Abstract.
The
main purpose of this treatise is to discuss the parts of Michael Dummett's
philosophy of language and of Bas van Fraassen's philosophy of science that may
concern Treatise No 1-3. Thus, the
text explains 1) what can be substantiated and for which reasons, 2) what can
be criticized or rejected, and 3) what can instead be substantiated. This is
clarified in the section “Introduction”.
First,
however, the importance of the semantic element in the debate between realists
and anti-realists is explained in the section “The semantic approach the
realism debate”.
The
section “Berkeley’s phenomenalism” points out the realistic elements of Berkeley’s
phenomenalism is pointed out and opposed to an anti-introductionistic view. It
is made clear how Berkeley’s project is different from that of Dummett.
The
section “Dummett's anti-realism” presents and discusses Dummett’s definitions
of realism and anti-realism. However, first it discusses Dummett’s fundamental
ides of discussing statements and entities rather than entities.
The
section “A clarification of the concept of object” first clarifies Dummett’s
view on the concept of object, which is compared with Berkeley’s philosophy. Thereafter
is looked at the anti-introductionistic view on the concept of object.
The
section “Subjects in the realism debate” looks at both general subjects and specific
examples. Realism and anti-realism including anti-introductionism must be
defended by means of concepts that do not presuppose the respective directions themselves.
This demand is satisfied unproblematically to the latter. It is pointed out
that Kant with his concept of a thing in itself, which builds on semantically
unclarified ideas, expresses an extraordinary realism.
The
section “States and imaginations” first discusses some examples of statements
and imaginations that according to realists can be simply true. Next, our fault
is made clear: Inspired by what we know of, we imagine situations whose
descriptions only apparently have semantic meaning.
The
section “Fictions and meaning” examines our possibilities for making
semantically meaningful
imaginations. Our imaginations of the past are yet an example of the
above-mentioned kind of imaginations. Dummett’s view of the reality of past is
criticized.
In
“The correspondence theory of truth”, the principle of this name is criticized;
in this way the anti-introductionistic view of anti-realism is clarified. Thus,
it is concluded that the correspondence of a statement with reality is
identical to what the justification of the statement consist of.
In
“The importance of the top down line
of thought”, the top down principle
as an argument for anti-realism is treated.
The
section “Truth and semantic meaning” compares statements about truth and
existence with statements about statements’ semantic meaning. This can be
exemplified with e.g. statements advanced by Friedrich Nietzsche and Henri
Bergson. Statements about statements’ semantic meaning can be concluded to be
the most strong of these. Realistic criticism of this is discussed.
The
section “Bas van Fraassen's
constructive empiricism”
looks at basic elements of van Fraassen’s philosophy of science. Van Fraassen’s
literal understanding of scientific statements can be criticised for not having
semantic meaning just as his criticism of Grover Maxwell’s view of
observability is insufficient. Contrary to this, van Fraassen’s view of the
concept of model can be a solution. Problems with that are discussed however.
In
“A critique of realism as a good explanation”, the contention that realism is a
good explanation of the progress of science and thus must be assumed, is
criticized, as this issue is important in the discussion of scientific realism.
1. Introduction.
The
treatises Treatise No 1-3, especially
Treatise No 1-2, were a presentation
of a number of points of this hic-et-nunc-anti-realism
and the argumentation for them. To the author of this treatise, at last they made
up an entirely new, clarified wording of a long series of notes, as this result
was both more systematic and coherent.
For
that reason, I will here make up my accumulated debt to the philosophy of Dummett
and that of van Fraassen, i.e. partly what my own philosophy is and partly what
I have taken a position on in these treatises, including what I have criticised
or endorsed (”adopted”). This account does not deal with the question whether I
have understood these philosophers correctly; for that matter, I may have let
me inspire by a misunderstanding. Instead it is a matter of justifying my
results with due reference to these sources of inspiration, whether I have understood
them or not. Moreover, a number of examples and related subjects are analysed
in order to clarify the presented viewpoints from another angle.
It
will appear that my usage and many of my formulations are influenced by those
of Dummett. On the other hand, my philosophy is especially inspired by
Berkeley's phenomenalism, though I have given it a different angle. Primarily,
however, my philosophy started some time before these papers with the idea that
the conception of the world as “a box or space with something in it” was
completely unacceptable. My intention has been to elaborate and substantiate
this idea in these articles.
The
result has become a top down-based
anti-realism, that rests on minimal assumptions, i.e. assumptions about what we
can at most put into our statements, and thus into a number of concepts of
existence in different respects. As explained in “The aim of the treatise” in Treatise No 1, this is why this
anti-realist view is called “anti-introductionism”.
Finally,
it must be remarked that when statements like “this point of view presupposes
realism” are put forward in these treatises, the term “realism” is not meant as
a critique in itself. The aim is only to put forward that the premise realism has to be substantiated so that
the argument for the viewpoint in question can be completely valid.
However,
it must be emphasized that it is the realist who introduces entities whose
existence he asserts, but which the anti-realist or at least the
anti-introductionist does not want to introduce. For that reason, the realist has
to be able to make good the necessity of these introductions. On the other
hand, it is of course in the interest of the anti-introductionist to argue for
the unnecessariness of introducing the entities in question, just as he has to justify
the semantic point of view.
2. The semantic approach to the realism debate.
For
the sake of brevity, the debate between realists and anti-realists can be
called “the realism debate”. This choice can be justified by the fact that it
is a debate for and against realism, though it may be misleading to let it take
place solely in realist terms and concepts, as the discussion of these is exactly
a part of the substance of the debate.
The
semantic approach in the realism debate is not to be understood as a direction. It is simply a necessity. For the central question in
the realism debate is to find out what justified semantic meaning we can at most put into a number of different statements.
Besides
among other things, time and consciousness, these can concern the following
five sorts of entities whose ontological justifications are at issue:
1) Objects that are observable in principle,
namely
a) things just in front of us,
b) things in the next room,
c) objects in distant galaxies,
2)
so-called “microscopic objects”, i.e. objects seen in microscopes, and
3)
so-called “not-observable entities”, e.g. under name of “elementary particles”
and “force fields”.
In
fact, both what are referred to under point 2 and point 3 are so-called not-observable
entities. For the term “microscopic objects” at point 2 refers to putative
entities imagined to exist as small versions of objects that are observable in
principle (cf. point 1), while the term “not-observable entities” at point 3
refers to putative entities imagined resemble the scientific, constructed
entities, somehow or other. The realist will put forward that the mentioned
putative entities are real entities, while on the contrary the anti-introductionist
will assert that the references are putative too, as there is nothing to refer
to in this connection.
A
single specification: Physical objects we can sense are observable in principle. If we send such an object
out in the space by means of a rocket, the object will be practically observable some time after the launch, but not
practically observable forever. According to some realists, however, the word “electrons”
denotes entities that actually exist but are just not-observable in principle, while other realists will think that
they are indirectly observable, and some realists do not want to distinguish
between indirectly and directly observable entities.
Realism
concerning physical entities is sometimes defined as the view that the
existence of objects is “independent
of our consciousness”. Here, it is natural to ask what it means that objects
exist “independently of our consciousness”. However, this question cannot be
answered directly as it presupposes that the formulation has a meaning altogether,
which is not given and wherefore the semantic meaning first has to be procured
in order that the formulation can be used. In addition, this construction is
based on a mistaken or inadequate way of expounding the realism debate, as it
presupposes that also the realist’s opponent can accept the interpretation.
This becomes especially obvious, if anti-realism is simply defined as the
opposite of the above definition of realism, i.e. as the view that “the
existence of objects is dependent of
our consciousness”; for the anti-realist will often define his position in
another way and argue for it on other premises. As an example, the author of
this text has by no means described anti-realism as the latter position, i.e.
the view “that objects only exist in a way that is dependent on our consciousness”.
Moreover, it has already been emphasized in this text (Treatise No 1, “Minimal suppositions”) that of course we do not
master our sense expressions. Instead, we can only observe the regularity of
nature. Finally, the decisive question in the realism debate is a clarification
of the meaning of our statements about existence. Thus, the here criticized definition of realism, is itself at issue.
It
is simply a trivial truth that every discussion has to begin with a
clarification of what is at issue, as most persons probably want to avoid
discussing assertions that may turn out to have no meaning. Moreover, talking about
something that proves not to have the meaning it is supposed to have is likewise
a delusion. We have to ask the questions: What is presupposed, what do the
concepts imply, and is the formulation reasonable? For these commonplace
reasons, the semantic approach is not a philosophical direction, but a necessary
clarifying part of any philosophical discussion. Therefore, the realism debate
is not primarily about existence, but about the meaning of statements about existence.[1]
Moreover,
this talk about the idea that the existence of a certain kind of entity is
dependent on consciousness, is already misleading because the concept of consciousness
is redundant, cf. the argumentation in the chapter “Our own statements and
their meaning” in Treatise No 3.
Apart from this, as mentioned, we do not master the regularity of nature (the
sense impressions), so already for that reason and in this sense our sense
expressions are not dependent of consciousness. Therefore, it is of course in
agreement with anti-realism to talk about an external world; the question is
just what we can put into this talk about an external world, which is a
semantic question. Obviously, it is necessary to put forward these basic
remarks.
On
the other hand, it is a fact that what
we can assert about nature among
other things depends on what we can sense. This is not in advance the same as the
allegation that what exists depends
on what we can sense. Instead, it implies that what we can assert the existence of depends on what we can sense. However,
it has no sense at all to say that what we on grounds of principle cannot
assert the existence of exists.
3. Berkeley’s
phenomenalism.
As Berkeley’s phenomenalism is
referred to several times in the following, it is shortly accounted for and
criticized here.
According to Berkeley Berkeley
With other words, it can at most be
the case that science is able to make theoretical constructions concerning our
apprehension of tings. For that reason, this process cannot be a reality, if it
is understood as something that takes place in itself, but must be understood
as a structuring of our experiences at the top-level of the explanation
structure. This means that our cognitive situation is such that we simply sense
things as wholes and that we speak of them in terms of things - as our concept
of things can be developed further and as our conversation about it works.
It is exclusively under unfavourable
circumstances (in darkness or fog, or if the object is more or less hidden)
that we only gradually realize what kind of object we see, i.e. almost as we do
according to Berkeley Berkeley
In spite of the above pointing out
of the realistic element in Berkeley’s idealism, it is nevertheless Berkeley’s
project to answer the question, “What is a thing”. That is, Berkeley is occupied
with the semantic meaning of the statement “This is a thing”, while Dummett, in
his treatment of the subject, rather aims at the truth of statements about
things than at understanding the concept of things.
According to the top down line of thought, we can elaborate our statements about
observations of things. As already touched, we may for instance sense that this
is a blue ball, after which we can substantiate
the statement that the object in question is a blue ball, etc. We do this by
use of statements we in their turn can explain more precisely, possibly in a
general way. Thus, we can conclude:
The concept of thing belongs at a step higher in our explanation structure than Berkeley
4. Dummett's anti-realism.
From Dummett's philosophy, the author of these treatises has especially
taken a position on some of the views Dummett has accounted for in “Realism”
(1963) and “The Reality of the Past” (1969) in Truth and other enigmas ([Dummett 78]). Though these accounts have
certain common features, they give occasion for different considerations.
4.1 Entities
versus statements about entities.
In “Realism” Dummett advances that rather than understanding the realism
debate as being about certain entities
or terms he finds it more suitable
to characterise it as a debate about a certain class of statements, “the disputed class”, e.g. statements about the
physical world or about the past. Dummett bases this choice on the view that
the different realism debates have different sorts of content with the
justification that e.g. phenomenalism is rather about reduction than about the
existence of material objects or the reference of their terms, while the realism
debates concerning the past and the future are not about the references of
expressions at all. ([Dummett 78], p. 145-6)
As regards the latter premise, however, it is not evident that the
realism debate concerning the past cannot deal with reference. For this debate
deals with which semantic meaning we
can attribute to statements about the past. The question is whether there is
something to refer to at all, that is, whether we for instance refer to the
past understood as former present time, somehow, or whether the past can at
most be said to be a construction, which we just talk about in a referring way.
Furthermore, something similar can be said about our concept of the future.
(Cf. the subsection “Time” in Treatise No
1.)
As to Dummett’s conclusion that it is expedient to talk about statements
instead of entities, I understand this such that when the realism debate takes
place in names of entities, or rather names of types of entities, it deals with the question whether it has sense
at all to assert that the types of entities in question exist, and if so, what
it means to say that they exist. However this just corresponds to Dummett’s following conclusion that the
debate deals with what concept of truth is suitable for statements about the
disputed class, for which reason according to Dummett the discussion must be
about what kind of meaning the statements in question must have. ([Dummett 78],
p. 146)
The discussion of the truth of a statement must necessarily involve the
meaning of the statement in question as this discussion presupposes a
clarification of what it means that the statement is true. Moreover, the
determination of the meaning of statements is of course so fundamentally
decisive to the discussion of every philosophical statement or question that
what Dummett adduces about this cannot be quite new. Thus, also the empiricists
Berkeley and Hume involved the concept of meaning in their philosophy[2]. It has already been made clear
that the discussion of what we can say about the world also involves the
question of what is needed for different kinds of statements to have meaning, i.e.
in what this meaning consists. Cf. the section “The semantic approach to the
realism debate”, above in which it is put forward that the semantic approach in
the realism debate is not a philosophical direction but a necessity for every
philosophical discussion.
However, as it is trivially clear that it is only possible to talk about
entities by means of statements, this may seem to make unimportant Dummett's
distinction between discussing the existence of entities and discussing
statements about entities. Still, however, there is a difference between
focussing on entities and on explicit statements about entities, for if we
always specify which statements we discuss, focussing on the truth of
statements may clarify certain anti-realist views. Thus, this approach can make
these views less contra intuitive, in so far as they seem to be so. If this
discussion is not to be trivial, of course it ought not to be just about
existence statements of the form “There are entities of the type T”, but also
about other kinds of statements concerning entities of a certain kind. Moreover,
discussing specific statements about entities is more precise than talking
about the existence of entities, since such specific statements are
presentations of what the debate is about. In this way, in all its seeming
harmlessness, this forms a shared base of discussion for realists and anti-realists,
as it may make clear central sticking points.
An example of the usefulness
of discussing specific statements about entities, rather than debating the
entities themselves, may be replacing the discussion whether the table in the
room next door exists not observed with a discussion of this statement. “There
is a table in the room next door”, or perhaps with a version the realist could
use in his criticism of phenomenalism “In the room next door, there is a table
nobody senses”. However, also the anti-realist could say that the former statement
is true if e.g. the table has just been placed in there and nobody has been
seen removing it. So he can confidently say that the table stands in there making
a pressure on the floor, and even that it is making dents in the carpet, as we
have found that this is what tables do when they have stood at the same place
for some time, etc. Even if it can be documented that no persons are in there,
the anti-realist will of course say that version two is true, as this is in agreement
with the meaning of that sentence. However, the realist will insistently assert
that the table is simply there “in itself”, “on its own terms” or the like. He
will so to say just put more content into the sentence, that is, something that
we cannot use, as it does not make any difference. Though we, along with the
realist, can imagine the table in there, it is not necessary, however, to do
this if we want to put forward such statements.
Of
course, the optimal solution of the realism problem would involve a general
clarification of the realism debate concerning all types of entities. As to
that, in Treatise No 2 it has been
argued for certain general demands on the
principle of inference to the best explanation as regards not-observable
entities7. These concern a number of entity
types belonging to science and psychology.
4.2 Dummett's
definitions of realism and anti-realism.
4.2.1 Realism.
Based
on his considerations in “Realism” accounted for above, Dummett characterizes
realism as the view that statements about the disputed class possess objective
truth-values though we have no possibility of knowing them, as they are true or
false “in virtue of a reality existing independently of us”. Thus, according to
the realist, the meaning of the disputed statements consists in how their truth-values
are determined by states, which exist independently of whether we possess
evidence of them or not. ([Dummett 78], p. 146)
The
first part of this characterization of realism just seems to cover common sense realism, i.e. a realism
that is not necessarily philosophically based. However, as to the determination
of the truth-value of statements, in practise its adherents may have a more
radical point of view, as the realists in questions can simply assert that a
statement can be simply true, i.e. without it being necessary to state closely
specified conditions that have to be satisfied in order for the statement to be
true. They may thus say, “Dualism is no problem” or “Sympathy is situated in
the genes”, though it has not been clarified how. Some realists will not at all
accept the idea that the truth-conditions of statements have to be determinable
in order that the statements can be true, since it involves the human being and
its conclusions, and as this has nothing to do with the state of affairs in
themselves. This also applies to statements that contain terms of so-called not-observable
objects, states or accidents, and which therefore cannot be used to describe
our daily life phenomena directly as it erroneously is often done. Compare e.g.
the discussion of the use of the scientific concept of adrenaline in the
section “Fictions and meaning”.
In
“The Reality of the Past” in Truth and
Other Enigmas, Dummett assets that the realism debate generally is about which
kind of meaning we can attribute to a statement concerning a certain subject. Dummett
thus describes the realism as the view that the meaning of such a statement consists
in our knowledge of “what has to be the case for it to be true”, and our
understanding of the statement just consists in this knowledge. ([Dummett 78],
p. 358. my underlining)
It
is worth noticing that in the way it is
worded, this connection between the meaning of a statement and our
knowledge of what has to be the case in order that a statement can be true is
also in agreement with the hic-et-nunc-anti-realism,
but in another way than the realist conceives it. For without contradicting this, we can regard the truth
conditions as demanding actual experiences
concerning a discussed statement, and it will later be substantiated that we
must do that - especially in “The correspondence theory of truth”. Therefore,
this part of Dummett's account in itself does not express a difference between
realism and anti-realism; the difference consists alone in the idea of what
principally has to be the case in order for a statement to be declared true, i.e.
in what is contained in this concept.
However,
Dummett clearly expresses a difference between realism and anti-realism with
the following view of realism. According to the realist concept of the truth of
a statement, we are not always able to acknowledge whether the conditions of it
are obtained or not, as every statement is either true or false, “independently
of our knowledge”. ([Dummett 78], p. 358)
Nevertheless,
this talk about truth conditions that exists
without our being able to acknowledge it, is an obvious example of realism
expressed in the realists' terms, as it presupposes that it has sense at all to
talk about the world in this way. However, the discussion is exactly about this
very question. Moreover, it must be remarked that this talk about “what has to
be the case for a statement to be true (i.e. the truth conditions of the
statement)” presupposes that something can be the case in itself - unless we
are dealing with what is present to
us. For it is certainly meaningful to talk about what has to be experienced in order that a statement
can be declared true, i.e. the assertability conditions of the statement, but
not to talk about something that simply is considered true no matter how, and
whether it is possible to verify the statement or not. By truth conditions, we
must thus understand assertability conditions. According to the nature of the
case, the realists will not accept this, of course, since the realistic point
of view involves the idea that truth is independent on verification.
Moreover,
this implies that if what has to be the case in order for a statement to be
true is understood as something that exists and which is possibly discovered,
this is also to presuppose realism, of course. Thus, Newton did not discover his laws of nature, but found
out how to describe the phenomena by means of certain mathematical expressions;
in this way, the phenomena could be predicted, also. However, it is very common
to express oneself as if such a realist talk of truth has sense; and in many
cases, the anti-realist has to argue against this kind of ideas outside the
philosophical debate, i.e. in everyday or scientific contexts.
Dummett's idea that a statement about existence can be understood only
by reference to entities whose occurrence we may regard as evidence of the
statement ([Dummett 78], p. 146) can be regarded as agreeing with the view that
we cannot put anything else in our statements about existence than what justifies
them. This is expressed in Treatise No 3,
“Summary of Treatise No 1 and 2”. The viewpoint has subsequently been
elaborated in “Grover Maxwell’s argument for the existence of not-observable
entities”, Treatise No 3, in which
the following is concluded:
We cannot put more in statements
concerning not-observable entities than the content of the observations and
theories that is referred to in the deductions of the existence of these entities
- or rather, that cause the use of these concepts.
Thus, though we use the concept of microbe, we cannot assert that the
laboratory technicians handle with some microbes that exist in themselves, only
that the technicians change some microbe-pictures, as the technicians do not
relate to anything else, but only to these. This does not mean, however, that
these actions cannot have other consequences that can be observed.
4.2.2. Anti-realism.
In
“The Reality of the Past”, Dummett advances that according to anti-realism the
meanings of statements in the disputed class are not given by conditions
that can exist independently of our possible knowledge, but by “the
conditions we recognise as establishing the truth or falsity of a
statements of that [the disputed] class”. ([Dummett 78], p. 358/9, my square
brackets and underlinings.)
It must first be remarked that if this
characteristic, which the anti-realist idea of the meaning of statements is
here contrasted, is attributed to the realism, this can clarify the
above-referred description of realism ([Dummett 78], p. 358). For this would
imply that according to realism the meanings of statements in the disputed
class are given by conditions that can exist independently of our possible
knowledge.
Concerning
the very characteristic of anti-realism, it is not completely clear what is to be
put into the verb “to recognize”, which demonstrates the subtlety of this
problem and thereby exemplifies the fact that this debate particularly deals with
precise formulations. However, if the expression “conditions which we recognise”
is to be understood as referring to our experiences, it must be the present
facts that constitute the truth of a sentence, according to Dummett.
This
interpretation is supported by Dummett’s assertion that the constructivist thinks
that mathematical statements are true “in virtue of (…) 'Our possession
of a mathematical proof'”. ([Dummett 78], p. 361, my underlining) However, I am
not quite able to determine whether this example completely corresponds to Dummett's
view, but that is not my purpose either. On the other hand, a generalization of
this view corresponds to the anti-introductionism.
In
“Realism”, though, Dummett more clearly asserts about the meaning of statements
that according to anti-realism, statements about the disputed class can only be
understood by reference to something we could know and whose occurrence we may
regard as evidence of the statement, just as its meaning depends on what we
regard as evidence of them. ([Dummett 78], p. 146) This statement unambiguously
implies that if a statement is about existence or states, evidence of the statement
has to be available for the statement to have meaning. Moreover, it must also
be given what kind of evidence is required to assert the existence of a certain
type of entities.
(What
reason Dummett has for avoiding the term “verification”, I do not quite know.
However, though the term has a technical ring it shares with the verb “to
observe”, the word “verification” can be used synonymously with the term
“providing evidence”.)
Another
possibility of construction consists in the realist view that when the truth of
a statement has been proved, the statement was also true before the proof, i.e.
in itself, such that the statement would also have been true and thus have semantic meaning whether it
had been proved or not. This view is of course exactly contrary to the core of
anti-introductionism: that the question of existence presupposes that its
semantic meaning is clarified. However, also the anti-introductionist can put
forward - though in a quite different respect - that after the proof of the truth of a statement, it can possess meaning
to assert that a statement was true before
the demonstration of its truth. This means that we can now (after the demonstration) attribute a truth to the statement
that applies at earlier points on the time-axis, not just in the sense that we
can refer to the regularity of nature, but also because it is simply settled
now. This is the case in situations where we, based upon this regularity and present
premises, can relate a consistent fiction to a point on the time axis before
the now.
What
we ultimately refer to in our
explanations must necessarily be the phenomena and their regularity, i.e.
observable entities and their properties. This is also true of the disputed
statements concerning a scientific theory when we refer to - or rather, use
terms for - some of the so-called not-observable entities the theory
presupposes. Moreover, it applies if the terms of these concepts form part of
other explanations (at a lower level in the explanation structure). These
concepts are all introduced constructions. For instance, we do not know of electrons
understood as concepts by analogy with observable entities but only of certain phenomena that can motivate our talk about electrons. The fact is just that certain
observations have made scientists use the term “electron” in such a way that
they do not have to discuss the observational background of this talk, unless
something special happens that makes it problematic.[3]
As
expressed above, the conditions under which a statement is true or false, are
the conditions under which the statement can be declared true or false. Put in another way: The conditions of the
truth of a statement consist in the conditions of the decidability of the statement.
Thus, satisfied truth conditions are present, i.e. observable. In other words,
truth involves knowledge of evidence. From these innocent facts, the following
can be concluded:
Statements
that are seemingly about microscopic facts or elementary particles are actually
about observable – and thus macroscopic - facts.
5.
A clarification of the concept of object.
5.1
Dummett’s view of the concept of object.
According
to Dummett, reductionism is the idea
that statements about a certain kind of entities, the “disputed class”, can be
justified by statements about another kind of entities, the “reductive class”;
the idea is that statements about entities in the disputed class can be
paraphrased into statements about entities in the reductive class. Thus, if the
disputed class of entities is material objects the corresponding reductive
class is sense data. ([Dummett 78], p. 156.)
In
“Realism”, Dummett presents an argument for the view that phenomenalism does
not have to be a reductionism - a defence on behalf of the phenomenalists, so
to say. As far as the author of this text understand it, Dummett begins his
argumentation rather obscurely and circumstantially by asserting that if we
disregard the phenomenalist’s “claim that there exists a reductive class” and
that material-object statements can be translated into sense data statements,
the following view remains. Any statement about a materiel object (that does
not deal with an observation of the object or somebody's handling it) can be
understood as a counterfactual of which the constituents are expressed by
reference to material objects (in a material-object language). According to
this view, such a statement can only be true if there are actual
observations that can serve as a ground for its truth. ([Dummett 78], p. 159,
my underlining) As far as I can see, Dummett’s idea must be that otherwise we
(according to this view) are in a situation, where we can neither determine the
truth of the statement nor its falseness (the truth of the negated statement).
Here I am without an example, but perhaps the statement “There is a white table
in the room next door” or “There is a monolith on a planet in the Andromeda
constellation” could be one. However, the decisive point must be that at such a
statement can only be declared true under the demanded conditions, while otherwise
the statement does not have semantic meaning but is only random.
From
his argumentation, Dummett concludes that reduction is not necessary to reject
realism as regards material objects, and assets that the realists have not yet
put forward any counterarguments against that argument. ([Dummett 78], p. 159)
At least, however, against this view the allegation has been put forward that
the case is not about proofs but about existence in itself somehow. These will
be commented below.
In
“The Reality of the Past”, Dummett argues for the same point of view, he has
argued for above, as far as I can see, but in a somewhat clearer way, viz. that
reductionism is unnecessary if we are able to judge whether the conditions of
determining the truth or falseness of the statements about the disputed class
are present. For, according to Dummett, we are then able to express the
occurrence of these conditions directly by the use of statements about the
disputed class. ([Dummett 78], p. 361/2) However, this does not concern
Berkeley’s phenomenalism, as this is not a reductionism, but deals with what we
can put into the concept of a physical object.
The
anti-introductionism is not a reductionism either, but aims at clarifying the
semantic meaning the semantic meaning of statements concerning the existence of
a number of types of entities and thus avoiding introduction of more then what
can be accounted for. This project involves references to our knowledge of the
regularity of nature: It has thus already been argued that we explain and
justify our statements by references to observations and to our knowledge of
the state of things and especially of the regularity of nature. In addition,
this means that the assumption about the literal existence of theoretic not-observable
entities is superfluous. In these treatises, it has even been argued that it is
semantically meaningless to assert their literal existence. (Cf. “A. Realism”,
in “Three different views of the concept of a physical object” in Treatise No 2.)
However,
the realist will neither be satisfied with this nor with Dummett's elucidation
expounded above. According to the realist, the important point is not that the
things are in the room next door because
we know that they are placed there and not removed afterwards. He will assert
that the case is not about substantiation of our statements, but about the
reality of things, as the objects are there in a way he terms “in itself”, “on their own”[4], etc.
We cannot make use of these concepts,
however, and we do not need them
either, just as it - as already stated - is doubtful whether we always think of
them in praxis (which certainly cannot be used as an argument for anti-realism).
The realist will also assert, however, that the anti-realist does not succeed
in arguing against this realist viewpoint but that the anti-realist in fact just defends a different viewpoint that
concerns the regularity of things and our possibility of drawing conclusions
from this regularity. Moreover, he will put forward that also he (the realist)
may agree with this view but at the same time assert that the viewpoint does
not concern the discussed realism.
Against
this, the anti-realist may assert that there is not more in our existence
statements than what the anti-realist defends, and that the realist can only
defend his viewpoint if he can point out something in the world that makes a
difference that can prove his philosophy. The fact that the realist cannot
carry this out, though, does not mean that the realist does not express a difference, viz. a standpoint
that is compatible with accept of absurd claims about e.g. sensation of adrenaline
in the body and literal views of physiological theories. Such a viewpoint can
influence persons’ view of man and thus decisions about our society.
Moreover,
Dummett states that it is often adduced against phenomenalism that if the
phenomenalists can remainderlessly translate statements about materiel objects
into sense-datum statements, the phenomenalists will only be able to say the
same things as the realists, just in a different language. However, Dummett
thinks to have justified that there is in fact a difference between
phenomenalism and realism, as there will be conclusions the realist will not
accept. ([Dummett 78], p. 159-60.)
Here
some examples are again needed. It is certain, however, that at least the
anti-realist will reject some assertions as meaningless and ascribe to other
statements a different meaning than the realist will. An important example is
statements about the states of the brain, which is already touched on.
As
mentioned above, there is anyhow an important difference between realism and
Berkeley's phenomenalism, as Berkeley does not intent to reduce anything, but to adduce a new construction of the concept of object. Such a construction is
philosophically interesting in itself and is likely to have consequences for
other philosophical discussions.
Thus,
Berkeley is engaged with what constitutes
our physical surroundings, i.e. what we can put in our ideas of it. Dummett
disregards this problem in “Realism”[5], as
he only talks about objects as physical wholes, without further argumentation
or clarification of the subject. The question of how we shall relate to
Berkeley's philosophy thus remains unanswered by Dummett. Though the
anti-introductionism has rejected Berkeley's idea in its specific version, we
cannot pass over this problem in general, understood as the metaphysical view
of the question what kind of world we live in. Thus, a different view of this subject
has been put forward and defended in Treatise
No 1.
5.2
The anti-introductionistic view of the concept of object.
Unlike
Dummett, thus, anti-introductionism takes a decision on this part of Berkeley's
phenomenalism but just reaches a different view than Berkeley has. In this
connection too, it is a matter of assuming only what we have to assume. Here
Dummett only takes a position to the phenomenalist conception of the table in
the room next door, because he regards it as a prerequisite of phenomenalism
that statements about material objects can be translated to statements about
sense data, “the reductive class”. ([Dummett 78], p. 159) However, such a line
of thought drains the substance out of phenomenalism, because as mentioned,
what the phenomenalist - at least Berkeley
As
to this term, as mentioned in Treatise No
1, we could just as well characterise the realist counterpart of phenomenalism
as “introductionism” - a philosophy in which more than necessary is introduced.
Moreover, Berkeley is not bound to be able to translate statements expressed in
a “material object language” to statements formulated in a “sense data
language”, whatever that may mean, cf. the argumentation concerning that in Treatise No 3, in “Usage and existence”.
However,
for the sake of completeness, it must be mentioned that in his introductory
remarks in “Realism” (1963) Dummett states that the realism debate is about
either the real existence of particular entity types or “whether they are among
the ultimate constituents of reality”. According to Dummett, the quoted
question can be clarified by reductionism, as it can be answered negatively if
the entities concerned can be reduced. ([Dummett 78], p. 145) According to this,
it is true that Berkeley’s phenomenalism can be said to be a reductionism, as
it implies that statements about things presuppose reasoning about sense impressions
and therefore that our sense impressions and their regularity (based on which
we reason) constitute the concept of thing. However, this is not its purpose.
As
to anti-introductionism, this anti-realism is not dependent on whether reduction.
The question is simply what we can at most put into our existence statements.
Thus, it has already been argued that the reality of constructed entities -
regarded as something different from the constructed entities or concepts they
are - cannot be justified in a non-circular way2. For that reason, constructed
concepts are neither something else nor more than just constructed concepts.
With other words, they cannot be introduced as something that has a reality
beyond their construction, at all. Therefore, it is misleading to characterize
them as reduced away in
anti-introductionism; on grounds of principle, on the contrary, it is simply
unjustified to introduce them as
referring to literally existing entities. Thus, anti-introductionism is
anti-realistic in two ways:
Firstly,
these terms of so-called not-observable entities can at most refer to
constructed entities. Secondly, statements containing terms of these entities
have to be justified by reference to phenomena, i.e. to principally observable
entities or states. For that reason, of course, these statements come under the
anti-realism concerning the types of the entities
in question.
The
so-called principle of inference to the
best explanation can be used to show the presence of observable properties
and entities, as our talk about these already has meaning. However, it cannot
be applied to proving that our statements about constructions or invented types
of entities refer to real physical objects, for this presupposes that our talk
about them possesses literal meaning, which is just the matter at issue in this
dispute between scientific realism and anti-realism. This is not a criticism of
the very use of terms of so-called not-observable entities, only of the understanding
of this use as if it is on equal terms with the use of terms of known physical
objects.
6.
Subjects in the realism debate.
6.1
Realism – anti-realism.
The
many examples stated in this treatises of realism taken for granted in the
arguments for it, raises the question whether realism can be meaningfully accounted
for at all, i.e. without assuming in advance that realism has semantic meaning.
Thus, the realist often just puts forward statements about existence and
states, whose meaning he thus presupposes. However, what the statements mean he
doe not explain, though the meaning of such statements is exactly what is discussed.
The
question is whether it is at all reasonable to demand from the realist that he
must be able to describe his point of view in a way that is independent of the
principles of realism. However, this demand must be satisfied, for it would be
circular to define realism by means of concepts that only gain an alleged
meaning by means of this definition. That is, the realist must be able to
define realism by means of meaningful concepts, viz. by concepts that have a
meaning in advance.
Against
this demand, it may be adduced that it can also be demanded from the
anti-realist that he has to describe his view in a way that is independent of
the principles of anti-realism. However, that demand is fulfilled because
anti-realism does not have to be introductionism. The anti-introductionism, for
instance, does not introduce something extra that has to be defended; on the
contrary, that is simply a central part of its substance. For that reason, such
a demand is satisfied in advance.
6.2
Immanuel Kant’s realism.
Immanuel
Kant's concept of das Ding an sich is a
marked example of realism. The concept of the thing-in-itself is introduced by
Kant in Kritik
der reinen Vernunft ([Kant 90], B XX), but without further
discussion, as far as the author of these treatises can see. Thus, Kant takes
the meaning of this concept for granted, as he neither defines the concept nor
discusses what underlies it. Such an omission is of course a principal philosophical
failure, because discussing fundamental assumptions is an important part of
philosophy, as mentioned. However, we do not need this concept, which must
appear from the considerations in these treatises concerning the concept of the
regularity of the phenomena and thus of the concept of time, cf. e.g. the
bathroom example in “The past” in Treatise
No 1.
As
far as I have conceived it, Kant's concept of das Ding an sich has not
only to be understood as
1)
something that exists without our ever being able to perceive it, but also as
2)
something that is uncognizable in the sense that it cannot be characterized by
the concepts we have at our disposal.[6]
This
makes Kant’s realism into an extraordinary realism, an intensified realism.
According
to the anti-introductionism, however, it is the other way round:
We
can talk consistently by means of statements that contain terms of things and
deal with our sensation and handling things. This is why we have a talk about
things that exist, and why the concept of thing
has sense.
Thus,
it is not possible to justify the opposite statement, viz. that the fact that
our talk is consistent is due to the existence of things in themselves, or put
in a different way, a literal existence of things. We can talk consistently
about our experiences, simply because our experiences are consistent. This
needs nothing extra in our external world.
We
can even talk about the existence of not yet seen objects. Thus, we can draw
the conclusion that when we by tipping a box can hear something that sounds
like a rolling ball, there is probably a ball in the box, which can afterwards
be confirmed or disproved. However, this does not mean that in any similar way
we can conclude, “Things in themselves exist”, neither with the justification
that it is a useful assumption. For that would be to apply a scientific way of
concluding in philosophy – and lead to a conclusion that at the same time
introduces a new concept.2 In this connection, it must be
remarked that this talk of the regularity of nature certainly implies that
there are in fact regularities in nature, but not in the literal sense that the
nature of nature is regularity, only that we can regard some phenomena as
regular.
6.3
Fundamentally not-observable entities.
There
are entities that can be observed in principle, but just cannot be seen at
present, and there are in principle not
observable entities. Among the latter entities, there are the known kinds of
objects and their behaviour when we have no “access” to observing them (any
longer) but just imagine our being literally able to visualize them as objects
in distant galaxies, for example. As regards statements about these objects, it
is striking that the discussion of them is less trivial than the discussion of
statements about constructed, theoretical entities.
Concerning
these fundamentally not-observable entities, of course, we can have an idea of
e.g. what it means that a bust of Beethoven is drifting around in a distant
galaxy. In addition, the realist may assert that a statement about this can be
confirmed by means of space flights or production of special instruments we can
imagine to be done in the future. However, pro term we do not now know how to
carry out these projects. Thus, statements about them have no meaning. The
demand of available evidence is just not satisfied for the statement about the
bust, wherefore this statement has no meaning according to the anti-realism.
Moreover, the ontological status of distant galaxies is disputed, i.e. the
meaning of statements that contain terms of distant galaxies is unclarified. As
the requirement of evidence is not a part of a realist understanding of the
meaning of statements, the realist will assert that the statement about the bust
is either true or false - just as it is the case of any statement.
Nevertheless, this explanation is empty, as the opposite statement can be
asserted with just as good (or faulty) reasons.
6.4
Theoretic not-observable entities.
As
regards terms for the so-called not-observable entities, the constructed, on
grounds of principle they these cannot refer to physical entities of any kind at all[7].
Therefore, it has no sense to talk about objects corresponding to these
concepts and being observable as units in the same way as the objects we deal
with in everyday life. Nevertheless, they are called “not-observable” as if
there just for fundamental reasons are unfavourable states that prevent us from
that.
The
idea of the so-called not-observable entities is almost moulded in the same way
as Berkeley erroneously thought that we construct concepts of physical objects:
When weight and charge are observed as belonging together, science may fashion
a concept of a not observable entity with these properties, i.e. a theoretic entity.
Apparently,
we may imagine (or perhaps we can
only imagine that we are able to imagine) that we see these entities on equal
terms with the entities we can actually observe in everyday life. The entities
in question may be nuclei of atoms and their components or rotation of the
central mass of the sun - or rather sentences containing the terms “nuclei of
atoms” or “rotation of the central mass of the sun”. As stated, the
above-mentioned imaginations may be useful visualizations (of what takes place
in a mathematical model) - though they may also mislead or restrict the thought
- but they are not imaginations of something known or physical, even if we may
try to make them so. Instead, we just again have a communication that works to
a certain extent. In this case, however, it may be said to concern visualizations
of what takes place in a mathematical model.
6.5
Statements about the brain.
An
example of common sense realism is
made up by the widespread realist view that our knowledge and our abilities can
be encoded in our brains somehow - and that it thus has meaning to talk about
it. However, statements about this are nonsensical as nobody can acquire
knowledge of what happens in the brain at a logical level, and in that way
track down this alleged coding and its biological implementation and provide a
depiction of it, analogously to a printout of a computer program. For that reason,
the following must be emphasized:
No
matter how far the research will go, we will never be able to prove that there
is a one-to-one correspondence between our content of consciousness and parts
of the structure of the brain, for that reason alone that we will be able to describe
our feelings in every detail, and not even our grammar.
As
already mentioned in Treatise No 3,
we cannot finish of a description of the grammar of a language - at least we
will never be able to prove that we have finished our description of the
grammar definitively at a certain stage of our study, which has the same
consequences. The assertion that we have finished the grammar without knowing
it is meaningless if it cannot be proved (which it cannot, in the nature of the
case). What is more, this is not an expression of scepticism but an example of
the potentially infinite depth of the explanation structure, which is not a
human shortcoming, but a fact without which the world would be as clockwork or
program consisting of if-then-structures. Something similar would apply to all
finite kinds of explanations.
Moreover,
this problem involves the dualistic problem how such an encoding or
constellation in the brain could become sense experiences etc., whose
possibility it is at least a great philosophical problem to justify. The
assumption of such a possibility is not in agreement with the top down philosophy, which has already
been defended in these treatises, but is contrarily a case of bottom up extrapolation according to
which the so-called matter is the basis of the world, against which it has correspondingly
been argued.
6.6
Reality and faculties.
As
another example of realism is the idea that some persons have unusually good
faculties of seeing special aspects of nature or mankind and expressing this
insight correspondingly, because
they are artists - that is, that the cause of their faculties is their being
artists. However, this is an unnecessary introduction. For the explanation is
the other way round: it is because the persons in question exhibit such abilities that we call them artists at all. The
rejected idea presupposes that the persons in question have certain qualities,
an “artist core” of some kind, which manifests itself in artistic activity. However,
references to such a unit, or whatever it may be called, have no use, and the
concept of it has no sense for reasons similar to the rejection of the principle of inference to the best
explanation as regards not-observable entities. Of course, this does not
imply that the abilities in question cannot be related to other qualities and circumstances, but that does not imply
the existence of an “artist core”.
7.
States and imaginations.
Examples.
Example 1. If an explorer is found dead near
his route, does it then have any semantic sense to ask whether he had arrived
at his goal before that? It is tempting to answer affirmatively, as we think
that we have knowledge of our conditions in the world and that we walk about in
it under such and such conditions, and thus, we are easily induced to imagine
that this knowledge can be extended into applying to circumstances of which we
have no knowledge. Expressed without presupposing realism, we are easily induced
to assume that it has sense to talk about the existence of circumstances we do
not know, but which resemble the circumstances under which we move. However, it
can at most have meaning to talk about and to start to search for so-called evidences.
It
is also tempting to try to exe exemplify the problem with the following
question: “If an explorer reaches his goal, but dies without having told anyone
of it (and without it becoming known by the external world in any other way),
does it then have meaning to assert that the explorer had reached his goal
before that?” Given the premise that the explorer reached his goal, this
assertion do of course have meaning - as the premise can only be given by being
proved, and as the question is exactly about whether the statement that
describes the premise has meaning. Thus, we cannot talk about this question in
this way that presupposes not only the meaningfulness but also the truth of the
statement discussed. Only if we find so-called “evidences” or “traces” that
together with our theoretical knowledge can give rise to a fiction about the
past that is consistent both internally and with our knowledge, can we insert a
thought event at our time-axis.
Example 2. Corresponding to this example,
without any scruples an author may write a fictional story told in third person
about how a scientist makes a scientific discovery, but dies without having
told other persons about it. This omniscient account, too, presupposes
philosophical realism. If we understand such a story as a fiction, we may be
strongly inclined to accept it; but since it is based on imaginations of other
persons’ experiences and impressions, which we naturally cannot experience, it
cannot correspond to any part of reality, i.e. our possible experiences. As to
the occurrence of realism, there is no fundamental difference between this
fiction and the example above with the real explorer.
Example 3. Describing in detail in a
short-story what happens in a character's blood vessels just before this
character is struck by a brain haemorrhage (in the fiction) expresses realism,
just as the example with the distant traveller above, only in this case we are
dealing with scientific realism. First, the substantival expression “what is
happening in a person’s blood vessels” - understood as referring to an ongoing
process - only has the same grammatical
structure as the expression “what is
going on in the factory section”. However, this may seduce us into attributing
the former expression a meaning based on some idea of a possible analogy to
observable events. Instead, it has only sense to talk about a person having
been found unconscious and a medical investigation later having showed that the person in question has been struck by
a brain haemorrhage. A description like that in the short-story simply does not
correspond to anything directly perceivable, but is only a scientific construction
or an idea inspired by our knowledge of directly perceivable entities, it being
assumed that there exist not-observable entities analogously to these, and with
it that this assumption has semantic meaning.
7.1
Our situation in the world.
In
order to get on with a deeper understanding, we may try to look at the problem
from a general level. However, it is misleading to ask whether what we cannot
know because it in principle surpasses our possibilities forms part of our situation at all. For this formulation
presupposes that it has sense to talk about something existing of which we cannot know, which is realism. Put more
precisely, the question is about whether statements we cannot verify (or
falsify) concern our situation. The answer to this must clearly be that
statements that cannot be verified do not concern
our situation in the world, for if a statement cannot be decided it cannot express
any difference concerning our
situation compared with the opposite statement. For the evidence of the truth of
a statement consists in the demonstrations of the statement, and if this cannot
be done, the statement cannot be relevant either. If the statement did concern
our situation, the evidence of its truth would thus be present, namely by
virtue of these consequences, for there is no other difference that the
evidence of its truth can make up than the very demonstration of it. On the
other hand, if the description of the evidence of its truth did express a
difference this would indicate a decision. Therefore, the sentence in question
has no meaning:
Statements
that we can neither prove nor disprove have nothing to do with our situation in
the world.
7.2
Imaginations.
However,
we can understand statements about
the physical, sensuous world, even when we cannot decide them; for we can
imagine situations in which they has been proved or disproved, for instance by
virtue of our directly sensing that which is the case according to the
statement. Apparently, we can also imagine something that is impossible, e.g.
that we are standing on a distant planet and see how plants are growing on it
almost like on the earth, only with a little different forms and colours. We
think that we can understand the sentence “There is a distant planet where
trees with blue leaves are growing”, but it does not deal with any reality,
only an imaginary situation as such, as we cannot here and now confirm the
sentence or make justified fictions about how we can prove it, and to disprove
it is likewise impossible. In this sense, it has no meaning.
It
may be objected that we must explain how we can talk about certain kinds of entities,
at all, if our talk about these types of entities has no meaning. It may be
argued that when we assert that our talk about the type of entity A has no
meaning, we use the concept of this type of entity A, for which reason the concept
in fact must have meaning. At least it may be advanced that it is problematic
how we can talk about such entities in this way, for instance talk about proving
or disproving statements about them.
To
this, the following must first be remarked: The types of entities this
criticism concerns can e.g. be some that according to anti-realism only exist
as abstract entities in a theory, but which realists can assert to exist on
equal terms with the physical objects that surround us. We may be dealing with
electrons and gravitational fields, understood in some literal way. This can
exactly be due to the fact that the idea of these originates in our
understanding of the physical objects in our surroundings, as the imagined
ideas is supposed to be like these, but with different properties that make
them into entities that are not-observable, without our considering the concept
of meaning. This may result in our applying our way of speaking about observable
objects to constructed entities, as if they were observable entities. However,
this does clearly not imply that this talk has a literal sense, only that some persons
may believe that it has such a sense.
As
to other types of entities, we form some complex ideas based on other ideas
that have meaning in isolation, but whose combination has no meaning. Thus, it
is clear that we cannot meaningfully talk about a ball that is both red and
blue on its entire surface, or a dice that shows both three and five (at the
same time)[8].
However, statements can also be logically impossible without our being able to see
through this immediately because of their complexity. (This is why
simplification of logical expressions teaches us something new though the result
strictly speaking expresses the same as the original expression.)
As
to the argument against anti-introductionism above, according to that, we
cannot assert that our use of terms for any type of entity A has no meaning, as
these terms must have meaning if we can speak in an understandable way by means
of them. Against this it can be advanced that that which we do in such cases is
just to agree to use some terms, which does not presuppose the existence of
anything. By this, we just have a communication that works to some extent, for it
has already been showed how we use terms for entities that in fact do not exist.
(Cf. Treatise No 3, “Usage and
existence”.) It is only for convenience that we talk as if we refer to entities,
as we can avoid this by using some more circumstantial wordings. Thus, we do
not presuppose either that these terms have a literal meaning, i.e. that the
use of them implies existence.
In
fact, the opposite can be argued for: It is the opponent, the scientific
realist, who introduces a talk containing terms that only have a putative
reference. Before he can use these terms in a literal sense, he has to prove
that they have references, as it is impossible to talk meaningfully by use of
terms in a literal sense if these terms do not have reference or meaning. For
that reason, we can only introduce concepts whose usage we can somehow explain by
means of other concepts that already have meaning.
8. Fictions and
meaning.
An
essential question is what it means that statements about the physical world
have meaning, besides being grammatically and semantically correct and
logically consistent. Provisionally, a possible definition may be that any
statement that by use of correct language refers to our everyday experiences, i.e.
our sense expressions such as they are separately, or to a logically consistent
combination of these, has a meaning we can call fictional meaning.
However, to have a real meaning, in
addition a statement has to be able to refer to experiences that with a
starting point in our hic-et-nunc-situation
can be considered as possible in the
past, the present time or the future. Thus, statements about blue trees on a
distant planet have a fictional meaning, but no real meaning.
As
already touched on, we can also ask the question whether it is strictly
speaking possible to imagine a distant planet, or whether we can only believe
or imagine that we imagine it. We imagine possible experiences or combinations
of possible experiences and call these
prospective, but whether the combinations are consistent, we cannot estimate.
We can also imagine a thing in a smaller and smaller version, but having such imaginations
does not imply that there is a real version of the imagined object. Thus, we
have knowledge of pictures in a microscope, but not knowledge of any entities in
a form we can call microscopic. Generally, we cannot really imagine so-called
not-observable entities, for the logical reason that this presupposes that they
exist – and are observable. Perhaps we can imagine scientific constructions,
but that is problematic: Naturally, it has sense to talk about any kind of
picture we can make of scientific constructions, as they may belong to a
scientifically usable model. The question is what meaning it has to assert that
there exist not-observable objects that resemble the macroscopic ones, i.e. the
observable objects. These problems can be made clearer by the following clarification:
Imagining something is strictly speaking just to recall some connected sense
expressions - or possible combinations of these[9]. In a
weaker sense of this, these combinations of sense impressions do not even have
to be actually possible; we may only be dealing with descriptive statements without having especially
clear, attached imaginations. Thus, we can conclude:
To
imagine not-observable entities is just to imagine something observable and to supplement
this by stating either that what is imagined either has a size that makes it
not observable or is situated under circumstances that make it not-observable.
As
regards the distinction between fictional and real meaning, it can be concluded
that statements that only have fictional meaning do not actually have a
semantic meaning, because per definition they depend on combinations that
cannot be judged possible in the past, the present time or the future. Thus, fictional
meaning is only apparent meaning – in
the sense that it is only apparent that fictional meaning is meaning at all. Often,
we accept statements that only have fictional meaning just because of our lacking
overview. We simply do not always have the imaginations we think, as parts of
them are actually statements and not imaginations. Alternatively, they are not concern
what we suppose, as they are only illustrations of statements, and not of
realities. As already mentioned, we often put forward statements that prove to
be contradictory because not all the involved statements and their consequences
can be true at the same time. We often do that because the proof of these
inconsistencies may demand a careful investigation.
Concerning
meaning generally, one of the conditions for a statement to have meaning must
be that it refers to something that makes a difference compared with what is
referred to in statements that assert something that contrasts with this. Thus,
both the statement “There is ink in my ballpoint” and its negation “There is no
ink in my ballpoint” and has such a meaning, while neither the statement “There
are blue trees on a distant planet” nor its negation has such a meaning. In the
following further examples are adduced.
Naturally,
we can easily talk about the gravitational field of a celestial body, viz. the
area around this body towards which other bodies “are attracted” with a certain
force, i.e. the area in which they move with a definite acceleration, compared with
how they would otherwise move. However, a gravitational field understood as
something different from just such an area, i.e. as something real “out there”,
as if it were a body without any known sorts of properties, but added invented
properties, is an unnecessary concept, for it just does not make any difference
in relation to the opposite conception. Thus, the laws of physics can be applied
without the idea of such a supposed entity.
The
application of the terms of the constructed entities and concepts in science
can only have meaning in scientific contexts in which they are related to other
concepts in the scientific theory in question. For that reason, it can only appear
absurd to transfer the usage of these concepts to everyday concepts.
Thus,
the statement “I must certainly have a gene that makes me feel that I must read at least one novel a day”
does not have more linguistic meaning than it has to explain something by referring
to a postulated “steering spirit”. What is meant is perhaps “There must be
something about me that…” or “For some reason or other my nature is such that…”,
however the concept of genes cannot be used in this way.
Equally,
statements like “I felt how the adrenaline was pumping around in my body” are
absurd. For it is impossible literally to feel something like that in the
situation, because this would involve the experience of the truth of or rather
the applicability of a great
coherent biochemical theory, of which this concept of adrenaline and its relations
in the theory forms part. Perhaps it is possible to be aware of a special
sensation in one's body, but it is not possible to feel something that can be
expressed as “a sensation caused by adrenaline”. The reason to this is that statement
“I have an increased amount of adrenaline in my body” has the semantic meaning
that a certain measurement has shown
a result corresponding to a certain theory in which the term “adrenaline” enters
in a certain way - as far as this measurement can be made under the present
circumstances. It stands to reason that this cannot be experienced at the time
when the statement is uttered.
With
other words, it only has sense to use terms that refer to scientific concepts
if they are used in their scientific sense. For instance, if some asserts that
the person NN died because of the cause H, where the description of H contains
terms of so-called not-observable entities, this would only have semantic
meaning if the cause H could be pointed out. For this scientific talk about the
“cause” of the death has only to be understood as the idea that the case can be
classed with known, demonstrated regularities that can be described by
reference to scientific theories and observations.
Concerning
things we have left in “the room next door” or turned our backs towards, we can
have a conversation that works - also, as regards what the interlocutors will
say about their experiences. Our line of thought concerning such things also
works as to what we may later experience about them. Contrary to our
imaginations of such not-observable states, the very literal idea of things
that exist in themselves, which can be expressed in by the phrase “As if I see
them here and now”, is not crucial in this talk about things the speaker does
not observe here and now. For at any rate, the description of them has to be based
on our hic-et-nunc-situation, which
simply is at the top level in the explanation structure, the experiences we can
refer to here and now. To summarize:
At
any rate, our talk about things that we do not observe in our hic-et-nunc-situation has to be based on
what we know of the connections in which the things enter, because the description
of this hic-et-nunc-situation belongs
at the top level of the explanation structure. This level exactly concerns our basis
of explanation. For that reason, ideas about things existing in themselves “as
if I saw them here and now” play no decisive role. What matters is just whether
the terms of the things in question form part of a conversation that works.
In
the two discussed articles from Truth and
other enigmas, Dummett does not make much of clarifying what philosophical
world picture his philosophy implies, if any. In The seas of language, however, Dummett adduces that reality is not only determined by what objects exists,
but also by what statements are valid, and he concludes that the world is the
totality of facts, not of objects. Therefore, Dummett asserts, he has found the
logical concept of bivalence important in his philosophy. ([Dummett 93], p.
465)
This
can partly be compared with the world picture to which anti-introductionism
leads, such as it is described in the section “The aim of the treatise” in Treatise No 1. As it has already emerged,
however, anti-introductionism diverges from Dummett’s anti-realism. This means that
anti-introductionism implies a different world picture than Dummett’s
philosophy does. Moreover, anti-introductionism is able to reject assertions that
its philosophy of time implies “holes in reality”. For it does not maintain
that reality consists of all what our constructions may seem to imply, such as the
existence of a linear time, but that our constructions only exists as the
constructions they are.
According
to anti-introductionism, what makes up our reality is something that involves
our experiences and not something that is just out there. Thus, reality can be
understood as something that is more present according to anti-introductionism,
to which it is what we are confronted with, than it can according to realism,
to which it is something that exists in itself.
Against
Dummett's viewpoint described above, however, it must be remarked that the
metaphysical status of what we call facts is uncertain, as the concepts of
facts seem to have a linguistic character. Certainly, facts cannot be sentence members,
as it would be a category error to assert this, though we can say, “It is a
fact that the table is placed in the room next door” or “The table is placed in
the room next door. That is a fact”. On the contrary, what is said about the
world, i.e. the semantic content of the sentences is rather what can be called
facts, i.e. neither the sentence members nor parts of the world. Thus, facts
are not what we can point at: a bucket is not a fact. We can say, “It is a fact
that this is a bucket”, which just corresponds to the use of the word “truth”.
8.1 The past.
Realism
concerning the past is based on an idea of a past that resembles the present
time - corresponding to the realist idea of physical objects that exist in
themselves and resemble those we in fact can see.
In The Seas of Language Dummett asserts
that we are inclined to think that statements in the past tense are made true
or false by virtue of a reality to which we do not have access any longer -
except perhaps partly by means of the memory - though we think, however, that
this reality still exists in the past somehow. ([Dummett 93], p. 60)
This
can be contrasted with the anti-introductionistic view concerning past and
future, which in Treatise No 1 is justified
by considerations concerning what we can at most assert. The results can be
summarized thus:
We
only need to base our concept of the past on what exists now, such as memories
and finds, and not on any past in itself. It is entirely because of our
experiences now that we can talk about a past. The time axis is just one of various
means of talking about what exists now, on equal terms with other means within
other areas. Thus, when we talk about the past, what we are talking about is a
special arrangement of what are present (including our knowledge and our
memories) and a relating of this to the mathematical entity that a time-axis
is.
Therefore,
any reference to “history as a description of our origin” is not relevant as
regards any aspect of reality. For when we cannot remember these historical
details, or - less realistically expressed - when we cannot conjure up that
sort of memory impressions, such a reference presupposes a reality whose
existence we cannot prove. Thus, we are dealing with a continuity that is at
most a constructed conception. For that reason, these hidden and forgotten
circumstances, or rather again, the absence
of these memory impressions is without importance. Yes, according to this view,
referring to historical details understood as real entities rather becomes
completely without meaning. It is possible to be grown up in London
and nevertheless feel literally at home in Rome
So-called
historical founds can be systematised and related to each other in the making
of fictions about an imagined past, for which they can be used as a source of
inspiration. Such a past does not have more reality than fictions about the
future. Based on the regularity we observe in nature, we can relate the description
of certain events to definite points on the time-axis. We do that with an
accuracy corresponding to how sure we regard this regularity to be. It is worth
mentioning that this is not the denial of the past as a reality, but a description
of what kind of reality we can
attribute to the past, or rather how to interpret statements about the past, or
even more precisely, interpret statements that contain apparent references to
the past understood as a former present. In Treatise
No 1, “Time”, ethical arguments against consequences of this view have been
refuted.
Concerning
the sciences that deal with past events and diachronic courses (history,
theories about evolution, palaeontology, and cosmology), we only need to understand
them so that within each area they are occupied with fixing justified descriptions
or statements at an abstract time-axis, or with other words, relating these
descriptions statements to points on a time-axis. Objections that the
statements in question are nevertheless about something real do not have sense.
The reason is that this idea has no consequence to the activities of science,
and thus, it is without reason and not least useless. Consequently, as already
touched on (Treatise No 2, “Differences
between past and future”), it has no scientific relevance to talk about a
specific point on the time axis attached a time t0, e.g. five years ago now, such that statements are
about something more real when they refer to points with higher values than t0 than when they refer to points
with lower values.
Concerning
the possibility of a reductive interpretation of the past, in the section
“Realism” Dummett asserts that there neither is a reductive class of statements
about the past nor about the future. Dummett substantiates this with the viewpoint
that a memory cannot be characterized independently of what it is a memory of,
and that an intention cannot be characterized independently of what it is an
intention of doing. ([Dummett 78], p. 157)
As
far as I understand this, Dummett’s line of thought is that statements about memories
are the only conceivable possibility of a reductive class of statements about
the past, but that this possibility must be rejected, as every talk about
memory involves this talk about the past that the reduction just have to
remove. As it is true of the memories and the past, it obviously also applies
to the intentions and the future.
However,
we do not need to assert that memories are memories bout something, such as
Dummett assumes in his argumentation: Our descriptions of memories do not have
to be descriptions of a literally existing past but of our memory impressions
themselves. Therefore, a memory is not characterized by something it is a
memory of, but by how it looks. Our
usage by description of memories is the same as by descriptions of what we
experience now, as the former looks like the latter.
Moreover,
anti-introductionism concerning the past is not about reduction. (Here it must
be assumed that it is possible to use the term “the past” without being a
reductionist for that reason.) Instead, it aims at clarifying whether we attribute
too much to reality and, in that case, how we are instead to understand our
talk about the past. Thus, earlier in these articles, without use of
reductionism it has already substantiated that we cannot state any reasons for
attributing to the past an existence in its own right of any kind. (Cf. Treatise No 1, the section “Time”.)
As
an elaborated argument for this view, it can be advanced that we cannot at all
refer to experiences of situations in which we literally move the past forward
and live in it. Such situations would certainly be analogous to situations in
which we bring to the light physical objects that are out of sight behind other
objects and thereafter look at them, but we do simply not possess any memory of
such kinds of experiences. In contradistinction to memories of moving objects
forward, memories of literally moving the past forward are thus not a part of
our life.
As something
secondary, then, the only thing left of our talk about the past, is a concept
of a past according to which our use of the term “past” is just based on a
construction based on memories and finds. In addition, as just mentioned, we
cannot say that memories are literally memories of anything, only that they are memory impressions with a memory
content here and now that can be subordinated known regularities. In a similar
way, the future is only a fiction. The fact that an intention is an intention
of doing something definite in the future can be said to be true, viz. in the
sense that it only concerns a fiction,
which we call “the future”. Thus, we
cannot point out any future event to which it can be related; it simply has no
meaning to talk about a future understood as new nows at all.
8.4.1.
Past and learning.
In
“The Reality of the Past”, Dummett involves the concept of learning several
times in his explanations. Thus, he puts forward that we learn to use the past
tense by learning that the assertion of certain statements in that tense is justified
by certain situations. These situations may be situations in which we remember
the occurrence of an event at which we have been present. ([Dummett 78], p.
363)
However,
this talk about previous learning presupposes that we corresponding to each of the
faculties that we may possess and can be said to have learned, have memory
impression that depicts learning situations in which we acquire this faculty. This
does not apply to all cases, for instance not to our use of the past tense. Talking
about a not-remembered, but existing learning situation (possible records are
hardly available), just as Dummett does, presupposes realism - analogous to how
Berkeley's idea of our construction of the concept of thing presupposes it.
Thus, I do not have any memory of “how I learned to use the past tense”, just
as I am not conscious of my somehow constructing my sensation of a certain
thing in the way that Berkeley describes it. In the same way, this part of
Dummett's explanation is based on a pure construction or fiction. The fact is
that here and now, I do not refer to some kind of learning that I remember when
I use “the rules of the past tense”, not even if I am able to expound the rules;
as already touched on, it has no meaning to say that they are not-conscious
(so-called unconscious). Perhaps educationalist can investigate in children's
learning and work out usable theories about a constructed childhood, but that
is something different.
As
mentioned, it expresses realism to talk about the remembered event as having
happened, if that talk has to be understood as referring to more than just the
very memory impressions and the relations into which we let them enter - which is
less than what is literally implied by the talk about the fact “that the event
has occurred”. Thus, we can say about our memory:
If
I assert that I remember to have been witness to an event, this can only be
literally understood as my having memory impressions representing my being
witness to the event, in which case, however, these memory impressions must not
be inconsistent with other memory impressions and facts.
Dummett
concludes that the only concept of the truth of statements in the past tense
that we could have gained in our practice
of using such statements consists in the possibility of justifying them, i.e.
in the existence of recognizable situations that justify these assertions.
([Dummett 78], p. 363, my underlining)
Here
too it must be emphasised that this talk about training is neither necessary
nor successful as also this presupposes the reality of the past, which is
circular in a discussion about realism - in particular concerning the past. The
fault consists in the very reference to previous learning, which presupposes
both 1) that we have already clarified what time and especially what the past
is, and 2) the idea that this clarification results in a justification of this
talk, i.e. a realist result, which however is what we have to prove, and thus
not given. For if we are only justified in talking about relations between our present knowledge and our memory impressions,
there is nothing else to refer to than our cognitive situation here and now,
including our knowledge. Thus, we cannot talk about the past as an entity in
itself and, with that, we cannot talk about a learning process either, as if it has meaning to assume that it has literally
existed in the past. This talk is semantically meaningless. As mentioned,
however, this does not mean that we cannot talk about an arrangement of memory
impressions and past fictions that motivate us to making assertions about the
past.
It
is not necessary at all to refer to the
concept of learning in the realism debate about the past, as statements
about the past can be understood analogously to statements about physical
entities. For in both cases, the truth conditions of the statements consist of
verifications that are based on our knowledge of the regularity of nature and
reliable memory impressions; these verifications have to be present in the now.
Certainly, the realist may still protest by asserting that there can be other facts
than those we can know about. However, that presupposes the meaning of talking
about a reality existing in itself, and thus realism. As to that, in this treatise
it has already been argued that the meaning of statements about circumstances
we will never know about is absent.
From
what is just referred, Dummett concludes that we are not justified in asserting
that a statement about the past is either true or false independently of our
possibility of knowing its truth-value. Dummett’s reason is that since something
that justifies such a statement can later emerge, we cannot either say that the
statement is neither true nor false. ([Dummett 78] p. 364) Here, an assertion
about the age of the last existing dinosaur must be an example of statements
that can never be declared true or false, as we do not have possibility of
knowing its truth-value. However, such an assertion is also a counter example against
the other of Dummett’s two conclusions, as nothing capable of justifying it can
ever emerge. Therefore, this conclusion is not quite general.
Our
statements about the past can be based on our present knowledge and not on a
concept of past training, because that concept is usable in this contest, as
mentioned. However, Dummett advances a view that is far more in agreement with
anti-introductionism than the above referred views: According to Dummett, it is
a fact that that we are immersed in time, which means that we cannot describe
the world such as it would be to a person that were not in time, but only such
as it is now. ([Dummett 78], p. 369)
True
enough, also this formulation and especially the use of the metaphor “to be
immersed in time” expresses a realist conception of time, as it is conceived as
a continuity of which some exists before us and some after us, in the same way
as the now - for how else could we be “immersed” in time? Apart from that, however,
in a reasonably way the description expresses that aspect of our cognitive situation that our point of departure of
understanding the world is the now. We can only assert that here and now we
have knowledge of different sorts that originate from different places, or
rather, knowledge we can relate to different kinds of conditions or circumstances.
It must be remarked, however, that Dummett’s talk of being in time suggests that
the now is a part of time, i.e. that time is more the now, which expresses realism.
A
more clear expression of realism consists in Dummett’s assertion that to the
anti-realist the past only exists as the traces it has left to the present,
while to the realist it still exist as past. ([Dummett 78], p. 370) What is
wrong is not that Dummett only speaks about one of many realisms about the
past, but that he is not thoroughly anti-realistic here. In fact, his talk presupposes
a realism according to which it has sense to assert that the past has once
existed in the same way as the present time exists now, while only the
consequences of the past have been left as “traces”. Thus, this talk about the
past having left traces clearly expresses realism. However, it is certainly
difficult to avoid realist formulations completely.
9. The correspondence
theory of truth.
The
correspondence theory of truth can be described like this: “Statements are true
just if they agree with reality”. However, it is problematic what it means that
a statement agrees with reality:
We
cannot meaningfully our talk about the truth of a statement, purely and simply,
as talk about the truth of a statement demands specific reasons for its being truth.
This involves our clarifying its truth conditions and how they can be satisfied.
For
that reason, it is immaterial whether we talk about the justification of the truth of a statement or just about the justification
of a statement, for we cannot put more
into the former concept than into the latter. Justification of a statement is
justification of its being true. Thus, the concept of truth is not a
fundamental concept, but at most the opposite of falseness. Instead, the
concepts justified and unjustified are more fundamental than
that of truth. (Rather than just talking about justifying, we can also talk
about pointing out or proving. This is not crucial.) It must be noted that here
we are dealing with persons’ assertions, not with what can in fact be said to
be true. While there is no difference between asserting the statement p and asserting that p is true, there is a difference between
asserting p and asserting that p can be justified. For a person can simply
put forward a particular statement, P,
without asserting, “P can be justified”,
which he must do if the statement are assumed to be interesting and not random.
(P may be the statement “a > b”
concerning certain terms a and b.) On the other hand, the statement “P can be justified” implies the statement
P.
In
addition, it is evident that the statement p
only can imply “p can be justified” provided
that p can be declared meaningful.
As
we cannot meaningfully talk about the truth of statements purely and simply, we
cannot either talk about reality purely and simply, in so far as it is understood
as something other and more than what we can tell of with justified statements.
Thus, we can assert:
The
only kind of reality whose existence it has sense to assert is that which we
can refer to in our justified statements.
If
we instead of talking about possible assertions talk about the logically more
interesting subject what our assertions commit us to asserting, then the
statement “p is true” (A) is
equivalent to the statement “p can
be justified” (B); for our assertion of A commits us to assert B, and B implies
A. This means that if we assert the correspondence theory of truth, “p is true (A) just if p agrees
with reality (C)” we are committed to asserting the statement “p can be justified (B) just if p
agrees with reality (C)”[10]. This
means that we cannot put more (or less) in the statement about p agreeing with the reality (C) than in
the statement about p being justifiable
(B). Put in a different way: asserting the statement “p agrees with reality (C)” is the same as asserting the statement “p can be justified” (B). Thus, if we
can assert that a statement P is
true, this is not because its agrees with a reality that exists purely and
simply - as if this idea has sense at all - but because P can be justified and
thus agrees with what is present. The conclusion is:
We
cannot attribute meaning to the concept of some “reality in itself” which a
statement agrees with. What we can at most put in our talk about the idea that
a statement “agrees with reality” is that the statement can be justified. We
can only do that by means of our experiences and argumentations. Thus, the only
reality that exists is the experiences we are dealing with, including the
regularity we can demonstrate.
This
also applies to the theoretic constructions of science, which are ultimately
based on our experiences.
Examples.
Example 1: When does the statement “NN has a
book written by MM” correspond to reality, in fact? It does that, for instance,
if NN can show such an exemplar in his home. When does the sentence “NN has
returned a certain book to the library” correspond to reality? It does that, if
we have seen the delivery. On the other hand, it is not sufficient that the
book can be found at the library and that it is recorded as delivered by NN,
because the book could have ended up and been recorded at the library for different,
unknown causes (evidences of which could emerge). Even if we would say that the
statement also agrees with reality if NN can show a receipt for the delivery of
the book, we do not in this case experience the very reality concerning the
delivery with which the statement has to agree, viz. the very delivery of the
book. Our knowledge of the regularity of nature and the procedures at the
library does allow us to regard this
as an agreement. However, the very
act of returning the book to the library we have not witnessed just for that
reason.
Let
us assume that the book is not at the library, that no delivery has been
recorded, and that NN does not possess any receipt for the delivery. Under what
conditions does the statement “NN has delivered a certain book at the library” (P)
then agree with reality? As there are no evidences or so-called traces, only NN's
memory or his description of it and possible testimonies are left. Some realists
will simply answer, “Under the condition that NN has delivered the book at the
library”. However, this line of thought presupposes the reality of the past,
and in Treatise No 1, it has been
disproved that this concept has meaning. At least it has been argued, that such
a talk is senseless understood as reference to a former, literally existing present,
and it has already been clarified what we can at most put into it. There cannot
be attributed meaning to the concept of a literally existing past, just to the
concept of a part of an imaginary time axis to which events and states are
related to each other by means of memories and so-called evidences. Primarily,
however, the realist's answer is useless because it is only a paraphrase of the
referred statement based on the formula: “When is P true?” - “If P”. Therefore,
this is not an answer to the question. The question is about what has to be present
in order that we can assert that the statement agrees with reality.
Instead,
we can try to answer this question: “When do we have reasons to relate a
description of an event (P) to the constructed time axis?” To this, we need
memories, possibly present observations and our knowledge of the regularity of
nature and so-called evidences, in this case as regards receipts and the
function of the registration systems of libraries. As regards such invented
systems, however, we are at most dealing with so-called good reasons, which are not actual verification. As to memories of
the delivery, as mentioned, we can say that they make up correspondence to the
reality.
Thus, we cannot point out anything real with which
the disputed statement agrees, except in the simple case where such a part of
reality is directly present (the delivery of the book or our recollections of
it); the rest is constructions. At any rate, the verification takes place in the
present time. This is also true if the evidence is based on recollections or if
it exists on an available record. For that reason, we can conclude:
The
justification of statements about the past exists now, as true statements cannot
contain nothing but what can be concluded on the basis of present evidences.
With
other words, statements are true because of what is present now. For the same
reason we can even infer that the justification of any kind of statement about existence is present in the now.
Example 2: In order to verify the statement “The
acceleration due to gravity is 9.806 ± 0.002 m/sec² at the place S on the Earth”,
we can measure the acceleration of a body at the place S. If for instance the result
is 9.805 m/sec² the statement is true. This may be said to correspond to
looking after a certain book at a certain place in a bookcase, where the act of
“looking after” corresponds to making a measure. (In fact, this measure may be
complex, but that is not the important here.)
Unlike
the discussion in the library example of turning conclusions based on observance
of regulations into evidences of past acts, the acceleration due to gravity is
defined in such a way that there is a direct correspondence between the meaning
of the concept and a number of certain measurements followed by well-defined
computations. These specified acts constitute the reality with which the above-mentioned
statement about the gravitation due to gravity has to agree in order to be
true.
Though
these acts are observable, there is no (we cannot point out any) definite part
of the surrounding reality that can be called acceleration due to gravity and
with which this statement has to agree in order to be true. (Cf. also the characterization
of the gravitational field above.) According to this account, the found size of
the gravity acceleration is an abstract result based on a number of definite
acts. In spite of this, scientists may take the acceleration due to gravity as given when they make computations in
other connections, which may lead to our talking about the gravity acceleration
as if it existed in the same way as the air around us. However, this does not
change the conclusions above, and as argued in Treatise No 3, our usage is not ontologically committing. This also
applies to statements about electric charge; cf. the description in Treatise No 3.
10. The
importance of the top down line of
thought.
The
central point of anti-introductionism is the question what can at most be said
to be among the constituents of the world, and especially in what way it can be said to be it, if so. A part of the basis of
this semantic inquiry can be bases on the top down-view, according to which we
are in a clarifying situation. Usually we only perceive an object as a whole, when
we perceive that a thing of a certain kind is present, a ball or something
else. Then we can clarify this whole by describing our sense experiences and
their relations to each other, empirically and rationally, including
mathematically. Thus, the top down-view
involves the idea that our explanations have a starting point in our
experiences and arguments, and that they can be clarified in such a way that
they involve still more accumulated premises, that can be structured by means
of precedence relations of the form “This theory, T, is based on the theories T1, T2,... Tn and the experiences E1, E2,...
Em by means of the reasoning
R”. These experiences may for
instance be sensations of our body, including our sense organs. In this way,
the theory may be about the relation between our auditory organs and our
hearing as described in Treatise No 1,
“Our body”.
Here,
it was exemplified how, according to the top
down-view, only our experiences and reasoning are the main things in our
explanations. The descriptions of the theories at the top of the structure are made
on the basis of these. On the other hand, we can collect related statements and
theories in conglomerates of explanations further down in the explanation
structure. In the precedence relations, we can thus restrict ourselves to referring
to these conglomerates by their names.[11] Above,
Dummett concluded that “the reductive class” is unnecessary to reject realism. ([Dummett
78], p. 159 and 361/2) This
also follows from the top down-way of
reasoning but in another way, because our first perception of a material object
– e.g. the impression of a blue ball - is simply the most primary thing to which
we can relate. Thus, the concept of the ball belongs at the top of the explanation
structure. In this way, it is our impression of wholes that is the primary
thing in our sensation, not our impression of their parts. Therefore, there is
no question of reduction. Instead our detailed conception of what a blue ball
is (that is our explanation of how a blue ball can be characterized as a blue
ball), follows further down in the explanation structure. Below this level
follows the description of how the detailed characteristics of the blue ball
can be related to other circumstances, such as scientific theories. As
suggested above, this means that we are not dealing with a Berkeleyan
construction according to which a whole is understood as the result of piecing
details (different sense impressions) together to a whole, but with an
elaboration that goes from a whole to its details. In addition, we can involve
our knowledge of the regularity of nature already at the top levels of the
explanation structure. These levels contain descriptions of our observations or
sense impressions of objects and their regularities. Thus, we know that if we
put the ball into an empty desk drawer and close it, we can find the ball again,
and that if we pull the drawer, we may hear a sound like the rolling of a ball.
Thus, references to the whole thing
are situated at the upper level in the top
down structure, as the perceived object is a starting point of our
cognition when we under ordinary favourable conditions perceive it as an object.
In
the same way, the fact that a stone falls towards the ground belongs at the top
of the explanation structure, and only below there it follows how empirical
investigations lead to scientific theories about the fall of physical objects. Further
down, below this level of the structure, a closer description of these theories
of gravity follows, e.g. in relation to a general rule concerning the force of
attraction between two masses. It follows from the exposition above that the
explanation structure forms an arranged network in which not only simple entities,
but also parts of structures are related to each other.
The
details concerning the explanation structure are philosophically secondary; in
this connection, the most important thing is that the explanation structure has
a number of superior points of departures while the rest are constructions.
Thus, sounds, or more precisely our sound impressions, are real. For it is a fact that we receive impressions of
sound, whereas many of the concepts, such as our concepts of vibrations,
electric voltage, and microscopic details, that form part of the theoretical
explanation of sounds and our hearing only form part of theoretic constructions
based on other sense impressions and on reasoning. These constructions may be
said to have a reality understood as useful constructions but they have no
reality that can be put on an equal level with our ears or hands or with the
sounds that we hear.
According
to the top down model, the source of
the justification of the truth of a statement consists of something about which
we know, because the truth of a statement is simply implied by the facts we know, or rather by 1) the statements that
describe our experiences, and 2) what we have already concluded. It is our
knowledge and our reasoning that makes us express statements we can justify. As
argued for above, we justify
statements by means of facts, wherefore they can be said to be true because of
facts. Thus, it is not so that statements in
themselves possess a truth that we possibly later can find by means of
other facts. This would be an empty allegation, as even the adherents of the bottom up model have to conclude top down.
It
has already been put forward (Cf. Treatise
No 2, the subsection “Anti-introductionism”) that the assertion of the truth of a statement requires the involvement
of our current possession of
evidence, while the condition that a statement can have meaning only has to include our possible identification of evidence.[12]
Thus, existence statements only have meaning
if we know that we are capable of justifying
them (or their negations), while only existence statements that are in fact justified can be declared true, and only as regards their
verification procedure.[13] Even
if a meaningful statement is not verified, we may nevertheless talk about the
situation that would be present if
it proves to be true or the one that would be present if it proves to be false.
As
mentioned above, in a wide sense the top
down reasoning deals with the description of the construction of our
theoretical understanding, which implies involvement of still more details (as it
is described in Treatise No 2 in “Top down and bottom up explanations”). The top
down reasoning does not need the realist's assertions about things in themselves,
as according to the top down reasoning
we are only dealing with observations of objects and their regularities. From
these we draw conclusions about the relations by means of which the behaviour
of the phenomena can be described. (The theoretical explanation structures we
create can for instance be descriptions of cosmological theories or theories
about the elementary particle structure of matter.) The introduced concepts presuppose
and are explained by reference to our sensation, thought and possibly already
introduced concepts (about which this also applies).
Of
cause, it does not have meaning to talk about introduced concepts before they
are introduced, and they cannot be attributed different meaning than what is
implied by their explanatory conditions, i.e. what has led to the introduction
of them. Thus, they remain just ideas or useful concepts.
What
is true of the concepts also applies to their actual occurrences or
applications, of course.
As
mentioned in the section “Introduction”, perhaps the anti-introductionistic top down-philosophy is most of all inspired
by Berkeley's empiricism, in spite of the reservations mentioned in Treatise No 1 and above, and though some
of the wordings in this text are more or less influenced by those of Dummett. In
relation to several fundamental questions about existence, however, the anti-introductionism
does not take the concept of evidence as a starting point - in contrast to
Dummett's definitions of anti-realism and realism. (Cf. e.g. [Dummett 78], p.
146.) These questions about existence deal with the semantic meaning of terms
for the so-called “not-observable entities” in science, such as elementary
particles and the human being's detailed way of functioning at a micro-level.
Instead, it has been focused on what we can at most put into these concepts.
As
regards statements about entities observable in principle (entities of a not-fictive,
already known kind), the condition of their possession of meaning is that they
must be able to be directly verified. This applies to statements such as “A
bust of Beethoven is floating about in a distant galaxy” that refer to entities
that are observable if they are presented to us, i.e. observable in principle.
We can only meaningfully talk about such a statement if, based on our present
knowledge, we are able to construct a fiction about a number of acts that in
principle can lead to observations that justify us in asserting that the
statement is true according to its verification criterion.
The
realism debate about entities observable in principle is not about “whether an
object exists when it is not possible to observe it”, but about what meaning we can put into statements about such an existence. For
we cannot ask the question about existence before the question about meaning has
been answered. Only then does the question about existence have meaning.
For this very talk about “whether an object
exists when it is not possible to observe it” presupposes that this talk has
been semantically clarified, such that it has literal meaning to say the object
in question may exist somehow. Instead we can talk about whether it has meaning
at all to say that at a certain place
there is an object with certain qualities, when it is not possible to observe
whether such an object exists at that place - and which meaning this talk
should have in that case.
11.
Truth and semantic meaning.
The
definition of realism and anti-realism that was criticized in “The semantic
approach to the realism debate”, and according to which realism is the point of
view that the external world exists independently
of consciousness, while anti-realism is the opposite view, is formulated on the
basis of the realist's point of view. For this definition presupposes that the
question of what we can say about the external world is a question of existence
versus not-existence - or “lacking” existence, “holes in reality”, etc.
Moreover, this definition presupposes that the anti-realist necessarily has the
opposite view of that of the realist, which must be the view that the external
world does not exist dependently of
consciousness, and that this idea has semantic meaning. In short, this formulation
presupposes that the anti-realist can accept the realist's view of the problem
and his terminology. As a secondary comment, it must be pointed at the obvious
truth (for which Berkeley has been quoted in these texts) that consciousness
does not determine the content of the physical world, but that, other things
being equal, it is subordinated the phenomena, including their changes.
Here,
it must also be emphasized that of cause lack of meaning of a statement about existence is not identical to lack of
existence as some realists may thinks, for the talk about that presupposes that
the existence statements has meaning. The realists’ reason to this may be that
they understand anti-realism on the assumption that realism is true and in this
blend of realism and misunderstood anti-realism a contradiction is easily
found. These realists commit here the fundamental error of making familiar what
is unfamiliar instead of letting what is unfamiliar remain unfamiliar, as long
as it appears unfamiliar. By this, they defend themselves against an understanding
of anti-realism that is a condition of being able to argue correctly against
it. Cf. Nietzsche and Bergson's erroneous counterarguments:
In Jenseits von
Gut und Böse, Friedrich Nietzsche adduces that sense organs cannot
be phenomena in the sense of idealism, and that in that case they cannot be
causes. He argues that if the surrounding world is created by our organs - our
sense organs, I suppose - these have also created our body including
themselves, which is a contradiction. ([Nietzsche 88], p. 20-1)
Obviously,
Nietzsche presupposes a realist concept of cause. This is not necessary,
however. We only need a concept of reckoning events among described regularities,
which is an anti-introductionistic concept of cause. Clear contradictions like
the one Nietzsche arrives at, ought to make anyone that “discovers” them
reconsider whether he has understood his opponent's point of view quite
correctly. In this case, the contradiction is based on the realist bottom up-idea that we sense by means of
our sense organs understood as something basic, these being regarded as
something material in a realist sense. However, this is not one of the premises
of idealism, and in these premises, Nietzsche and other opponents have to find
a contradiction in order to refute idealism. Expressed in
anti-introductionistic terms, reference to our sense perception is primary in
the explanation structure while reference to our sense organs is secondary in
this structure. This subject has been touched on in Treatise No 1, “Our body”.
When
Henri Bergson in Matière et mémoire asserts that Berkeley's philosophy not only
made it impossible to talk about the success of physics, but also made it
necessary to regard the metaphysical order in the universe as a pure
coincidence ([Bergson 88], p. 2-3), this may be because he has the same realist
concept of cause as Nietzsche. At least it can be rebutted that Berkeley
Of
course, anti-realism does not need to be consistent with parts of other world pictures,
including realism - on the contrary, it implies a different world picture than
realism does:
The
statement, that a certain statement is meaningless, does in fact imply an
important statement about our surrounding world, namely the statement that our
situation in the world has such a nature that the statement in question cannot
have meaning. This means this if the statement S expresses that an entity of a certain kind exists, the statement
(P) that asserts that S is
meaningless is a stronger statement than the statement (Q) that asserts that S is false. - Of course, this does not
imply that the former statement (P), that S
is senseless, implies the latter statement (Q), that S is false, and possibly something more. It simply deals with the
fact that every statement about existence or not-existence (as e.g. Q) presupposes
that the statement in question has meaning.
This
means, as already touched on, that if what a statement expresses in any respect
does not refer to a difference that can be pointed out, then the statement in
question is meaningless. This applies not only to statements such as “Possible
worlds exist”, but also to statements according to which objects in the room
next door exist “in themselves”. Such statements express no difference compared
with statements in which the starting point is our experiences, our memories,
our knowledge of the regularity of nature, and our conclusions. Understood as
statements with a real meaning they are not usable. In support of our thinking,
we certainly can have specific imaginations about the existence of unobserved
and unobservable objects with the same qualities as these assumed objects would
show if they were present in front of us. What we do in that case is to imagine
that we are experiencing such objects in front of us. Of course, this cannot
imply existence.[14]
However, the realist will put forward that we are not only dealing with
imaginations, but that it is a question of factual existence. Before he can justify
this, however, he has to expound what meaning the philosophical statement he
will justify has, i.e. what difference
the possible truth of the statement should express compared with its falseness.
As
regards objects that are in principle observable, the realist will thus assert
that a statement as “The object A is at the place B” has a truth-value also
when it is not possible to observe whether the object is there. However, this
point of view demands a clarification of the meaning of the actually disputed statement
“The object A is at the place B also when it is not possible to observe whether
the object is there”. To this, as already touched on, the realist will put
forward that this statement means that in any circumstances the object A is at
the place B with the same qualities that it would show if it were in front of
us. However, this was exactly what he should expound as being more than just an
imagination, whether it is useful or not. The question deals with the meaning
of the statement:
The
realist can assert that he can explain a crash from the room next door by
referring to a big painting hanging in there in an old string, which is
possibly worn over just now. However, the anti-realist can also assert this, as
his usage is not ontologically committing; for concerning this sort of
verifiable statements, the realist does not have a patent on language: we talk
in that way. Certainly, the realist will put more in his explanation, but
ultimately both parts can explain their statements by references to the
regularity of nature. The discussion deals with this extra the realist will put
in his explanations. For as to the painting that is just seen in the room next door,
according to the realist as well as the anti-realist it will indeed be true to
say “The painting is in the room next door”. However the realist will, in a
philosophical context, add “But the painting is there also in the same way as a painting in this room is here” - whatever the
expression “in the same way” means.
The
anti-realist can assert that what exist are our present sense impressions,
memories, thoughts and knowledge about e.g. the regularity of nature. We do not
need more to expound our experiences. - The realist's extra statements do not
concern anything of this, but express an assumption that does not refer to something
that can make a difference.
It
is obvious that anti-introductionism expresses or implies a different
understanding of the world and of what it means to be in the world than realism
does. This view implies, to a higher extent than realism, that clarification and
knowledge are important as to what statements about existence we have reasons
to assert.
According
to the top down model, the
understanding of a statement containing the term of a certain type of entities
demands knowledge of what may be expressed about it at the above levels in the
explanation structure. This consists of the precedence relations of the
concepts and of the statements, i.e. references to the concepts and additional
statements that successively must be understood in order to understand the statement
itself. As regards the realism debate, these demands are philosophically relevant
to the semantic clarification of statements about particular entities and their
existence, as the talk about these exactly presupposes this clarification.
12.
Bas van Fraassen's constructive empiricism.
The
essential thing in the philosophy of science, at least in relation to the realism
debate, does not deal with the methods of science, but with what we can at most
put in scientific statements about existence. This is meant to complement what has
been advanced about the latter subject in Treatise
No 1-3, and relate it to Bas van Fraassen’s philosophy of science such as
he presents this in The Scientific Image
([Fraassen 80]). The results
of these considerations have implications as to what we can meaningfully say
about the purpose of science.
12.1
Constructive empiricism and literal truth.
According
to Bas van Fraassen, the anti-realist's point of view is that when the scientist
presents a theory, he does not assert that the theory is true. Instead, “he displays it, and claims certain
virtues for it.” ([Fraassen 80], p. 9-10)
From
this, it follows that van Fraassen - as he himself declares - endorses the
latter of the following two groups into which he divides scientific
anti-realists, viz.:
1)
Those who think that science aims at being true understood in the right way.
2)
Those who think that science has to be literally understood but that it does
not need to be true in order to be good. ([Fraassen 80], p. 10)
Against the latter view, which van Fraassen
shares, the decisive objection must be advanced that scientific statements cannot
be literally understood, if they contain terms for in principle not-observable entities.
For in these anti-introductionistic treatises it has been argued that it is
impossible to substantiate a literal understanding of statements about such
entities and their existence.7 Thus, there simply does not rest
anything substantial to put into the concept of a literal understanding of the theories, when the concept of literal existence of entities that are in
principle not-observable actually does not have any meaning. As it will be commented
on later, the only thing that remains of this at all is just some ways of
speaking that work to some extent.
In
addition, Bas van Fraassen asserts that the purpose of science is to propose
theories that are empirically adequate, i.e. true as regards what they say
about observable objects and events. According to van Fraassen, this is exactly
true of a theory if it is in agreement with the phenomena, and acceptance of a
theory involves only the assumption that it is empirically adequate. van
Fraassen calls this view of science “constructive empiricism”. ([Fraassen 80],
p. 12)
This,
I suppose, can only apply to the experimental
sciences and to experimental subsections
of other sciences. In the different diachronic sciences, such as history, evolution
theories, palaeontology, and cosmology, there are only limited possibilities
for experiments and only in subsections of these sciences, but not in the theories
as a whole. In this sense, there is only one history, though there can be several
interpretations.
The
diachronic sciences can be said to deal with construction of fictions that, contrary
to literal fictions, are completely bound by evidences and theories in such a
way that they both 1) has to be brought about by these, and 2) must not be in
contradiction to these. Point 1) means that even though a fiction is not in
contradiction to the evidences it must not contain not-necessary elements.
Moreover,
van Fraassen maintains [1] that though we do not understand the scientific
statements as literally true [2] we do not need to think that the entities that
exist according to the theories are real. ([Fraassen 80], p. 11-12, my
underlining and numbering.)
As
mentioned, the former idea (1) is in contradiction to the anti-introductionism
as such statements are semantically senseless. By contrast, the latter idea (2)
is in accordance to the top down-principle,
according to which science analyze and describe the phenomena as the primary
reality. This reality belongs to the physical world, while the helping
structures belong at a lower place in the description structure – but only as helping structures in the form of a
theoretic set of concepts and not as something primary.
In addition,
it must be emphasized that not only do we not have to think that the introduced entities exist. For in the
critique in these treatises of the extended use of the principle of inference to the best explanation, which has been
referred to several times, it has been argued that it cannot have meaning to
assert that such entities exist either. According to the top down-principle, they can only be constructions introduced with
the purpose of describing the regularity of connections between events. For
these reasons, it cannot either have any meaning or be interesting to seek after
an underlying reality in a literal sense, but only to search for usable
theories. However, this is just what results from empiricism in general. As
regards van Fraassen point of view, it just seems as van Fraassen he refrains
from taking a position to the above-mentioned problem about the semantic meaning
of scientific statements, for some reason or other.
As already done (in Treatise No 3, “Usage and existence”), however, it can also be
argued that we are not ontologically committed by our way of speaking. For this
reason, scientists can speak about electrons and force fields and (in
scientific contexts) about the existence of these without having to assume that
they actually exist, or that this talk has semantic meaning at all. However, it
only has meaning to talk in this ontologically non-committing way when the use
of these terms has a scientific purpose and thus a linguistic meaning. This is
why the way in which some persons talk about for instance adrenaline in the
body outside the scientific coherence is grotesque, cf. the already given
examples. Moreover, when some schoolteachers teach children that we consist of
cells that consist of molecules that consist of atoms, etc., this indicates a naive
and misleading view. This is certainly scientific concepts, but what is interesting
is their usage.
The
view that not-observable entities may exist has already been refuted as linguistically
meaningless in these Treatises.7 For that reason, talking about “the underlying
structure of reality” has no sense; therefore, the only physical thing that may
be said to exist is the phenomena. Even though for this reason there cannot be
any intellectual interest in the misleading metaphor “the underlying structure
of reality”, it is nevertheless possible to find the physical theories thought
provoking of cause. The interesting structure that science is dealing with is
therefore not an underlying reality, but simply consists of the
scientific analysis and description of the phenomena.[15] Therefore,
the structure of these is neither underlying nor hidden; on the contrary, it is
present in the shape of scientific models. Talking about anything underlying in
a literal sense does not have any meaning at all, for the only thing we can observe
and point out is the phenomena and their regularities (however subtly they may
be described). This is not due to a deficiency in our construction but to the
fact that observing something is to sense something and to make out the structure
of our sense perceptions. Thus, if we could point out something underlying, for
this very reason, this would not be underlying but present, exactly in virtue
of its being pointed out. Thus, observations are sensations, and the content of
the latter is phenomena.
12.2
van Fraassen and Maxwell’s discussion of observability.
As argued
for above, the phenomena are the real entities in our surrounding world, while
the rest are auxiliary concepts and constructions based on these sensations and
on reasoning. In reality, the laboratory technician does not sensuously relate
himself to microbes, as also referred to, but to something he sees in a
microscope, while the result of the computation of what we call “the size of
the microbes” are only a construction based on these sense impressions. Cf. Treatise No 3, “Maxwell’s continuous
arrangement of kinds of observations”.
Maxwell
argues for the reality of such entities in that what can be seen through a
window glass can also be seen without and what can be seen through a telescope can
be seen too if the observer comes close enough. However, it is insufficient
what van Fraassen puts forward against, viz. that the following principle
roughly applies:
”X
is observable if there are circumstances which are such that, if X is present
to us under those circumstances, then we observe it.” ([Fraassen 80],
s.16, my underlining)
It
is true that the anti-realist will conclude that as microbes cannot satisfy
this demand, microbes cannot be observable according to this definition. Nevertheless,
van Fraassen’s definition is defective, for we must also ask the questions:
When is a body present to us? When does it exist? Besides, what is the difference
between being present and being observed? However, the most serious objection
to this definition is that the very talk about the possibility that “X is
present to us” presupposes that there exists something we can call “X”, and
which is principally observable.
Moreover,
realists like Maxwell may unconcernedly object, “A microbe is present to us if
we see a microbe in a microscope”. Certainly, to this can be stated that it is
a problem what it means, “to see a microbe”. However, as mentioned in “The semantic
approach to the realism debate”, to some realists there is no distinction
between direct observation and something they would call “indirect observation”,
i.e. direct observation (of the microbe picture) plus conclusion of existence.
It
can here be referred to what has been put forward in Treatise No 3, “Maxwell’s continuous arrangement of observation
types”: We cannot say that microbes exist in the same way as the things we see
with our naked eyes or with glasses. Because when we look at microbe figures we also relate to them as if
they had the size we see. Or rather, we relate to the very size we see, whereas
the size we attach to our imagination of a microbe is the theoretic result of a
calculation based on the size of these figures and the magnification factor of
the microscope.
12.3
Usage and observability.
The
above-referred argumentation that we are not ontologically committed by our way
of speaking (Cf. Treatise No 3, “Usage
and existence”) consisted in a disproof of the opposite statement that we are
ontologically committed by our talk, i.e. committed to assuming a literal
existence of the imagined entities about which we are (seemingly) talking. As a
counter-example, the argument used statements about a special type of objects
that are observable in principle, viz. pictures on a PC screen. These may exist
in certain situations (visible on the screen), but not literally behind other screen pictures, though we
may express ourselves as if this is the case. This means that when we have to
explain what takes place on the PC-screen, we can do that in terms of pictures
behind other pictures without having to assume that such pictures exist on a PC
screen (what is actually not the case).
Some
may object that the argument does not necessarily apply to all sorts of talk,
such that certain kinds of talk can
be ontologically committing. Even though such a talk could be documented, this would
not be an objection against the argument, since it consisted in a disproof of
the idea that we are ontologically committed by of our usage (viz. by
presenting an example of the existence of ontologically not-committing talk).
The
important question is what we (at most) can put into our talk. As regards our
talk about screen pictures, this is easily understandable: The statement “The
screen picture is covered by another picture” just means that a part of the
former picture is replaced by a copy of a part of the latter, and that we can
recreate the original picture by the opposite operation. Thus, the statement “The
picture is uncovered” means that the picture parts belonging to the place where
the removed object was, are created again. Moreover, the statement “The picture
has been moved” means that the picture is removed and recreated at another
place (as just described), while the parts of the picture that belong to the
place where the removed object was, are created again.
However,
it is more difficult to answer the question how to interpret our talk about
not-observable entities, such as electrons (i.e. a talk containing terms for putative,
not-observable entities). For we must mean something specific by our statements
about electrons, I suppose. Moreover, there must in fact exist something specific that we talk about, when we talk
about electrons - also in an anti-realist sense - though this (such as it has
been argued for) cannot be that there are literally electrons.
When
we talk about electrons, i.e. make statements containing the term “electron”,
we express ourselves as if we refer to well-known macroscopic entities. However, our statements about these
elementary particles cannot deal with physical entities in the space, as the
semantic meaning of this lacks a clarification. For the same reason, the solution
of this problem cannot be either that we talk as if there were
physical entities with qualities like electrons in the space. (Cf. also the critique
of van Fraassen's idea that science has to be literally understood.) As already
touched on, a clarification is also necessary because understanding statements
about electrons by analogy with well-known physical entities has a limiting effect
on our fantasy and thus on our scientific creativity.
Though
our talk about “letting a screen picture cover a part of another picture” can
be translated, we do not think along those lines when we express ourselves in
such terms. In this case, our talk even corresponds to what some persons might
naively think about what happens at the surface level of the computer screen. When
our talk about a screen picture that are hidden and shown can be translated to
a talk about removing and recreating pictures on a PC-screen, this is because the
implementation aims at this. On the other hand, it is not guaranteed that our
talk about putative, physical entities can be translated into semantically
meaningful talk. Thus, it is not certain that the scientists' talk about
electrons can be translated, as this talk can be a result of a number of cases
of conceptualizations that have immediately immediate scientific advantages. For
that reason, the claim for translatability may be unreasonable.
Compared to this, translation of well-defined
programming languages, e.g. from high-level programming languages into
low-level programming languages, are unproblematic as they are defined by
already defined rules. On the other hand, rules for translation from mother
tongue A to mother tongue B are based on what statements, τ, a person that
speaks language B would express if he were in the same situation as another
person who speaks language A and who asserts the statement σ. However, as the
languages are different and a part of the situation, it is of course a question
whether it has meaning to talk about the same situation at all. For this reason
alone, translation can be made difficult or in principle rendered impossible.
As
to the screen picture-example, we may be said to talk as if physical pictures,
such as diagrams on paper, were present. However, we can also interpret our
talk as statements that would be true if they concerned events in a
mathematical model. (Such a model can for instance contain one representation
of each picture, a number of planes of which the topmost plane is called “plane
1”, the plane second from the top “plane 2”, etc. Here, the separate picture-representation
can be placed on a specific plane, i.e. attached its plane-number and position
on the plane, and in which the screen has a position in “plane 1”, etc.) As to
the terms of elementary particles that enter into theories about so-called
traces in bubble chambers, they can also be understood as entities in a
mathematical model, analogously to the description of a trajectory of a ball.
Nevertheless,
it is reasonable to ask, as done above, what we really speak about when we
speak of covering a screen picture with another one, for - though there is no
screen picture behind the latter screen picture - we must be dealing with
something that a grammatical object in an interpretation can refer to. In a
similar way, it is reasonable to seek to clarify what is present when
scientists talk about an elementary particle - though the concept does not have
meaning understood as a term of a physical (not-observable) particle. The result
needs only to be the description of a situation of some sort in which some definite
phenomena form part.
The
scientific usage, in which different terms of elementary particles with certain
qualities and behaviour patterns enter, has a causal history that is influenced
by scientific realism and conclusions by analogy with the behaviour of things
surrounding us. Therefore, this usage is not necessarily consistent. And thus,
it may seem remarkable that it can be used - that the scientific theories
simply do not contain more inconsistencies than they actually do. Despite their
diachronic coincidences, of which there may be good explanations, the ideal
must be a consistent, synchronic description of the connection between the
theories (i.e. their structure).
Concerning
the understanding of our talk about so-called not-observable entities in
science, in the previous text it has only been arrived at a demonstration of
this: 1) Bas van Fraassen's agnosticism cannot be maintained at all since it
allows semantically meaningless statements. 2) It is tempting to think that
corresponding to each of the statements the above talk consists of, there must
exist certain physical incidents and qualities that can be described in a
literally understandable way. 3) Nevertheless, it must be concluded that this
talk about not-observable entities does not necessarily constitute an entirety,
as we at most can expect ad hoc
connections between our talk and the discussed situations.
Certainly,
Bas van Fraassen asserts that the talk about underlying causal mechanisms can
be interpreted as a talk about the intern structure of the models ([Fraassen
80], p. 80), and this concept of models may in fact be a way to a solution of the
problem of what scientists' talk about elementary particles and their qualities
deals with. However, the problem concerning a connection between these
structures in models and the phenomena remains. For the question is still
whether there exists a one-to-one correspondence between the models and the
physical world, understood as the phenomena and their regularity.
Example: Based on Newton
In
contrast to the movements of observable objects, however, the only thing we can
put into the concept of the so-called traces of the elementary particles is the
concept of a figure. Here is again
an instance of how we allow ourselves to be misled by our usage, as our
understanding is influenced by our everyday talk about drawn traces, which presupposes that a movement has been made. However
if we have never observed such a movement of a particle we are not justified in
assuming that it exists as a part of reality – in contrast to the known curves
of falling bodies. On the other hand, we can introduce a concept of movement in
a mathematical model if it is useful.
The
scientific model line of thinking can be illustrated with the following example,
which also leads to critique of realism: According to realism, the Brownian
movements can be explained by the literal existence of pushing molecules. Therefore,
molecules must exist (according to realism). For molecules can only push if
they exist. On the other hand, according to the anti-realistic opponent, we can
only talk about an abstract model with attached mathematical rules and
operations, the results of which can be applied to the phenomena, including the
Brownian movements. However, the interesting thing is that the realist
explanations at the beginning of this paragraph themselves are about a model,
for they just turn on something we can only picture to ourselves. Thus, we are
only dealing with our imagining something that serves as a vague model. The
realist must give reasons for stating that this model represents reality also
literally. They must explain what it means that molecules exist.
13. A critique of realism as a good explanation.
In
the realism debate, it is stated that it is an argument for realism that since realism
can explain the progress of science realism must be a good explanation.[16] In
this conclusion, it is obviously presupposed that good explanations are true,
since the mere idea that realism is a
good explanation cannot be philosophically interesting. Only the idea of
its truth can be that. For the sake
of convenience, we can call the argument the “success-argument for realism”.
According to this argument, realism has superior explanation ability than other philosophical directions, including anti-realism.
This means (according to the line of thought) that realism is not just the
direction to be preferred in the philosophy of science but also that it is true.
However, what is it that has to be explained at all?
As
far as I have understood, the extent of scientific progress rather depends on
the scientists' effort and abilities to make out usable theories than on some reality
behind the phenomena. The progress of science does not call for another explanation[17].
Here
we are obviously discussing the realism debate
itself. The discussed realism debate is about what semantic meaning we can
attribute to our existence statements. The above success-argument for realism presupposes that no problem is left
over as to this question about meaning. However, if this matter were already clarified,
the very realism debate would obviously be settled, as its task exactly consists
in answering this question. Alternatively, if the realists reject this view of
the realism debate, they have first to put forward a definitive argument
against the here advanced view of the importance of the concept of meaning, and
thus against anti-realism, in order that the success-argument can be applied at
all. However, this argument against anti-realism would also render the success-argument
superfluous, of course. So, whether this premise about the content of the
realism debate can be obtained or not, it is superfluous to involve the success-argument
in the realism debate as such.
As it
appears, the realism debate is semantic and not a subject of natural science. Nevertheless,
it is interesting to discuss the success argumentation on its own premises,
i.e. disregarding the above counter argument. Under this condition, we can
construct an imaginary argumentation for the contention that realism is able to
explain the progress of science.
This
constructed argument is partly inspired by Hackings “argument of the grid”, which
thus will be discussed first.
In
the chapter, “The argument of the grid” in the article “Do we see through a
microscope” in Images of science,
Hacking puts forward the following argument against realism. Hackings uses as a
premise that it is possible to reduce a drawn grid (by processing a disc by
means of a suitable apparatus and the drawn grid) and later seeing this grid
“again” in 12 different kinds of microscopes. From this Hacking concludes that
the disc must really contain the grid structure then, and that this cannot be
due to “a gigantic conspiracy” between the involved 13 varied processes. This
is a case against anti-realism, according to Hacking. ([Churchland 85], p.
146/7)
Bas
van Fraassen refutes the argument of the grid by arguing that the grid is
constructed such that the desired result is obtained. ([Churchland 85], p. 298)
However, by this he does not give a satisfactory answer to the question Hacking
seeks to answer, viz. how this “recognition” is possible at all, i.e. how it is
possible to construct the grid in the desired way. Thus, van Fraassen’s argument
can be countered by the question how it is possible at all to construct - not
the so-called reduced grid – but the instruments, such that they meet the definite
demands Bas van Fraassen’s argumentation presupposes. The answer to this is
that they are simply based on the same theories. The fact that these can used
to create different instruments that gives the same results is not surprising.
However,
the case is not at all about the problem of existence that is created by
Hacking and discussed on its own terms by van Fraassen – and so far by the
author of this treatise. Instead the decisive thing is that the circumstances
Hacking points out can of course not make the theories literally true, no matter
how elaborated they are, as the case is about establishing of semantic meaning.
This must be present before we can talk about existence versus not-existence.
It
is possible to picture a more simple argument that just as Hackings is intended
to show the reality of the so-called not-observable entities: “When now
Millikan’s determining of the least size of the charge of the electron gives
the same result time after time[18],
electrons with fixed qualities must simply exist.”
Of
course, this example is not an attempt at refuting Hacking’s argument for the reality
of these not-observable entities just by inventing another example that is just
as bad. This thinned example is just intended to make clear the weakness of the
argument. It is a trite objection that its conclusion just moves the
explanation problem from one object to another, as it generates the question
about what is the reason to electrons
possessing a charge and similar questions ad infinitum.[19] Less
trite we can ask the question how electrons and other entities that possess qualities
can exist. What is the semantic meaning of assertions that something possesses
a quality or that qualities exist at all? To the anti-realist it is only a
question of using the same helping concepts for describing some of the regularities
of nature in what is supposed to be the same situations.
A
thought argumentation for realism: According to realism, there exists a reality behind the phenomena. This
reality science tracks down. Thus, as science is dealing with something real,
it follows that its good explanations are literally true. Therefore, science is
able to frame descriptions of this reality and make use of them as a firm and
certain starting point for further research. It is for this reason that science
makes certain advances. These are thus explained by realism. In this way,
realism has given a good explanation, and therefore realism is true.
By
this, we are back at the allegation that introduced this chapter. Besides the
facts that the progress of science does not call for a different explanation
than the one the scientific work involves, the argumentation is fundamentally
erroneous, for the criticized argumentation has a structure or form that belongs
to science, and not to philosophy. The reason is neither that it is about the
theories of science nor that it is about the physical world, but that it takes
as a starting point the question whether some assumptions are good or not. For
that reason, it cannot prove anything fundamental; on the contrary, it reflects
a lacking understanding about what scientific and philosophical argumentation demands:
13.1
Criticism of the concept of good assumption.
In
science, a good assumption is just only
a good assumption, and not something different or more, and only when it is an
assumption that can prove useful – i.e. capable of predicting the phenomena.
Philosophy, however, has to explain what interpretation we can at most put into
our statements including statements about good assumptions. For this reason,
the following can be concluded:
As
a philosophical direction, realism must be able to explain what good assumptions
are. However, if realism itself is just a good assumption, like the assumptions
of science, it itself demands an explanation of the same sort as the
explanation it claims to provide.
With
other words, the adduced argument for realism is itself an argument of the
discussed kind. (The argumentation resembles the attempt of explaining
induction in science by the argument that the principle of induction has worked
until to now, and for this reason must work in the future - so that the validity
of the principle of induction is justified by means of induction.)
Example: In Representing
and Intervening [Hacking 83], in one and the same argumentation Ian Hacking
succeeds in both 1) arguing scientifically for a philosophical point of view
and 2) using a circular argument for scientific realism:
“Experimental
work provides the strongest evidence for scientific realism. This is not
because we test hypotheses about entities. It is because entities that in
principle cannot be ‘observed’ are regularly manipulated to
produce a new phenomena and to investigate other aspects of nature.” ([Hacking
83], p. 262. my underlining)
Re
1: The debate concerning scientific realism is about the meaning of scientific
statements. This also applies to statements that are inspired by experiments. Scientific
experiments are as such not about philosophy. The relation between philosophy
and science is that philosophy can be useful to clarification of the theories
of science while science is a supplier of problems to philosophy.
Re 2: Talking about entities that cannot be observed
presupposes that they exist or can exist and that this talk has meaning. Therefore,
that cannot be part of an argument for their existence. (It must here be
assumed that Hacking's existence statement has to be taken literally, as it
cannot otherwise be an argument for scientific realism.) The same objection applies
to Hacking's concept of “manipulating entities”, though Hacking assigns a
special status to this concept according to the quote.
The
above argumentation against “realism as a good explanation” is sufficient, so
far. Nevertheless, the following specifying remarks can be added:
As
mentioned, it is the task of science
to predict the phenomena by means of good assumptions, viz. theories, whereas
it is one of the tasks of philosophy
of science to clarify what we can at most put into the good assumptions of
science. While scientific theories do not have to be literally true to be used,
philosophical conclusions are aimed at being fundamentally true.
Justifying
realism by asserting that it is a good assumption and thus have to be true is
to commit two fallacies:
1) The use of a scientific argument
in philosophy.
2) The use of the way of arguing that is discussed.
2) The use of the way of arguing that is discussed.
Re 1:
1.1)
The purpose of scientific work is to find usable theories about the regularities
of the phenomena, while the purpose of philosophy is more fundamental, such as
making out what we can at most put into our statements, including scientific
statements. Philosophical questions demand justified, certain answers, as they
have to be fundamental. Asserting that realism is a good explanation or
assumption is not fundamental and thus simply too unambitious.
1.2)
Scientific statements containing terms for so-called not-observable objects do
not have to be literally true in order to be usable. Thus, even if the success-argument
for realism had been valid, this would not lead to the conclusion that realism
is true, only to the one that realism is a good assumption - which does not
have to be true, only usable so far – and this is hardly satisfiable to the
realists.
Re 2: Philosophy ought to be able to explain what it
means that an explanation is a “good explanation” including what it may imply. However,
this leads to the following reasoning:
If
the predicate “good explanation” can be applied to the philosophical direction
realism and thus substantiate it, this realism must be able to justify
something by which it is justified itself - i.e. be able to justify its own justification.
This is circular, of course, since realism itself has to be justified in order for
something to be justified by means of realism as a premise.
The
discussed argument for realism thus presupposes that it is clarified what we
can put into the concept of good explanations.
However, this is just what realism itself ought to be able to explain.
Put
in another way: The
good assumptions of science are good as to predicting the phenomena. On the
other hand it is difficult to see what possible “good assumptions” should be
used to in philosophy, since what is philosophically interesting is to discover
what statements we can justify, and
what semantic meaning they have, and not just what we may assume. For that reason, the result of accepting the discussed argumentation
for realism would at most be that realism is just a good assumption. The argument
for realism presupposes that the good assumptions of science are literally true,
in order that it could be interesting, but it is this thesis that has to be
proven, as it just expresses the just discussed realism.
Very
naturally, however, this presupposition about realism does not enter as an explicit
part of the presented argument. If we want to follow and examine the
argumentation on its own limited terms, i.e. as it is formulated and thought,
we can allow ourselves to presume that it does not have as a premise that the
good assumptions of science are literally true.
The
idea is that the alleged conclusion that realism is a good explanation has to
establish the truth of realism. To investigate this, let us compare this with
the fact that scientific theories are based on good assumptions. Just as this
precondition does not release the
good assumptions of science from the state of just belonging to the set of good
assumptions, in the same way, the allegation that realism is a good explanation does not release realism
from the state of just belonging to the set of good explanations, which just means good assumptions. This refutation is
even based on the premise of the criticized argument, viz. that realism can be
discussed in this way at all, which has already been rejected as been due to a
misunderstanding. For as mentioned, two basic errors are committed in the
discussed argument for scientific realism, viz. treating realism as a theory
(on equal footing with the scientific ones), and letting the realism debate be
about existence and not about the meaning of our statements, such as existence
statements.
As
regards everyday assumptions, it is in our interest that the statements that
describe them are true: it forms part of our assumption that what we assume is
actually true (as we do not assume something that we regard as false).
Moreover, we can often examine whether our assumptions about everyday phenomena
and their regularity are in fact true, as by definition phenomena are something
to which we are presented. As to scientific assumptions, these just involve the
assumption that we can describe the phenomena in a certain way - by means of
theories involving different parameters such as the acceleration due to gravity
or electrical charge. This means that we can also predict the phenomena by means of these theories, such that the
theories can also be verified, but only as regards their ability of predicting
the phenomena (and not necessarily forever). However, some of the applied
parameters are only to be considered ingredients in the mathematical
description of the theories. (Cf. also [Fraassen 80], p. 11-12.)
13.2
Hacking’s positron-spray argument.
As
an example of an argument for scientific realism of the above discussed kind,
we can look at Hacking's assertion (quoted below) that if we can spray with
elementary particles they are real. This can be regarded as an illustration of
what Hacking talks about when as above referred he says that manipulation with
entities can prove that they are real. In Representing
and Intervening [Hacking 83], Hacking first quotes a scientist’s description
of an experiment and then describes his own reaction to it:
”'we
spray [the niobium ball] with positrons to increase the charge or with
electrons to decrease the charge'. From that day forth, I've been a scientific
realist. So far as I'm concerned, if you
can spray them then they are real.” ([Hacking 83], s.23)
Here,
the author of this article will assume that declaring that something is real,
is the same as declaring that it exists.
However, the very assumption that Hacking's way of concluding has sense presupposes
realism. The reason for this is not just that the existence of elementary
particles - and thus the possession of meaning of the corresponding statement
about existence - is already presupposed in the premise “if you can spray them”, which
is sufficient for declaring that the argumentation is circular. It is also quite
unproblematically presupposed that the very conclusion “they are real” (said about elementary particles) has meaning, which
also reflects realism. For it is an unclarified question what semantic meaning
this conclusion has.
Hacking’s
reasoning cannot be compared to what we do when we infer that there is a ball
in a box, for then we just infer that the box contains an object that in a
fundamental sense is of the same kind as the other things we know. However, we
do not draw the extra conclusion that there exists something fundamentally new
- which Hacking does in his example.
Moreover,
Hacking's argument does not only presuppose - the way it is expressed - that we
are ontologically committed by our talk (which has been refuted in Treatise No 3) but obviously also that
our formulations simply implies
qualities of the physical reality, and thus the truth of statements about existence.
Hacking may not really intend that, but it follows from his text.
The
following example has the same argumentation structure. If we are looking for
an object by a stick of wood at a not easily accessible place, we can say for
instance, “Now I can feel a stone with the end of the stick” (because it feels
in that way). However, this does not imply that we have to assume that we do literally
sense something with the end of the stick, in the same way as we can do with
our big toe. This is yet an example of the fact that we are not ontologically
committed by our usage. However, corresponding to Hacking's example we should
be able to conclude, “If you can sense an object by means of something, then
that something do exist” - viz. sensation
at the end of the stick - because we talk as if such a sensation exists.
Literature.
[Aristotle
68] Aristotle:
The Metaphysics,
English translation by Hugh Tredennick,
Harvard Univ. Press, William Heinemann Ltd. (Great Britain
English translation by Hugh Tredennick,
Harvard Univ. Press, William Heinemann Ltd. (
[Bergson 82] Henri Bergson: Matière et
mémoire
Presses Universitaire de France, 1939 (Paris 1982)
Presses Universitaire de France, 1939 (Paris 1982)
[Berkeley
Penguin Books (London
Penguin Books (
[Churchland 85] P.M.Churchland
& C.A.Hooker (eds.): Images of
Science, Essays on Realisms and Empiricism with
a Reply from Bas C. van Fraassen,
The University of Chicago Press
(Chicago/London 1985)
Science, Essays on Realisms and Empiricism with
a Reply from Bas C. van Fraassen,
The University of Chicago Press
(Chicago/London 1985)
[Dummett 78] Michael Dummett: Truth and other Enigmas
Duckworth (London
Duckworth (
[Dummett 91] Michael Dummett:
The Logical Basis of Metaphysics
Duckworth (London
The Logical Basis of Metaphysics
Duckworth (
[Dummett 93] Michael Dummett: The Seas of Language
Oxford UP (Oxford
[Favrholdt 99] David
Favrholdt: Filosofisk Codex
Gyldendal (Denmark 1999)
Gyldendal (Denmark 1999)
[Fraassen 80] Bas C. van Fraassen: The Scientific
Image
Oxford University
Press (Oxford
[Hacking 83] Ian Hacking: Representing and Intervening
Cambridge University
Press (London
[Hume 90] David Hume: Enquiries
Clarendon Press (Oxford
Clarendon Press (
[Kant 76] Immanuel
Kant: Prolegomena
Philosophisce Bibliothek Band 40,
Verlag von Felix Meiner (Hamburg, 1976)
Philosophisce Bibliothek Band 40,
Verlag von Felix Meiner (Hamburg, 1976)
[Kant 90] Immanuel Kant: Kritik der
reinen Vernunft
Philosophische Bibliothek Band 37a
Felix Meiner Verlag (Hamburg 1990).
Philosophische Bibliothek Band 37a
Felix Meiner Verlag (Hamburg 1990).
[Klausen 04] Søren Harnow Klausen:
Reality
Lost and Found
University Press ofSouthern Denmark
(Denmark 2004).
University Press of
(Denmark 2004).
[Kuhn 70] Thomas S. Kuhn: The Structure of Scientific Revolutions,
Second Edition, Enlarged
TheUniversity of
Chicago
(Chicago 1970).
The
(Chicago 1970).
[Maxwell 62] Grover Maxwell: 'The
Ontological Status of Theoretical Entities' in
Minnesota Studies in the Philosophy of Science,
vol. III:
Scientific Explanation, Space, andTime
University of Minnesota
Press (USA
vol. III:
Scientific Explanation, Space, and
[Millikan 24] Robert Andrews Millikan:
“The electron and the light–quant from the experimental point of view”, Nobel
lecture May 23, 1924, from
Nobel Lectures, Physics 1922-1941.
http://www.nobel.se/physics/laureates/1923/millikan-lecture.html
Last modified October 23, 2001.
Nobel Lectures, Physics 1922-1941.
http://www.nobel.se/physics/laureates/1923/millikan-lecture.html
Last modified October 23, 2001.
[Nietzsche
88] Friedrich
Nietzsche: Jenseits von Gut und Böse
Reclam 7114 (Stuttgart 1988)
Reclam 7114 (Stuttgart 1988)
[Putnam 75 A] Hilary Putnam: “The
Meaning of ’meaning’” in Mind Language
and Reality
Cambridge University Press (USA
Cambridge University Press (
[Putnam 75 B] Hilary Putnam: Philosophical
Papers vol I, Mathematics mather and method
Cambridge University Press (Cambridge
Cambridge University Press (
[1] The reason for these hackneyed considerations
is found in Søren Harnow Klausen's dissertation Lost and Found ([Klausen 04]). This applies to the considerations
concerning these two facts:
1)
The above-described unsuccessful definitions of anti-realism, and thus of the
realism debate as such. Cf. [Klausen 04] p. 13: “Since antirealism is the denial
of realism, the definition of this term becomes absolutely crucial”.
2)
The necessity of involving semantic considerations in the discussion. These are
necessary in order to avoid statements of the following kinds: “Realism is of
course the belief in the existence of reality.” ([Klausen 04] p. 14, my
underlining) This statement is based on a misunderstanding:
For
according to anti–introductionism, there are not only good reasons to believe that reality exists, but also
reasons to know that reality in fact
exists. It is simply senseless to deny the existence of reality! For we sense and handle it and talk
with other persons about this in a way that works. Instead, the question is
what we can put into our statements about entities of different kinds. This subject
has already been discussed as regards physical objects, time, consciousness,
and scientific constructions.
[2] Berkeley also talks about
"whether he [the reader] means anything by the actual existence of an
idea, distinct from its being perceived" and refers to this concept with
the words, "those propositions which at bottom have no meaning in
them". ([Berkeley 88], §45, my square brackets.) Cf. also §§77, 79-80.
In
[Hume 90], §58, Hume talks about the impossibility of finding a causing force
and infers "that the necessary conclusion seems to be that we have no idea of connexion or power at all, and
that these words are absolutely without any meaning, when employed either in philosophical
reasonings or common life".
[3] Though it is not important in this connection, and perhaps not the
cause of this comment, it must be mentioned that this proviso is necessitated
by Thomas Kuhn’s talk about normal science and paradigm shift. [Kuhn 70]
[7] Cf. the criticism of the principle of inference to the best explanation as regards
introduction of new types of entities, or the so-called ”abduction” in ”A. Realism”, in “Three different views...” in Treatise No 2.
[8] Here
shall be referred to Aristotle’s arguments for his principle of contradiction.
([Aristotle] 1005b)
[9] It must here be referred to
David Hume's treatment of what it means to imagine something, in ([Hume 90],
§13 ff).
[10] This follows from the transitivity
of implication, which implies that (A ó B & A ó C) ≡ (A ó B & A ó C & B ó C).
[11] This may cause associations to
Hilary Putnam's talk about "division of linguistic labour" in "A
sociolinguistic hypothesis" in "The Meaning of ’meaning’" in
Mind Language and Reality ([Putnam 75 A], pp. 227). In the explanation
structure, however, we are dealing with division of knowledge about all kinds
of theoretical explanations.
[12] I owe to mention that this may be inspired
from what has been accounted for above from ([Dummett 78], p. 358/9, 361),
where Dummett assets that the meaning of a statement is given by the conditions
which we regard as proving the truth-value of the statement.
[13] Cf. also the
formulation of this in the subsection “Maxwell’s continuous arrangement of
observation types” in Treatise No 3.
[14] I owe to mention that Berkeley discusses this case in ([Berkeley
[15]
van Fraassen also touches this subject. Cf. ([Fraassen 80], p. 80) and the
subsection “Usage and observability”.
[16] In Matière et mémoire Bergson
praises the phenomenalist Berkeley for having demonstrated that the secondary
qualities are at least as real as the primary ones, but then criticises him for
having happened to "transporter la matière
à l'intérieur de l'esprit et en faire une pure idée" ([Bergson 82]
p. 3), after which Bergson concludes that.
"(...) Berkeley se vit incapable de rendre compte du success de la
physique (...)." ([Bergson 82] p. 3)
According
to Bergson, obviously Berkeley
Hilary
Putnam puts forward a similar but stronger statement in Philosophical Papers vol. I, Mathematics matter and method:
"The
positive argument for realism is that it is the only philosophy of science
which doesn't make the success of science a miracle." ([Putnam 75 B] p.
73)
Putnam
supports his argumentation by putting forward that the same existence statement
can be used by different theories.
[17] Here I owe to say that Bas van
Fraassen very ironically criticizes this demand of explanation put forward by
Putnam ([Fraassen 80], p. 38-39).
[18] Cf. the account
of the experiment in Treatise No 3 or
in ([Millikan 24], p.54–61).
[19] Similarly, a
nine-year-old child that is told that some god has created the world will immediately
ask who or what has created this god, then.
