Let’s call verificationism the thesis that the meaning of any sentence expressing a synthetic proposition is just its method of empirical verification. It follows that a statement is meaningless if it has no method of empirical verification, and if neither it nor its negation is an analytic truth. (Obviously, this characterization of verificationism could be fleshed out, and most if not all verificationists would refine it in some way.) Empirical verifiability (or falsifiability – I’ll henceforth refer to the former term for ease of explication) is a matter of a proposition’s having some sort of probabilistic or deductive relationship with observable states of affairs. For example, a statement regarding the existence or behavior of an unobservable entity can still be empirically verifiable if it entails or has some sort of probabilistic relationship with observable states of the world (see, for example, Ayer 1952: 13, 38; Carnap 1936: 425-427). Thus, a statement about, say, an atom, can be empirically verifiable if it is evidentially related to observation sentences – things like “under such-and-such conditions the cloud chamber will look like this,” or “under such-and-such conditions the screen connected to a scanning tunneling microscope will look like this.” Alright, here’s my question: what does the verificationist mean by “observable” and “unobservable”? Or, more specifically, because these are the terms the verificationist is more likely to use, what does the verificationist mean by “possibly observable” and “possibly unobservable”?
Maybe the verificationist will say at this point that what we should mean by “possibly observable” is “nomologically possibly observable” – possibly observable, given the actual laws of nature (this view is maintained by many positivists, e.g., Carnap 1936: 423). Here is the problem with this response. Galileo is said to have dropped two cannon balls of different weights from the leaning tower of Pisa to see if one fell faster than the other (they both fell at the same rate). But shouldn’t the verificationist concede that the following statement is meaningful? “Galileo dropped the cannon balls and the heavier one fell faster than the lighter one.” But apparently the verificationist can’t say this statement is meaningful, because it does not represent a nomologically possible observable state of affairs.
Sources:
Ayer, A. J. Language, Truth, and Logic. 2nd ed. New York: Dover Publications, 1952. Print.
Carnap, Rudolf. “Testability and Meaning.” Philosophy of Science 3.4 (1936): 419-71. Print.
Great question. A great deal of theoretical physics and mathematics is called into question by it.
Andrew,
Re “But shouldn’t the verificationist concede that the following statement is meaningful? ‘Galileo dropped the cannon balls and the heavier one fell faster than the lighter one.’”
The verificationist who understands possibly observable as ‘nomologically possibly observable’ could respond in a few ways, the most promising of which would be the following:
If a proposition p is nomologically impossible, it does not follow that ‘p is confirmed’ is likewise nomologically impossible, because ‘p is confirmed’ does not deductively entail ‘p is true’. Thus, a probabilistic confirmation of p would not require its truth. It would be sufficient to make observations that contradict the empirical facts on which p is inductively based. The logical lesson here is that if p contradicts a physical law and some observational evidence that confirms p, it does not follow that the evidence contradicts the physical law because the evidence may confirm p without p being a necessary consequence of the evidence. Therefore, verificationists who explicate possibly observable in terms of nomological possibility are not committed to the consequence that the negation of a proposition expressing some physical law is meaningless.
Re “verificationism [is] the thesis that the meaning of any sentence expressing a synthetic proposition is just its method of empirical verification.”
If defined in this way, the verifiability theory of meaningfulness presents particular problems. In particular, the formulation seems to confuse the meaning of a proposition with the evidence that would confirm it. Presumably, one cannot gather evidence for a proposition if one does not already know what would constitute as evidence for a proposition. However, most verificationists who care to be careful enough do not identify the meaning of a proposition with the states of affairs that would constitute indirect evidence for the proposition.
Re “Observable”
Why cannot the verificationist say that to observe some object is to receive sensory stimuli or to have some physical interaction via our sensory organs? This would permit the verificationist to explicate uncontroversially what it means for some term to obtain cognitive significance. The story would go something this:
A descriptive term is cognitively significant iff it is either ostensively definable- some particular physical object or some physical relation or process- or reducible to ostensively defined terms. This leads to the principle that philosophers and theologians forget because they are all too absorbed in linguistic abstractions and lose sight of the fundamental basis of sensory experience from which all significant language originates and to which it must inevitably return. One may define one word by means of another, but until one comes to words whose meanings are explicated by exhibition or ostentation, no word possesses semantical meaning. One may define an Euclidean isosceles triangle as a three-sided geometrical shape comprised of three straight lines, two of which are (at least) equal in length, and two equal interior angles, but if one does not further explain what an ‘angle’ and ‘straight line’ are via indicating examples, one will fail to define ‘isosceles triangle’.
Hi Aaron,
You’ll see below that I’ve retracted my argument because of Clayton’s response. However, the argument in your first long
argumentparagraph seems like an adequate response as well.For what it’s worth, I don’t see much reason to think that “One may define one word by means of another, but until one comes to words whose meanings are explicated by exhibition or ostentation, no word possesses semantical meaning,” at least where “exhibition” or “ostention” only operates with the aid of normal sensory experience. Counterexamples (which, of course, I don’t really expect you to regard as good counterexamples): moral intuition, religious experience, innate concepts (or, more accurately, those innate dispositions to form certain concepts that are studied by cognitive scientists). Furthermore, even if all of our words are ultimately learned through, say, ostention, we can edit these concepts and recombine them to form meaningful predicates that nevertheless do not have a method of empirical verification. In any case, those sentences that verificationism regards as meaningless are often, since they are clearly meaningful, simply counterexamples to verificationism. Don’t feel obligated to give a lengthy response to these points, as it’s sort of off topic in the comments to this post. But, since you offered a brief justification for verificationism, I felt I should offer a brief (and, admittedly, vague) rebuttal.
Andrew,
I take your brief reply and I will not offer a lengthy response. I will, though, say a few things. First, the so-called counterexamples to verificationism often (always?) beg the question at hand. Second, moral intuition, religious experience, et al., again, merely beg the question at hand. Third, why do you not find such utterances as ‘god’, ‘spirit’, ‘soul’, and similar terms problematic? You and I can agree on how to define and give properly empirical import to almost all of our terms, so why the glaring lacuna here? I assert that you and your theologically-inclined fellow philosophers have been caught in two millennia of superstitious verbal abstractions.
Andrew,
I don’t mean to say that philosophically-mature theists such as yourself are believe in a ‘god’ because they are superstitious. I meant only that the belief in supernatural agencies has part of its historical origin in superstitions, which is not to impugn theistic beliefs today per se, of course.
Hi Andrew,
Can the verificationist make some headway in dealing with your problem by saying this:
For all p, if p is empirically verifiable, both p and ~p are cognitively significant.
If so, the false proposition you mention about one ball falling faster than the other counts as meaningful because its negation counts as meaningful. This seems like a natural move for the verificationist since it also seems natural for them to say this:
For all p, if ~p is not cognitively significant, ~~p is not cognitively significant.
You wouldn’t want: ~~ (The balls fell at different speeds) to come out as meaningless because that entails that the balls fell at the same speeds. Surely, ~~p is meaningful iff p is.
Dang, I think you’re right. I retract my argument. Thanks!
I think the principle of verification is a criterion of meaning of a statement which can be compared to a hurdle for a race. If any statement has the capabilities to come across this hurdle it is considered to be meaningful. But the case is not so much easy as it is thought to be. Possibly all the metaphysical utterances fail to overcome this hurdle and thereby it is meaningless.