words and observable behavior are all we have to go on, and the idea provides only the illusion of an explanation.
John Horne Tooke urged the point in 1786, ten years after Hume's death. I only desire you to read [Locke's] essay over again with attention to see whether all that its immortal author has justly concluded will not hold equally true and clear, if you substitute the composition, of terms wherever he has supposed a composition of ideas . . . The greatest part of Mr. Locke's essay, that is, all which relates to what he calls the abstraction, complexity, generalization, etc., of ideas, does indeed merely concern language. (Pp. 37-38) Tooke was a kindred spirit of his remote predecessor William of Ockham and other medieval nominalists, who had dismissed abstract objects as flatus vocis, vocal breeze. Tooke's was a major step toward what Richard Rorty has called the linguistic turn. If there is sense to be made for the compounding of ideas, clearer sense can be made of the compounding of language. Words, unlike ideas, are out where we can see what we are doing.
Much further freedom in the constructing of ideas -- or, now, terms -- was achieved by Jeremy Bentham a few years later in his theory of fictions. He observed that to explain a term we do not need to compose a synonymous phrase. We need only explain all sentences in which we propose to use the term. It is what is now called contextual definition. Bentham's motive was ontological: he wanted to be able to introduce one or another useful term without being charged with assuming some controversial object for it to designate, or objects for it do denote. The seeming object or objects could in this way be dismissed as innocent fictions.
[p.7] In India there has been debate since the seventh century over whether sentences or words are the primary vehicles of meaning. Since the lexicon or Wortschatz is finite whereas the realm of admissible sentences, the Satzschatz, is boundless, a systematic guide to a language must consist of a dictionary of words and a grammar for building sentences from them. The words, on the other hand, will mostly be explained in the dictionary by their use in illustrative sentences. The goal of the whole enterprise is to inculcate facility in understanding and producing correct and useful sentences. In learning our native language we zigzag similarly. We learn a simple sentence as a whole, and then we project a component word of it by analogy into the construction of another sentence. Nowadays an appreciation of contextual definition - Bentham's insight - has lent support to the view of sentences as the primary vehicles of meaning. It is a view that Gottlieb Frege vigorously espoused a century after Bentham. [George Boole's contextual definition of OPERATORS in differential calculus - e.g. d/dx can stand for an operator that cannot stand alone without the f(x). Boole's innovation was to take an expression such as d2/dx2 f(x,y,z) + d/dy2... ] in the compact fashion (d2/dx2 + d2/dy2 + d2/dz2) f(x,y,z), and then to manipulate it as if it stood for a sum of three genuine quantities multiplied into f(x,y,z).
It was this very example that inspired [PM v.1 p.66] Bertrand Russell's familiar contextual definition of the singular description (ix)Fx: G(ix)Fx for (Ey)(Gy ^ Vx(x=y EQ Fx)) [The singular description is ] one of the many deft contributions of Whitehead and Russell's three forbidding volumes of Principia Mathematica, which appeared in 1910-13. Their heroic project was to clarify the whole intricate structure of classical mathematics by deriving its principal concepts, step by step and definition after definition, from a slender basis of clear and simple primitive terms, and deriving its principal laws pari passu from a few postulates. . . . The economical foundation achieved in Principia, and further reduced by subsequent logicians [Frege/Peirce/Peano], now comprises only the truth functions and quantification of elementary logic plus the two-place predicate \epsilon of class membership. The whole conceptual scheme of classical mathematics boils down to just that. . . . The conclusion [Whitehead and Russell and Frege before them] drew was that mathematics was translatable into pure logic. They counted membership as logic. So mathematical truth is logical truth, and hence all of it must be deducible from self-evident logical truths. This is wrong, as transpires in part from Kurt Godel's paper of 1931 and in part from findings by Russell himself in 1902. Buoyed by their achievement, Russell reflected in 1914 on realizing the dream of empiricist epistemologists: the explicit construction of the external world, or a reasonable facsimile, from sense impressions, hence from simple ideas. He adumbrated it in Our Knowledge of the External World, and a dozen years later Rudolf Canap was undertaking to carry it out. Carnap's effort found expression in Der logische Aufbau der Welt (1928). [10]
Carnap's building blocks were to be sensory elements, as in the constructions dreamed of by the old British empiricists. But what sensory elements? Atomic sense data? The Gestalt psychologists claimed that we are first aware of various organized wholes, and then we abstract the atoms. Carnap circumvented that issue by settling for unorganized global experiences. Each of his elements was the individual's total experience at the moment, or perhaps during the specious present. These global units he called elementary experiences. Carnap's basic relation between elementary experiences was {\it remembering as similar}. I shall call it R. One elementary experience, x, bears R to another, y, if x includes a memory of y as partially resembing x. [p.10] Here, as in Principia Mathematica, economy was part of the game. Carnap allowed himself free use of logic and mathematics, and in addition just one two-place predicate, the one I am calling R. He does not need to assume a further predicate to denote the elementary experiences, for he can define it: an elementary experience is anything that bears R to anything, or to which anything bears R. This can be expressed in terms of R and logic. Next he defines part similarity of elementary experiences: it holds if one of them bears R to the other, forward or backward. This definition would seem to deliver not just "part similar", but "remembered as part similar"; but let us pass over that. By a more subtle definition Carnap introduces what he calls similarity circles. A similarity circle is any largest class of elementary experiences each of which is part similar to each [every one]. It is the largest in the sense that no elementary experience outside the class is similar to every member. [p.11] This brings him to the point where he can define the notion of a sense quality. This definition well displays Carnap's ingenuity. He sets his sights on quality classes, a quality class is the class of all elementary experiences that represent a given quality. All the elementary experiences in a quality class are part-similar, by virtue of sharing the quality. Still the class will be narrower than a similarity circle, for the members of a similarity circle need not all share any one quality. But Carnap argues that the quality class will always be the common part of all the similarity circles that it overlaps by more than half. This works out, he argues, if experience is reasonably varied and random.
Similarity of qualities is based on part-similarity of elementary experiences (share one quality) ==> manages to define the five senses: sight, smell, taste, hearing, and touch. Each sense is the largest class of qualities that are connected to each other by similarity. . . . each of the five senses can be singled out by dimensionality; for, Carnap points out, each has a different number of dimensions. Sight for instance, has five: the two spatial dimensions and the three dimensions of color. Dimension itself is definable mathematically, hence ultimately by logic and membership. . . . [As a first step in constructing a full system of physical reality from this sensory foundation, he envisaged] a projection of visual qualities of our two-dimensional visual field into three-dimensional space. Imagine lines projected outward from all points in the rounded front of the subject's eye. Each line is perpendicular to its immediate neighbourhood on that rounded surface. Each is the subject's line of sight from that point on the eye, and thus corresponds to a point in the subject's visual field. The colour of that point in the visual field is to be assigned to a point on that line of sight, out in three-dimensional space. Colors are to be assigned in this way to all lines of sight, one color to each. The remaining question is, how far out on each line? [p.12] These distances are to be adjusted, Carnap answers, so as to minimize the variegation of colors in the resulting three-dimensional space. And not only that, we must try also to minimize or retard the variegation of color over time at each point, thus going back and readjusting earlier assignments in the light of later input. In short, we are to so assign colors to points in space-time as to make for the drabbest and slowest possible world. A law of least action gets built into our very standard of what to count as ral. This was a deep insight of Carnap's. It is a stick-figure caricature of what the scientist actually does, early and late, in devising theories. It is the scientist's quest of the simplest solution. [p.13] For the subsequent construction of the physical world, one could not hope to proceed purely by definition; for minimization requires us to go back and reconsider past spatial allocations of qualities in the light of later ones. [Why not a 4D incremental algorithm?]
The idea of a self-sufficient sensory language as a foundation for science loses its lustre when we reflect that systematization of our sensory intake is the very business that science itself is engaged in. The memories that link our past experiences with present ones and induce our expectations are themselves mostly memories not of sensory intake but of things and events in the physical world. This led Otto Neurath, Carnap's colleague in Vienna, to persuade Carnap to give up his methodological phenomenalism in favour of physicalism.
[Carnap's "rational reconstruction" embraces] a Cartesian dualism of mind and body, if not indeed a mentalistic monism. Physicalism, on the other hand, is materialism, bluntly monistic except for the abstract objects of mathematics. [Carnap refused to pursue] the pragmatic alternative of a physically based constitution system for science. He refused for decades to permit an English edition of the Aufbau. [Quine sets out to pursue that physicalist alternative.] [16] Two directions suggest themselves. One, aimed solely at conceptual economy and clarity is pretty much what physicists at the theoretical pole have long been up to, though the logician might lend a welcome hand in the trimming and polishing before the final miniaturized model is cast in bronze. The other direction, more analogous to Carnap's Aufbau, is what I think of as NATURALISM. It is rational reconstruction of the individual's and/or the race's actual acquisition of a responsible theory of the external world. [Q. What is "responsible"? ] It would address the question how we, physical denizens of the physical world, can have projected our scientific theory of that whole world from our meager contacts with it: from the mere impacts of rays and particles on our surfaces and a few odds and ends such as the strain of walking uphill. [Q. What of deliberately constructed apparatus / experiments leading to observations in science] GLOBAL STIMULUS: Carnap's ground elements were his elementary experiences; each was the subject's total sensory experience during some moment, or specious present. Global Stimulus is its physical analogue, the class of all sensory receptors triggered at that moment, or, better, the temporally ordered class of receptors triggered during that specious present. [But what are the bounds in time for this "temporally ordering", or for the "specious present"?] "let me drop the memory factor here" [But then, how to compare temporally separated stimuli? How to abstract incrementally from stimuli?]
[Quine builds up the distinction between external phenomena and the signals at the nerve endings, called receptors. Perceptual similarity may not result in receptual similarity, but can be measured by looking at the sensory-action map. If similar receptory signals A and B result in opposite rewards, then whether C is more similar to A or B can be tested by seeing if the reaction to C is that of A or that of B.] RECEPTUAL SIMILARITY: Two global stimuli are more or less similar according as they comprise more or less the same nerve endings in the same order. PERCEPTUAL SIMILARITY: A and B are receptually similar signals, but the same action in A gets a reward and in B a penalty. Now when encountering stimulus C, if the reaction is the same action as in A it means C is more similar to A than to B. [18] (perceptual similarity is encoded by models of memory - see Missing Moment: How the Unconscious Shapes Modern Science, chapter 3.) [May be asymmetric due to the temporal recency of the experience of B]. EXAMPLE: Seeing a rectangle from different perspectives - different receptual similarity but high perceptual sim - "intrinsic propensity to associate perspectives" [Seems unlikely?] [19] SALIENCE: Global stimuli are excessive; need to focus on only the salient ones. May be receptually dissimlar but perc similar. Salience refers to those receptors within a g.s. that are shared with other perc similar g.s.. Thus, g.s. are perceptually similar by virtue of the shared salience. The inversion [??] is in the spirit of Carnap's constructions. Salience is the operative factor for "ostensive" definition. [ostensive = manifest; etym from L. ostendere, to show. WN: manifestly demonstrative? ostension = seeing?] [NOTE: Idea of salience is similar to what Langacker has called symbolic unit - how the semantic unit is related to a phonological unit. The semantic unit requires reification - coming up next. But is it also possible without reification? "Cold", "Thunder"? ] I have come this far in my physical mimicry of phenomenalistic epistemology without invading the percipient organisms more deeply than his sensory receptors. All that has mattered is the individual's distinctive responses to their activation. I shall not have to invade him more deeply, despite the illuminating progress of psychoneurology.
[Apes, birds, humans share information through cries or calls] Apes have a repertoire of distinct signals for distinct purposes. One signal warns the fellow apes of the approach of a lion or leopard, another warns of an eagle, another perhaps the sighting of some fruit trees while the troop is ranging through the forest. Each member of the tribe is disposed to broadcast the appropriate signal on experiencing a certain range of perceptually similar global stimulations, and to react with appropriate motor behaviour on hearing the signal. [ Thomas Struhsaker reported in 1967 that vervets gave distinct alarm calls in response to spotting three predators (leopards, eagles, and snakes), and the listeners would appropriately in each case. The calls appear to function as "representational, or semantic, signals" (125). "When one vervet hears another give an eagle alarm call, the listener responds as if it had seen the eagle itself. This behavior suggests that in the monkey's mind the call "stands for" or "conjures up images of" an avian predator even when the monkey has not yet see the eagle" (125). In the language of memory, Seyfarth and Cheney (1992) is making the proposition that vervets have evolved explicit categorical (semantic) recall ... But [the main difference with humans is that] they do not seem to recognize mental states in others. Seyfarth and Cheney, Scientific American, December 1992. ]
There is a puzzle here. Global stimuli are private: each is a temporally ordered set of one indivisual's receptors. Their perceptual similarity, in part innate and in part molded by experience, is private as well. Whence then this coordination of behaviours across the tribe? It requires that if two individuals jointly witness one scene, and subsequently jointly witness another scene, and the one witness's g.s's on two occasions qualify for him as perceptually similar, usually the other witness's gs's will also qualify thus. [20] So we see a pre-established harmony of perceptual similarity standards. ... This public harmony of private standards is accounted for by natural selection. The indiv's standards of percep similarity are inculcated by natural selection and due to shared ancestry and shared environment, will tend to harmonize across the tribe. There is also the discipline imposed by the vocal signals themselves, and later, language: random deviations among indiv's get canceled out by their hearing the signals from each other. [21]
A. Utterances reflect reified concepts - and may result in reification across the language group. B. Phonological standardization occurs when utterances are deemed non-meaningful (or same as other signs, or overly distinct) by others, focusing attention on the phonemic boundaries. C. Even if the signals change as a whole in the group (due to social power structures etc, these changes would also propagate and stabilize D. Perceptual similarity refers to the stimulus-action map. This requires that signals reflect the same USAGE patterns. ]
human counterparts of bird-calls and apes' cries... "It's raining", "It's cold", "Dog!" They are occasion sentences - true on some occasions. They report intersubjectively observable situations, observable outright. The pertinent language community is a parameter that we may take more broadly or narrowly according to the purpose of our study. VAGUENESS: What does "outright" mean? how many seconds of reflection? Take the "Swan!" or "Lo, a swan': would our speaker affirm it of a black one? [Australian swans are black] - gradations... Within the individual the observation is keyed to a range of perc similar g.stimuli - it is due to the pre-established harmony that they qualify as observation sentences across the community. Observation sentences are not only the prehuman counterparts of language, but they are also its inception as with the recruitment of each new child into the language community. Some o.s's such as "Milk!" or "Dog!" are what we in our sophistication see as terms denoting things, but to the child, innocent at first of any thought of reification and reference, they are on par with "It's cold" and "It's raining": just things to say in distinctive circumstances. [22] [NOTE: This is an assumption reiterated through the book - but psychological data on rigid objects would appear to indicate that perceptually, there is evidence for reification well before observational sentences appear. The low probability of coherent visual flow associated with coherent objects make them perceptually similar much before they are linguistically formed.] The child is innately more agile at learning new observation sentences by ostension, and beyond this soon learns connectives by which to compound new observation sentences from old ones. 'Not' and 'and' come to mind. A negative obsvn s, 'Not p' is probably learned when the child's mentor utters it as a correction to the child's utterance of obsvn s 'p'. The connective 'and' is painlessly acquired amounting as it does to the merest punctuation between affirmations. [All this can be tricky. Consider: "Walk out of that crib again and I will tie up your hands." Necessary to assume that the caretaker is deliberately using a simpler form of language, reflecting more standard usages of these connectives, and hence that the caretaker is aware of the centers for these semantic prototypes. ] On learning 'not' and 'and' the child internalizes a bit of logic, for to affirm 'p and not p' is to have mislearned one or both particles. The child masters further connectives : above, beneath, before, after, in, beside [NOTE: all are spatial. IDEA**: USE next quottation in Spatial learning work. ] Perhaps the child learns such a connective by learning a compound obsvn sentence outright as a whole by direct ostension. Then having learned also each of the components independently, he catches on to the trick and proceeds to apply the connective by analogy to other pairs. [24]
One of these primitive grammatical constructions in particular si significant as a first step toward reification of bodies. Applied to simple observation sentences 'Black' or 'that's black' or 'Dog', it forms the compound 'black dog' or 'that dog is black' - as an observation sentence the compound can be phrased either way, since terms are not yet recognized as denoting. I call this primitive grammatical construction "observational predication", in anticipation of the mature predication into which it will evolve. Observational Predication is quite a different matter from mere conjunction. The conjunction describes any scene in which black and dog are salient, whereas 'black dog' requires that they be situated together, the black patch engulfing the canine patch. The predication expresses the compact clustering of visual qualities that is characteristic of a body. Bodies are our first reifications: the first objects to be taken as objects. It is in analogy to them that all further positing of objects takes place... typically it contrasts with its visual surroundings in color and in movement or parallax, and typically it is fairly chunky and compact. If we make contact, it resists pressure. These traits distinguish it from [mere sensations] - the colour of the evening sky, the feel of a cool breeze, [the sound of an aria] or other details of the passing show. The mode of compounding observation sentences that I have called predication, then, is a step towards reification of bodies, in its stress on spatial clustering. However, I hold that at this point reification is not yet achieved. [NOTE: THIS last may prove to be incorrect. Given the perceptual salience of a coherently moving black blob, which subsequently is identified with "dog", the co-visuality of black and dog would be more salient and therefore more likely to be the interpretation of "black dog" even if the utterance were "black and dog". ]
Meanwhile I turn to what I picture as the first step beyond ordinary observation sentences: namely a generalize expression of expectation. It is a way of joining two obsvn s's A and B to express the general expectation that whenever A holds, B will be fulfilled as well. "When it snows, it's cold"; "Where there is smoke, there's fire", "When the sun rises, birds sing". The leap from observation sentences to observation categoricals to is a giant one, for o categoricals are the direct expression of inductive expectation, which underlies all learning. O c's bring us vicarious habituation, vicarious induction. One gets the benefit of generalized expectations built up over the years by some veteran observer or even by that veteran's own informant long dead. Observation categoricals can be handed down. 25 Even at this stage there is no denotation, no reference. The child's observation sentences 'Mama' and 'Doggy; register repeatable features of the passing show, on a par with 'Cold' and 'Thunder'. Nevertheless we have a [modest] sketch of a causal chain from the imprint of rays and particles on our receptors to a rudimentary theory of the external world. Observation categoricals are indeed a [miniature] scientific theory of the world, complete with empirical checkpoints subject to the experimental method. ... by waiting for an occasion where the first component of categorical is fulfilled, or even by bringing about its fulfillment, and watching ... [26]
to contribute some excerpts from your favourite book to
book
excerptise. send us a plain text file with
page-numbered extracts from your favourite book. You can preface your
extracts with a short review.
email to (bookexcerptise [at] gmail [dot] com).
We reply to all feedback!