Is there a genuine distinction between observable and unobservable entities? Why does it matter? How, and why, might one distinguish between theoretical and observational statements in science? I have decided to tackle both these questions because they feed into and relate to one another. They emphasize different aspects of a prevalent debate, all aspects of which I wish to touch on.
Whether the question of a distinction between observable vs unobservable entities is synonymous to the question of a distinction between theoretical vs non-theoretical statements is itself a matter of debate.Quine advocates semantic ascent, the shift in which the language we use to refer to the world becomes something we talk about in its own right. Semantic ascent is a shift from questions about objects to questions about words or statements. He says we should ‘drop the talk of observation and talk instead of observation sentences, the sentences that are said to report observations’ (The roots of Reference). So obviously Quine thinks the two questions are equivalent.
They have often been treated as equivalent questions, or at least not distinguished too carefully.I agree with Van Fraassen that we should at least note and respect the differences between the two ways of talking about what might be the same issue, and not make the category mistake of talking about theoretical entities, just for clarities sake. At any event Paul M Churchland disagrees with Quine that the two debates are parallel , He says “we agree (Churchland and Van Fraassen) that the observable/unobservable distinction is entirely distinct from the nontheoretical/theoretical distinction”.This disagreement / confusion as to the very terrain, layout of the questions of the debate, arises because there is the ordinary language question of how do we naturally apply the terms ‘observed’ and ‘observation’, as well as the question of whether a principled O/T distinction can or should be drawn; as Gerry Fodor’s Granny says: “True there is an epistemologically important distinction, that it’s reasonable to call ‘the’ observation inference distinction, and that is theory relative.And, also true, it is this theory-relative distinction that scientists usually use the terms ’observed’ and ‘inferred’ to mark. But that is quite compatible with there being another distinction, which it is also reasonable to call ‘the’ observation /inference distinction which is also of central significance to the philosophy of science, and which is not theory relative. ” It is this second principled O/T distinction that I will focus on as opposed to the ordinary language distinction, I do not think ordinary language arguments bear on the question of whether there is or should be a principled distinction.Although examining what inclines us one way or another in ordinary language usage may clarify factors that also influence us in an overall distinction, such as naturalness, entrenchment, flexibility and plasticity.
After semantic ascent the question of whether there is an O/T dichotomy becomes one of whether all observation reports presuppose some theory. This slightly ignores the question of the ontological status of the entities, whether observed or unobserved, but this will come up when I tackle the subsidiary part of each question the “why make a distinction, for what purpose? ”or “why does it matter if a distinction presents itself? I think the strategy of semantic ascent is useful and justified since the debate takes place in at least two domains, the perceptual/cognitive (internal) and the observational/inferential (public)“The strategy of semantic ascent is that it carries the discussion into a domain where both parties are better agreed on the objects (viz. , words) and on the main terms connecting them. Words, or their inscriptions, unlike points, miles, classes and the rest, are tangible objects of the size so popular in the marketplace, where men of unlike conceptual schemes communicate at their best.The strategy is one of ascending to a common part of two fundamentally disparate conceptual schemes, the better to discuss the disparate foundations. No wonder it helps in philosophy.
” Quine word and object. But it is a bit confusing and difficult to translate debates or points between the two, and certain debates are clearer at the ground level rather than the meta-level. There are three classes of arguments that bear on the T/O distinction: 1. Meaning holism arguments. Which tend to work against the distinction 2.Ordinary language arguments.
Which tend to work for the distinction 3. Psychological arguments. Which can work for or against As well as a specific argument by Grover Maxwell from the continuity of observation with inference which works against the T/O distinction. There are two extant modes for making the theory observation distinction – Fodor’s and Van Fraassen’s. Fodor defends the distinction against the implication from cognitive science that perception is continuous with cognition. VanFraassen defends the distinction against Maxwell’s challenge that it is impossible to draw the line between what is observable and what is only detectable in some more roundabout way. Fodor and Van Fraassen have different reasons for drawing a distinction, Fodor, to defend realism, Van Fraassen to attack realism, strangely enough. Fodor to defend realism against Kuhnian relativism, and Van Fraassen to defend constructive empiricism, a form of anti-realism, against incoherence, and so pit it against realism.
As Andre Kukla notes “It is not surprising that a realist and an anti-realist should agree on something; but it is curious that van Fraassens and Fodor’s defenses of the theory-observation distinction play diametrically opposite roles in their philosophical agenda’s. ” Andre Kukla the theory observation distinction. But should we be driven by a philosophical agenda in debating a question? Or should we resolve the question and then decide on a position which accords with our answer? Shouldn’t we be neutral when we make philosophical decisions?Unfortunately in philosophy there is so little ‘evidence’ making up your mind is more a matter of achieving coherence, it is legitimate to allow justification to flow in all directions.
The question of whether there is a T/O distinction is relevant to the debate between realists anti-realists and relativists in the following manner. So far as realists debate with anti-realists is concerned, the T/O distinction is optional for realists. They have everything to gain and nothing to lose by making it unravel.They have everything to gain, because the constructive empiricist position is incoherent without a T/O distinction. But so far as realists debate with relativists goes, realists have conversely everything to gain and nothing to lose by defending a distinction, they would defeat relativists.
Kuhnain Relativism requires the lack of a theory neutral language with which to adjudicate our differences, so we get incommensurability, incommensurability leads to the irrationality of theory choice thus we get relativism. But realists cannot have an easy victory against both parties.I suggest that the realist denies the T/O distinction and so wins against the constructive empiricist. The lack of a T/O distinction does not entail relativism; a theory laden observation can still test a theory. To return to the question of whether we should be driven by a philosophical agenda in deciding a point, it must be remembered that we are concerning ourself with the question of whether there is a significant or principled O/T distinction. Its significance comes from its position within a larger debate.Frankly, everyone can admit there is some sort of distinction or difference between direct and indirect observation, the question really is how significant the difference is, whether a distinction can be drawn at a position significant enough to support any theory, the significance depends on the work it is made to do by larger theories. Paul M.
churchland defines his scientific realism as a realism entirely in terms of his attitude towards the T/O distinction. He believes any attempt to draw the distinction, particularly Van Fraassen’s, is arbitrary.By any skepticism “our observational ontology is rendered exactly as dubious as our non-observational ontology” He is not an orthodox scientific realist; he is skeptical about the overall truth of our beliefs, the reference of scientific terms, and the convergence of theory towards truth. But he is skeptical about the success of all our theories, cognition at large, from a low to a high level not just scientific theories, and thus does not distinguish between the integrity of observables and the integrity of unobservables. He states that “global excellence of theory is the ultimate measure of truth and ontology at all levels of cognition”.
Although churchland has exactly the same attitude to observables and unobservables, a cautious skeptical attitude, relative to his peers he has a slightly pro attitude to unobsevables, and a negative attitude to observables. This pro – attitude to the unobservables of science makes him a realist and his slightly negative attitude to the observables of everyday life make him a scientific realist “the function of science, therefore, is to provide us with a superior and (in the long run) perhaps profoundly different conception of the world, even at the perceptual level”.I agree with Churchland as to the theoretical character of perceptual judgments, I agree that “perception consists in the conceptual exploitation of the natural information contained in our sensations or sensory states” . Having done part of a module on “the brain as a statistician” I know that our perceptual judgments are statistical decision problems akin to gambling or any decision based on uncertain evidence.Because inputs are noisy – the external world and inefficient transduction creates noise- the question of whether a signal is present or not will reflect the relative probability that a signal is drawn from distribution A(noise only) or distribution B(signal + noise).
Biasing factors are the probability of occurrence of a member of each category, information on which is drawn from memory. Perceptual decisions rely on perception and memory, or evidence and prior knowledge, prior knowledge being essentially a theory about the world.However I disagree with Paul Churchland as to the possibility of our being trained to make systematic perceptual judgments in terms of theories other than the common sense theory we ‘learnt at our mothers knee’ For one, I don’t think we learn our common sense theory rather it is built into our genetics. I do not think we are nearly as plastic as he makes out, on this point I go with Gerry Fodor, perception and cognition are not continuous, and perception can never make judgments in terms of grand theories which we can barely conceive. The boundary between what can be observed and what must be inferred is largely determined by fixed architectural features of an organisms sensory / perceptual psychology” Gerry Fodor Observation Reconsidered. Paul Churchland directly contradicts this saying “our current modes of conceptual exploitation (perception) are rooted, in substantial measure, not in the nature of our perceptual environment, nor in the innate features of our psychology, but rather in the structure and content of our common language”.How plastic the brain may be is an empirical point, and I think Gerry Fodor wins the debate with his analysis of the muller-lyer illusion.
Fodor says the robustness of the muller lyer illusion attests to the imperviousness of perception by cognition. There are both perceptual plasiticities and implasticities. Kuhn was impressed by the plasticities, but it is time to dwell more on the implasticities. “To the best of my knowledge, all the standard perceptual illusions exhibit this curious refractory character :knowing they are illusions doesn’t make them go away” However I don’t think Fodor is being entirely empirically accurate.Some illusions such as the concave – convex illusion, in which heavily shaded circles appear as concave when the shadow is at the top of the circle, and convex when the shadow is at the bottom of the circle, which occurs because we have a strong prior belief / prior assumption that light falls from above, can be reversed or at least nullified if you really try. The famous duck rabbit can definitely be flipped at will. And the old hag, young girl illusion, personally I can never see the old hag unless it is explained to e, then I can. But anyway Fodor makes his point, we cannot always see just what we want to see or think we should see.
I agree with Gerry Fodor that perception is fairly modular, and is not (probably) affected (much) by conscious explicit knowledge. Certainly the muller lyer illusion is fairly robust And I think far too much is made of the duck rabbit illusion – Kuhn says “it is as elementary prototypes for these transformations of the scientists world view that the familiar demonstrations of a switch in gestalt prove so suggestive”.But I do not think they are anything more than just that – suggestive – because a scientist, does not, cannot form an image or representation of quarks and leptons in any way analogous to a duck or a rabbit, so this image cannot ‘flip’. Paul Churchland seems to think we can form such images, but personally I cannot. I see the western sky redden as the sun sets not “the wavelength distribution of incoming solar radiation shift towards the longer wavelengths”.
However I would say our inability to alter our perception does not damage churchland’s essential point which was that perception relies on theory, implicit theory. A very entrenched embedded theory, but theory all the same. Churchland thinks the distinction between the theoretical and the non-theoretical is just a distinction between freshly minted theory and thoroughly thumb-worn theory whose ‘cultural assimilation is complete’.I think some ‘thumb-worn theory’ is actually entrenched in our biology. But maybe individual differences come into play here, maybe some people are more plastic than others, or innately sensitive to some aspects of reality than others, maybe our biology is not universal. Paul Churchland says that the person with perfect pitch is not a physiological freak but a practiced observer.
But I think it most likely that there is something unique about them.Maybe I am closed minded in the sort of visualization Churchland encourages, maybe that’s just me, I had no luck with seeing in the fourth dimension even after reading ‘Flatland’ and ‘speculations on the fourth dimension’ whereas other people (the authors) claim to have, still I’m a bit skeptical. Paul Churland’s thought experiments where he gets us to imagine various other beings, with radically different physiology, beings that can visually see infra-red heat for example, raises the idea of the possibility of other sensory modalities.And although we cannot communicate with them, so they are not part of our epistemic community, there are animals on our planet who presumably sense different things to us, such as bats and dolphins. Van Fraassen insists that is ‘observable’ must be observable to us unassisted, and as we currently are, an anthropocentric conception; “the limitations to which the ‘able’ in observable refers are our limitations qua human beings. ” It could be argued that Van Fraassens anthropocentric conception of the observable is not just anthropocentric, but parochial.Alternatively it could be argued that van fraassen draws the line arbitrarily: according to Van Fraassen we can observe planets using a telescope, but we cannot observe viruses using a microscope, because planets are something we could observe without any augmentation of the senses, where we close enough to them, and indeed some of them we can observe from earth, our natural position, (venus) whereas under normal conditions viruses cannot be seen.
I do not agree with this objection to Van Fraassen, I think where he draws the line is one natural place to draw it if it has to be drawn, but it is just that I don’t agree with him that the drawing of the line here is very significant. I am a realist and I believe unobservables are generally as real as observables. From his drawing of the line, van Fraassen only believes in observables.Fodor lightly passes over the fact that “perceptual analyses are undetermined by sensory arrays” and are only resolved by Bayesian reasoning from previous evidence / experience, and that “the appeal to background theory is inherent to the process of perceptual analysis” Fodor Observaiton reconsidered. I think this fact is indisputable, and it is in this respect that perception and cognition are similar as Paul churchland maintains, both are theories and “global excellence of theory is the ultimate measure of truth and ontology at all levels of cognition” .
The impossibility of our being trained to make systematic perceptual judgments in terms of theories other that the common sense theory we ‘learnt at our mothers knee’, the implasticity of actual human perception, is irrelevant in drawing a theory observation distinction, both perception and cognition are theory dependent. But Granted as Fodor points out against Kuhn scientific knowledge doesn’t actually percolate down to affect the perceptual. Kuhnian perceptual theory loading does not occur.
There is some natural barrier. Is this barrier the location of the O/T distinction? I think it probably is if there is one.It is significant, but not significant for the anti-realist, it does not decide our ontology. It is significant in the realists fight with relativism since observations are theory laden, but are not necessarily laden with the high level theories that they must adjudicate between.
So perceptions are laden with perceptual theory, but not laden with quantum theory. Fodor makes the O/T distinction in such a way that it is significant for realists against relativism, but not significant for anti-realists. “Fodor isn’t looking for a notion of observationality that underwrites our granting epistemic privilege to observation statements.He’s looking for a notion that will ward off the incommensurability arguments. And for that purpose anything that produces consensus will do” Andre Kukla The theory observation distinction. Now to explicitly tackle the questions, “why make a distinction, for what purpose? ”or “why does it matter if a distinction does or does not present itself? ”.
I have already touched on the answer to these questions when outlining the role of the distinction (or lack of) in larger debates between anti-realists, and relativists.The question of the O/T distinction has epistemological significance; it concerns the epistemic bearing of observational evidence on theories it is used to evaluate. This is part of the debate between realists and relativists. The relativists holding that observation is an inadequate basis for choosing between rival theories, the realists claiming it is an adequate basis, or there is at least something which is an adequate basis.
Observational evidence also plays important and philosophically interesting roles in other areas including scientific discovery and the application of scientific theories to practical problems.But we will concentrate on theory testing. It seems that if all observations are theory laden then there is no objective bedrock against which to test and justify theory. The classic or common view of science is that scientific knowledge is derived from the ‘facts’ or observations. Two schools of thought that involve attempts to formalize this common view of science are the empiricists and the positivists. An extreme interpretation of the claim that science is derived from the facts implies that the facts must first be established, and subsequently a theory built to fit them.This is the baconian method building a case from the ground up.
This is not how science actually proceeds. “our search for relevant facts needs to be guided by our current state of knowledge, which tells us for example that measuring the ozone concentration at various locations in the atmosphere yields relevant facts whereas measuring the average hair length of the youths in Sydney does not” A F chalmers What is this thing called Science?. But the fact that science is guided by paradigms does not support kuhnian relativism.Kuhnian relativism can only be established if incommensurability is, that is if high level theory-loading of observation were established. As I have already argued along with Fodor, observation may be loaded with low level perceptual theory but not with high level conscious and elaborate theory. Proponents of competing theories often produce impressively similar observational data, this indicates perceptual theory loading is not that great. If science were blinded by paradigms that would be a different matter. Against semantic theory loading; Often observations reported non-linguistically, pictorially with tables of numbers etc.
Late 20th century philosophers may have exaggerated the influence of semantic loading because they thought of theory testing in terms of inferential relations between observational and theoretical sentences. Against Salience or attentional loading – scientists under different paradigms attend to different things. Yes, but doesn’t always happen. And scientists may appreciate the significance of data that is brought to their attention that had not been noticed. Attentional loading is not inevitable and not irredeemable.
So observation is and adequate basis for adjudicating between theories (unless the theories are underdetermined by data).In conclusion I would say there is no absolute T/O distinction, but there is enough of a difference, enough bottom up flow of justification, to defeat relativism. A. F.
chalmers: what is this thing called science? Paul M Churchland: Scientific realism and the plasticity of the mind Paul M churchland: The ontological status of obsservables: In praise of superempirical virtues Gerry Fodor: observation reconsidered Andre Kukla: the theory observation distinction W. V. O Quine: Word and Object Bas Van Fraassen: the scientific image