I
Few philosophical movements have arisen with such bold
pretensions and ambitious hopes as those which attended the rise of
Logical Positivism. Here, it was claimed, were to be found the
insights that would revolutionize philosophy. No longer would
philosophy dispute, or be remote from the work of science. The
significant work of the human intellect could now be conveniently
classified into throe general areas: the search for empirical
("cognitive") knowledge, which is science; the logical
structuring and semantic clarification of science, which is philosophy, and the stimulation and satisfaction of human
perceptions and emotions, which are the arts. No longer need
time and talent be wasted in the interminable and meaningless
harangues of classical philosophy, or the futile search for "deeper"
yet cognitive meaning in the arts. Through the positivistic
insights, we might tidy up the work of intellect by separating the
cognitive from the emotive and the meaningful from the meaningless.
It was to be the work of the philosopher -- which is to say the
positivist -- to maintain the order and harmony in the intellectual
establishment, through the explication and application of
indisputable rules of logic, norms of language and criteria of
meaning.But such a schema was not to be established. It could not. It was
a plan for the gods, and not for scientists, philosophers, artists
or positivists. It was a regime believed infallible, since it was
thought to be based, not upon empirical theories (which are
fallible), but upon explicitly stipulated decisions and rules of
logic and or language. Thus, since they supposed that they had, not a theory, but an activity of regulation and
clarification, the positivists also believed that they could
profitably guide the scientific search for facts without themselves
being open to factual refutation.
The logical positivists attempted to orient the meaning of
propositions to the world of experience. To this end, they
formulated several criteria of meaning, some quite strict, others
much more generous. Among them: (A) "A proposition is meaningful
only if it is verified." This view, which has virtually no
adherents, would, deny moaning to all hypotheses not confirmed, and
would acknowledge only proven laws and facts. All theoretical
investigation would accordingly be meaningless. It is clearly
unacceptable: We raise it only because it has boon falsely
attributed to many innocent positivists. (b) "The meaning of a
proposition is the method of its verification." [Schlick 146] But if
one is to verify a proposition, does this not indicate a clear
distinction between the task of verification and that which is
verified -- the proposition? Meaning is not a method: methods
serve meanings., (c) "A statement has meaning only if it is
verifiable." The crucial term hero is "verifiable," which is offered
as a tool, not a synonym, of meaningfulness. Various analysts take
the term to mean "verifiable" as a technical possibility, as an
empirical possibility not contrary to natural laws), or as a logical
possibility (not involving a tautology or a contradiction -- which
is the mast liberal interpretation) (d) "A statement has moaning
only if it is confirmable" (again, the moat generous interpretation
would insist only that confirmation be logically possible – a
concept that we will explore later in this paper). Here the
requirement of complete verification, even as a logical possibility,
is replaced by a "weaker" requirement of incomplete evidential
support.1
II. THE LIMITS OF MY LANGUAGE
It would be impractical and unwise to discuss all of the criteria
outlined above. We shall instead consider some representative (if
sometimes extreme) views of Moritz Schlick, the "hub" of the Vienna
Circle. Schlick defends three concepts in particular that we shall
find reason to question: (a) an absolute conception of meaning based
upon (b) "logical possibility" which rests upon (c) the "rules of
language." (The test of "logical possibility is "conceivability," as
we shall consider below). But let us hear from Schlick himself:
The possibility of verification does not rest on any
"experiential truth," on the law of nature or any other true
general proposition, but is determined solely by our
definitions, by the rules which we which have been
fixed in our language, or which we can pick
arbitrarily at any moment. All these rules ultimately going
to extensive [i.e., ostensive] definition... and
through them verifiability is linked to [sense data] ... no
rule of expression presupposes any law or regularity in the
world ... but it does present both data and situations, to
which names can be attached. The rules of language are rules of the application of language; so there must be
something to which it can be applied. [Schlick 157. My
italics. EP]....
The height of the tower cannot be 100 feet and 150 feet
at the same time; a child cannot be naked undressed at the
same time – not because we are unable to imagine it, but
because our definitions of height of the numerals, or the
terms "naked" and "dressed", are not compatible with the
particular combinations of those words in our examples.
"They are not compatible with such combinations" means that
the rules of our language have not provided any use for such
combinations; they do not describe any fact ... The result
of our considerations as this: verifiability which is the
sufficient and necessary condition of meaning, is a
possibility of the logical order; [my italics, ep]. It
is created by constructing the sentence in accordance with
the rules by which its terms are defined ... Grammatical
rules are not found anywhere in nature but are made by men
and are, in principle arbitrary; you cannot give meaning to
a sentence by discovering a method of verifying it,
but only by stipulating how it shall be done.
[Schlick .55]
...
[A] fictitious world may be empirically impossible,
because incompatible with the actual laws of nature – though
we cannot at all be sure of this – it is logically possible
[if we are able] to give a description of it. [Schlick .64]
...
The dividing line between logical possibility and
impossibility of verification is absolutely sharp and
distinct; there is no gradual transition between meaning
and nonsense. For either you have given to grammatical rules
for verification, or you have not...
Empirical possibility is determined by the laws of
nature, but meaning and verifiability are entirely
independent of them. Everything that I can describe or
define is logically possible – and definitions are in no way
bound up in natural laws. [Schlick .156-7].
Apparently, Schlick’s absolutism rests upon language, or, more
specifically, upon the "rules of language." But I fail to find here
any acknowledgment of the imperfections of human language
which, I submit, are painfully commonplace. By "rules of language"
he means, I presume, definitions and grammatical norms. But whose
definitions and whose norms? Where do we find these definitions? The
dictionary? And whose dictionary and what edition? Dictionaries are
compiled from surveys of actual usage – usage which is, of course,
broad and fluctuating. Consider this observation of Allen W. Read:
In the compilation of the Oxford English dictionary,
the intermediate or transitional quotations were discarded
as being "ambiguous" or "not clear," and the resulting
patterns are false to actual usage. The theoretical fluidity
of these features that are abstracted from a social setting
is hard for many people to admit, especially philosophers.
[!!!] Bloomfield has stated the principle that "every
utterance of a speech-form involves a minute semantic
innovation."2 [Read 41]
Has Schlick thus ignored the fact that language (with the
possible exception of purely formal language – i.e. logic) is
inevitably vague and ambiguous – at least to some extent.3
What word do we understand so well that we know it’s "real" meaning
– wholly and completely? We are familiar enough with those endless
arguments over what such-and-such word really means; but if
we have the beginnings of philosophical sophistication, we recognize
these to be verbal arguments, that is, arguments over arbitrary and
variable usages.
It may be replied at the positivist is concerned, not with the
definitions that are found in dictionaries, but in definitions that
are completely and explicitly stipulated for our use. These meanings
may be complete and agreed upon. But this is only an apparent
escape. Stipulated definitions are made in terms of conventional
words, or if not, the definitions of the words in the stipulation
use conventional meetings, etc. Somewhere, somehow, conventional
natural language must be involved in any stipulation. One can
stipulate as much as he pleases, but eventually he must come
down to ordinary usage. The so-called ultimate ostensive definitions
that some analysts believed to underlie all meaningful empirical
discourse are in fact ideal fictions that logicians (and logical
atomists in particular) like to talk about when they speak of their
"perfect" or "logical" languages. But the bases of meaning in the
natural languages (which, at least to a significant degree, underly
the languages of the sciences) is convention and, only remotely and
vaguely, ostensive definitions. As John Hospers puts it:
A definition is no stronger than the words in it, just as
the chain is no stronger than the weakest link. Every time
you think you have an airtight rule for "X," it can turn out
that the very constituents of the rule are not airtight
themselves; the plugs have been put in to fill up the gaps
have to be filled up themselves. This phenomenon is
sometimes called "the open texture of language."
This is just a way language is built. As long as words
are built upon the basis of previous imprecise words, and
those have heard of others, this phenomenon will continue.
[Hospers 49].
In sum, and contrary to Moritz Schlick, the "rules of language"
do in fact "presupposed laws or regularities in the world:" they are
the regularities of how people who in the world conventionally use
their languages, and the fact that no empirical term can be wholly
free of vagueness and ambiguity. Apparently the meanings of words
are related to those who use them, and the clarity of terms is
always less than complete. "The rules of language" are shaky
foundations indeed for an absolutism: an absolutism that Schlick
alluded to when, as quoted above, he stated that "the dividing line
between a logical possibility and impossibility of verification is
absolutely sharp and distinct."
II. "LOGICAL POSSIBILITY"
Schlick’s concept of "logical possibility," we recall rests upon
"the rules of language" which designates the meanings of words and
relationships among them in phrases and sentences. Since he believes
these "rules" to be explicit and complete, Schlick further holds
that propositional meanings are likewise absolute. But even if we
accepted Schlick’s belief that carefully stipulated meanings and
grammatical rules (specific "rules of language") are simple, plain
and concrete – and, of course, we do not – we must still reject an
extension of this absolutism to propositional meanings. Propositions
are not merely organized aggregations of meanings, and thus
subject to the same rules and composed of the same arbitrary
qualities, as words and grammar. On the contrary, propositions relate
meanings and thus refer to relationships between the
referents of the component word-meanings; that is to say, they refer
to relationships in the world. What empiricist would wish to suggest
that relationships in the world can be judged possible or impossible
(even "logically" so) simply by analyzing the relationships between
meanings in a proposition?4 Of course, some terms
are so related in meaning that they might involve a contradiction if
they both appeared in certain types of sentences (to use Schlick’s
example, "the child was naked, ut wore a long nightgown"). But while
formal rules of contradiction can be expressed logically, the
question of whether or not certain terms entail contradictions in
certain forms of sentences is a question of usage – a contingent
empirical question. To answer these questions we must consult either
a lexicographer or the person who stipulated that particular meaning
and, in both cases, we must expect a less than absolute answer. For
example, Schlick believes that the sentence "the campanile is 100
feet and 150 feet high" is inconceivable and logically impossible.
He sees a contradiction and, on the face of it, so do I.. But I also
see a tower 100 feet high, with a 50 foot foundation below the
ground. If the words are interpreted so as to fit a certain logical
form, then it is a contradiction. If they are so interpreted as not
to fit this form, it is not. But this is a question of usage, or of
intention; that is to say, it is an empirical question.
This leads to our next question: just what does Schlick mean by
logical possibility? He says that "not only can the logician be an
empiricist at the same time; he must be one if he wants to
understand what he himself is doing." [Schlick 157] Or again,, "I
call a fact or a process ‘logically possible’ if it can be described; i.e., if the sentence which is supposed to describe
it obeys the rules of grammar we have stipulated for our language."
[Schlick 154] Logician Alonzo Church defines "logic" as an
investigation of "the structure of propositions and of deductive
reasoning which abstracts from the content of propositions
which come under consideration and deals only with their logical form. [Church 170] not logic a purely
formal enterprise,
and thus by its own rules, independent of experience? It would seem
that both tradition and current professional usage would support
this conception. Yet Schlick suggests that a logician must be
an empiricist, and that logical possibility can be tested by "describability."
Schlick’s criterion of "logical possibility" seems to have
wandered well beyond the province of "logic" on board as the term is
currently understood and used by logicians. Apparently, since he
believes that with due caution we can have absolute meanings of
terms, and combine these into absolute propositions, we need only
exchange such terms for the variable’s (the "X’s" and "Y’s") in
propositional functions to have propositions every bit as simple and
cut-and-dried as the logical forms. This, I submit, is a most
perilous assumption. The meanings of words are necessarily and
forever involved in the morass of human caprice and confusion. We
might strive for some position and succeed quite well for many
purposes, but a word in a natural language will never be so
simple and clear of connotation, vagueness and ambiguity as a
logical variable such as A, B, C, or x, y, z. I would suggest that
form, and form only, be the province of logic It may be wise to test
propositions logically for formal consistency and to analyze
and combine premises to discover what implications may be involved.
But, in so doing, we must standardize our terms and axiomatize their
relationships (i. e., transform our terms into logical variables) so
that, as our assumptions are logically examined, empirical
information will be systematically excluded, to be reinstated only
as the variables are interpreted as empirical terms. Anything else,
whatever it is, by stimulating or even scientifically productive,
but it will not be logic – at least a pure logic. Logic has
been wisely limited to a formal science. I see no warrant in
extending its application to either empirical meanings or facts.
IV. TRUTH
To the logical positivist, a proposition is
true if the
state of affairs designated by the proposition corresponds to actual
facts. Like meaning, truth is absolute – there’s no
such thing, says the positivists, as a proposition that is "more or
less true." For, he argues, is not a state of affairs either flatly
the case or not the case, regardless of whether we know is or
not? And, accordingly, is not a proposition either true or false?
This might be argued if we could know precisely what
is being asserted by proposition. But if we acknowledge that words
must, in principle, fail absolute precision of meaning, how are we
to argue that sentences may express propositions which assert states
of affairs with complete and unambiguous clarity. And without such
clarity how then can there be an exact correspondence between an
assertion and a fact? Let us consider as a simple example, the
sentence "there is a dog in the next room." This is as plain and
simple an assertion as we could reasonably ask for. Is not its
meaning (and therefore its mode of verification or conceivability,
not to mention its logical possibility) quite obvious – is not
plainly true or false? For practical purposes the meaning of the
statement is quite cut-and-dried. But even here, we are not totally
assured either of its meaning or of its verification. We
are told for example, whether the picture of the dog on the wall, or
the dog shaped paperweight will render the statement true. Probably
not, and with a slight clarification, we will then look for a "live
dog." But then further clarifications are in order: which adjoining
room is the "next room," or will either room answer the question?
Does the "room" include the closet, the window sill, or the doorway?
When the dog in the room? When its nose is in the door? When
all but its tail is through the doorway? Is this strange creature
"really" a dog, or is it a fox, a wolf, or whatnot? And so on. If
the person who utters the statement is available, he will perhaps
supply answers to our qualifications (and thus present increasingly
specific new sentences), until our questions become ludicrously
trivial. Then he will probably display a justifiable exasperation at
our pettiness, feeling that his original statement was precise
enough to express his essential message – as it probably was. If we
read the statement, then we will be able only to surmise what
may have been meant by the person who wrote it. In both cases, the
meaning might be identified as a state of mind, or intention, that
is expressed by the sentence.5 The meaning could be
expressed in different sentences, or even in a different language,
but the meaning, that is, the "intention" of the speaker would be
the same. And in all cases the meaning would be less than absolutely
precise. Some statements are "relatively precise;" such statements
as "there is a dog in the next room." Some are less precise in
meaning, such as; "the fall of Rome came with the barbarian
invasions," and "neuroses are commonly manifestations of child
repressions."
Is truth, then, an absolute quality of propositions? If "raw
propositions" cannot assert a state of affairs with total
completeness (and apparently they cannot) – if we cannot say with
absolute certainty, what would completely verify the proposition (e.
g . Seeing a "genuine" dog "really" in that room) – then we have no
exact replica in our minds of any actual state-of-affairs. We can
come close, very close, quite close enough for many of our
purposes, but an exact correspondence will, in principle, forever
elude us, and so too will "the whole truth." Accounts
of
facts are not the facts they assert, nor can they be.
Propositions do not possess the certainty, the "hardness," of the
facts-in-themselves which they purport to assert.
V. THE EMPIRICAL MEANING CRITERION
Meaning, as we have previously noted, is commonly associated
(by contemporary philosophical analyst) with verifiability or
confirmability. For some positives, the meaning of a
proposition is the conception of its verification; to others,
the conceivability of the verification is a necessary and sufficient
condition of meaningfulness. These criteria arise from a conception
of "cognitive meaning" as the quality of being either true or false.
For us to speak intelligibly of the truth value of a proposition, we
must be able to conceive of an experience that would indicate
whether or not the proposition is in fact true or false. It is not
necessary, say the positivists, that the verification be achieved,
or even that it be practically possible, but it must at least be
conceivable. To Schlick, this means that it must be consistent with
the meanings assigned to the terms and with the rules of logic and
discourse, which is to say, "logically possible."6
The verifiability criterion of meaning (or, as it is otherwise
known, "the empirical meaning criterion") has many features to
recommend it, but as we have before, many of its advocates have been
too bold in their hopes for its application and utility. It is
certainly worthwhile to distinguish between logical validity of
actual truth, as it is to point out the relevance of experience to
the question of the truth or falsity of our assertions. It is also
handy to have a ready guide with which we might determine whether or
not we are wasting our time when we ask certain questions. But this
final advantage, that of the "handy guide," has perhaps been too
eagerly sought, and much of the mystery and scope of human
experience and language has thus been slighted.
The exaggerated ambition of the early positivistic criterion has
been recognized and at least partially corrected. But in the
process, much of its early uniqueness has been compromised, and
probably wisely so. The early positivists called upon the
philosophers, philosophical analysts, and scientists to re-examine
their assumptions and their methods. Time has shown that the
positivists have accommodated themselves to other analysts and to
the scientists at least as much as these others have heeded the
challenges of the positivists.
Simply stated, the most significant revision of logical
positivism has been the following; it was first insisted that for an
assertion to be meaningful, it must be possible (if not technically,
but at least conceivably) for some experience, some fact, to settle
the question of its truth. Now all that is commonly asked is that it
be conceivable for some experience to have bearing upon the truth of
the assertion. The requirement of "verifiability" has been replaced
by that of partial "confirmability." As C. I. Lewis sees it:
Reference to verifiability as essential to meeting is
only a roundabout way of pointing out that unless you are
somehow prepared to recognize the factuality you assert, in
case that factuality should be, or could be, presented to
you, your verbal expression is not a matter-of-fact
statement because it affirms nothing intelligible. Any
conditions of verification over and above this one
requirement that a matter-of-fact assertion must have
empirical science – whether these further conditions be
"practical" or "theoretical" – are irrelevant to the
question of meaningfulness. [Lewis 391]
This conception is extreme in its liberality – few empiricists
(or scientists) would argue that it is too restrictive. Lewis is a
pragmatist, not a positivist, and there’s much more substance to
contemporary positivism than this. Nevertheless, this passage
indicates how generous the empirical meaning criterion has become
with some contemporary thinkers. The retreat from the requirement of
verifiability to that of confirmability is noted by Hempel:
The earlier insistence that each statement of empirical
science should be fully verifiable or falsifiable by mens of
observational evidence has been modified in two respects:
(1) by the recognition of the scientific hypothesis cannot,
as a rule, be tested in isolation, but only in combination
with other statements... and (2) by replacing the overly
rigid standard of complete verifiability or falsifiability
by the more liberal requirement that a system of hypotheses
must be capable of being more or less highly confirmed by
observational evidence. [Hempel 43].
This modification is both necessary and wise, for some very
significant scientific assertions would be, according to a strict
verifiability criterion, meaningless. "Natural laws" for example,
hold for all the indicated cases, past, present, and future. If all
the virtually infinite instances implied by a natural laws are not
individually confirmed as, of course, they cannot be, the
law is not fully verified; it may only be highly and, for
practical purposes, indisputably confirmed.7 Moreover,
the condition for verifying a statement about the past would be for
a person now to experience the event in the past – a
logical impossibility. Statements about the past are meaningful on
the basis of evidential traces of past events (e. g.,
records, artifacts, etc.), but these indications are forever
incomplete; that is, they can confirm but never verify a
statement about the past. [Hospers 97] Finally, if, as we have
suggested, empirical terms are in principle imprecise to some
degree, owing to the limitations of natural languages, it will be
impossible to conceive with total clarity, or to confront the total
certainty, a verifying event.
The positivist might reply that, while we cannot in fact, or even
with any logical possibility, be confronted with an historical
event, we can conceive of what such an event might be. Also
while we cannot confront all cases covered by a natural law, we can
conceive of what such a confrontation might be. Perhaps so,
but why use the terms "verifiability" or "verifiability in
principle" to denote this "conceivability?" Does the term
"conceivability" refer to anything other than that sometimes vague
intuition or picture-in-the-head that is popularly associated with
word "meaning"? Perhaps it may, but I confess that we might well
wonder just how much more is meant by the term "conceivability."
Have not the positivists, with this term "conceivability," arrived
dangerously close to their point of departure – the common sense
conception of "meaning?"
But could not a conceived-of state of affairs or verifying event
be stipulated to mean such-and-such an assertion? And if so,
would this not remove the stigma of fallibility and imprecision?
This interpretation might be acceptable, but it would be trivial,
and it would in no way involve objective verifying events. It is
nothing more than a matter of stipulating that "by proposition X I
mean this conception that I now have." Even here, the exact
conception (in the mind of the stipulator) will be both private and
forever gone in the next moment. Moreover, the communication will be
only partial and imperfect to all those to whom the definition is
expressed. Or, if the definition of the proposition is ostensive (i.
e., "by proposition X I mean that state of affairs), the
indicated state of affairs will be variably preconceived and
variably remembered by all who encounter it. Absolute meaning,
truth, verification, and conceivability, are unattainable ideals –
that useful fictions – and they must remain so. Such is the finite
lot of mankind, positivist to the contrary notwithstanding.
By attempting to base their concepts of verifiability upon what
they believed to be strict logical foundations, many positivists
overlooked the relevance that verification must have to contingent
empirical reality. This oversight is understandable since such an
acknowledgment would only undermine the absolute logical models that
the positives have set up. For example, let us now recall Schlick’s belief that "verifiability, which is the
sufficient and necessary condition of meaning, is a possibility of
the logical order; it is created by constructing the sentence in
accordance with the rules by which its terms are defined." Or
"everything that I can describe is logically possible." And finally,
"the possibility of verification does not rest on any ‘experiential
truth,’ on a law of nature or of ay other true general proposition,
but is determined solely by our definitions, by the rules that have
been fixed by our language. [Schlick 155-7]
But is concept "everything that I can describe" a "logical"
concept? It appears to be a quite explicit reference to "real"
things that exist out in the world of experience. Consider also the
work of verifiable." It seems to suggest (in the suffix "-able") a
possibility to be realized in the future. Moreover, a "verification"
could hardly involve anything less than a confrontation with the
empirical world, for such a confrontation were unnecessary, what
warrant would thre be for using this word? How, then, can Schlick be
so certain – so logically certain – about a reference to
future empirical events? If his "meaning" is logically determined,
what relevance should such empirical events have to it? Could we not
imagine a verifying event, only to find (when we encountered
reality) that such an event was impossible and that the question of
the truth of our proposition was in the way settled? Do we not
sometimes discover verifying procedures, long after
entertaining a proposition? Do we not indeed have to have the
"meaning" of a proposition somehow "in mind" in order to ask how it
is to be verified Is not the question of verifiability at least
partially contingent upon our weak and fallible knowledge of the
field of verification, namely, physical reality?
V. LOGICAL POSITIVISM AND SCIENCE
The positivist’s search for an absolute criterion of meaning and
verification has prejudiced his attempt to establish a close rapport
with the work of the empirical sciences, for much significant
exploratory theorizing (which often must utilize vague and tentative
concepts) is simply inadmissible by such a strict criterion. An
unyielding position that does not acknowledge the use of vague and
ambiguous language must thus be so confining as to deny significance
in a very wellsprings of original scientific thinking. The history
of science is too rich in examples of the refinement of humble
conceptions into significant in penetrating the scientific
terminology to warrant such a narrow analytic procedure. Such
refinement is indeed one of the most significant outcomes of
scientific investigation. We have learned, through the work of
science (particularly the life and social sciences), that the
failure of a scientific hypothesis to present exact terms with
precise meanings involving explicit conceptions and rules of
verification need not necessitate a rejection of the hypothesis on
the grounds of "meaninglessness." And so, we must add our knowledge
of the practical work of science to our reasons for rejecting the
absolutistic conception of meaning and verification.
On the other end, the fact that many early working hypotheses are
necessarily vague and tentative, should not excuse those
formulations which are so stated as to be irrelevant to any
empirical confrontation. Vagueness and tentativeness in scientific
hypotheses must be corrected with deliberate speed. While some
important theories might be at first quite primitive, and even
pre-scientific, those who examine such theories must adopt valid
procedures of investigation and must seek increasingly specific
formulations have significant confirmations. And so, the
positivist’s ideals of "absolute" meaning and verification, for all
their difficulties, remain useful (though fictional) goals toward which the proponents of primitive scientific hypotheses might
well aspire.
To a philosophical analyst, cognizant of the nature and weakness
of natural language and the workaday methods of scientific research,
a modified empirical meaning criterion remains a useful critical
device. As such an analyst, while properly critical of the extremes
of early logical positivism, nonetheless acknowledges his
considerable debt to the enduring contributions of the movement of
analytic philosophy and to the philosophy of science.
NOTES
1. It is of the very nature of [certain]
propositions that their truth cannot be established with certainty
by any finite series of observations." A. J. Ayer [Ayer 37].
2. Hempel too notes that "the conception of an
analysis of ‘the’ meaning of a given expression presupposes that the
conditions of its application are (1) well determined for every user
of the language and are (2) the same for all users during the period
of time under consideration... Clearly neither of the of [these
presuppositions] is fully satisfied by any natural language." [Hempel
9-10]
3. A. J. Ayer admits that he once made this
oversight: "[A] difficulty which I overlooked in my original attempt
to formulate the principle of verification is that most [I would say
all, ep] empirical propositions are in some degree vague.
Thus what is required to verify a statement about a material thing
is never that he occurrence of one or another of the sense of
contents that fall within a fairly indefinite range... There is
never any set of observation statements of which it can be truly
said that precisely they are entailed by any given statement about a
material thing." [Ayer 12].
4. Such a belief might require the positivists
to reconsider their rejection of synthetic a priori propositions.
5. "Meaning is not inherent in words and
sentences; it is given them by their human users. Strictly speaking,
it is not sentences that mean at all; we speak as if this
were so, but this way of speaking is and ellipsis; actually it is we
who mean various things by our sentences." [Hospers 75].
6. Schlick does not mean to say that
"conceivability" and "logical possibility" are synonymous. He
regards conceivability to be the essential test of such possibility
(i.e., while something might be logically possible but
inconceivable, perhaps to the frailty of human imagination, nothing
could be both logically possible and conceivably
7. "The application of a law is infinite in
extent; verifying it would mean making an infinite number of
confirmations; and it would generally be held that it is not
logically possible to perform an infinite number of acts of
confirmations in a finite length of time." [Hospers 198]. Ayer
substantially concurs with the above remark of Hospers: "The
question that must be asked about any putative statement of fact is
not, ‘would any observations make its truth and falsehood logically
certain?,’ but simply, ‘what any observations be relevant to the
determination of its truth or falsehood"?" [Ayer 38].
