New ‘new riddle of induction’

Logical positivists were interested in giving rules of confirmation which would establish scientific
laws. And these rules would be objective and formal like deductive logic. Also, how we get the
scientific laws is continuous observation and verification, which is empirical and inductive. On
the other hand, logical positivists try to let inductive conclusion form to be deductive premise and
to well confirm prediction. Like the following figure, logical positivists through inductive
argumentation conclude that all emeralds are green. And use this conclusion to objectively confirm
if there exists an emerald, it must be green.

For Goodman, who holds that inductive cannot be a fully formal, objective relation like deductive.
He gives the ‘grue’ example which is also called ‘The new riddle of induction’ to show that there
is no such theory to show inductive can be a fully formal, objective relation like deductive.

For ‘grue’ example, ‘grue’ is a predicate that when objects x is grue if and only if x is green and
is first observed before, or x is blue after observing it. As when we observe it, the object is green,
so we can get inductive argument comes from this observation. We can observe emerald 1 is grue,
emerald 2 is grue...., and conclude that all emeralds are grue.
In this inductive argument, the premise ‘emerald X is grue’ seems to be true, but it is difficult for
us to agree with the conclusion ‘all emeralds are grue’ as grue is the predicate that would be blue,
is only contingently true at that time. Therefore, all emeralds are grue may not be true. From this
example, Goodman tries to point out that there’s no language-independent way to distinguish
predicates if its ‘well-behaved’ predicates like ‘green’ or non-well-behaved predicates like ‘grue’.
And the problem of predicates will be leading us cannot be well confirmed to draw conclusions
inductively.
I agree with Goodman skeptic about the inductive argument that problem of predicates will be
leading us cannot be well confirmed to draw conclusions inductively. But for Goodman’ example
of ‘grue’, I don’t agree with that is good example to challenge inductive argument. As a valuable
inductively argument, the premise must be true, but based on the definition of ‘grue’, the premise
‘emerald X is grue’ may be wrong, this leads to the argument to be unvalue. Therefore, in this
paper, I would agree with Goodman conclusion that skeptic about the inductive argument. Under
this premise, I will point out the problem of ‘grue’ example that the premise ‘emerald X is grue’
is false. Also, I will point out some possible criticisms and respond to them. Then, I will give a
new example to replace ‘grue’ this problematic example.

Problem of Observation method & Time

The first problem of ‘grue’ example is that before we draw conclusions ‘all emeralds are grue’,
some of or all of the premise ‘emerald X is grue’ may turn out to be false. When we say 'Emerald
X is grue' which means we need to observe to verify it. Also, to conclude ‘all emeralds are grue’
which we need to verify all (at least most) of emerald is grue one by one. Then here may have a
problem that like figure 1.

Figure 1

As when we observe the Emerald 1 is grue, the premise is true at that time (time t1). Also, when
we observe the Emerald 2 is grue, this premise is also true at time t2. But at the same time, the
premise ‘Emerald 1 is grue’ will turn out to be false. Because of grue is the predicate that is blue
after observing it. When we start observing Emerald 2, which also means the fact that observe
Emerald 1 and get the premise ‘Emerald 1 is grue’ is in the past. According to the definition of
grue, Emerald 1 has been after observing, which means that Emerald 1 has been turned to be blue.
Considering that emeralds cannot be blue, the premise ‘Emerald 1 is grue’ is false when we are
observing Emerald 2. On the other hand, when we are observing Emerald X+1 at time t+1, the
Emerald X will always be false. Therefore, for the ‘grue’ example, base on the premise the
‘Emerald X is grue’ will always be false, it makes the ‘grue’ example of inductive argument
unvalue. And it is nonsense to keep using a unvalue argument to support Goodman’s inductive
skeptic view.
But someone may question why we need to go verify it one by one, why can't we verify all
emeralds at the same time and draw conclusions? Therefore, someone may respond that we can
imagine having a verification machine, it can verify ‘all emeralds are grue’ in the world at the
same time. In this way, the case that when we are observing Emerald X+1 at time t+1, the Emerald
X will always be false will not occur, as all emeralds will observing at same time.
Assume that this verification machine exists, also can verify ‘all emeralds are grue’ at the same
time. But it doesn't help, take look at figure 2

Figure 2

Pay attention to the predicate of ‘now’, now is the predicate that is keep moving. When we say
event E is now. This usually means ‘E occurs at time t’. According to the moving ‘now’, which
means that ‘E is now’ is the past that occurs at t. On the other hand, when use the verification
machine to verify ‘all emeralds are grue now’ is usually means ‘we verify all emeralds are grue
occurs at t’. Which means that ‘we verify all emeralds are grue’ is past that occurs at t. According
to the predicate of ‘now’, The premises ‘Emerald X is grue’ that we verify through observing will
all turn out to be false before we get the conclusion. Therefore, when we are observing Emerald
X, Emerald X will always be false based on the predicate of ‘now’, the ‘grue’ example of inductive
argument is unvalue.
Somone may say if we redefine the predicate ‘grue’ that x is grue if and only if x is green and is
first observed before 2050, or x is blue after 2050. I agree that redefine the predicate ‘grue’ can be
avoided the problem of observation method. But this response does not defend logical positivists.
Because redefine the predicate ‘grue’ which means that the inductive argument of ‘grue’ example
becomes valid again, and it will making sense of Goodman’s inductive skeptic view again.
But please note, this paper is focused on Goodman’s ‘grue’ example of inductive argument is it
value or not, not focus on the point of Goodman’s inductive skeptic view. Therefore, I agree
redefine the predicate ‘grue’ can let Goodman’s ‘grue’ example of inductive argument to be value
again, but this is only limit on if there only have the problem of observation method and time. As
the ‘grue’ example may still have the following problem.

Problem about feature of grue

Consider the proposition ‘snow is white’, which means that snow has property of white. In grue
example, ‘Emerald X is grue’ means that emerald has property of grue. But by definition of grue,
emerald is not grue, see figure 3

Figure 3

Predicate green is the property that is always green. But grue is the predicate that will turn to blue.
Before observing or 2025, green and grue have the same properties that is green, but they are
fundamentally different. As grue have a modal property that will turn to be blue but green doesn’t.
In the ‘grue’ example, the premise ‘Emerald X is green’ is true as emeralds are always green. But
when we say ‘Emerald X is grue’ which means that emeralds have the same properties as grue. On
the other hand, to say ‘Emerald X is grue’ means that emeralds have the modal property that will
turn to be blue. Obviously, emerald is not grue, it doesn’t have the modal property that will turn to
be blue. Therefore, the premise ‘Emerald X is grue’ is false, and it will make the ‘grue’ example
of inductive argument unvalue.
Considering there may be a problem of predicate grue, it makes the ‘grue’ example of inductive
argument unvalue. For Goodman, he may say that this is exactly what he wanted to do, there are
no language-independent way to distinguish predicates if its ‘well-behaved’ predicates ‘green’ or
‘grue’. But again, we will be more likely to accept value arguments to support our point of view
rather than using a unvalue argument like value argument. Therefore, I hope to create a new value
inductive arguments example to replace the ‘grue’ example. To summarize, ‘grue’ example may
have the following problem and I hope to avoid these problems as much as possible in my example:
1) ‘grue’ is a temporary predicate that create by Goodman, creating a new predicate can be
controversial, therefore the example will use existing predicate
2) ‘grue’ example is unvalue inductive arguments, it is not making sense for a unvalue
argument can support Goodman’ view, therefore the inductive arguments must be value.
3) According to Goodman wanted to achieve in the ‘grue’ example, all premises must be true,
and the conclusion seems to be true but is contingently true

Example of ‘water’
Water is H2O is a scientific fact. As a scientific fact, we need to use an inductive argument to
verify water is H2O. Which means we need to find kind of liquid it calls ‘water’ and verify it like

We use scientific instruments to verify that all rivers and lake, also it is true that rivers and
lakes are H2O. Then we can confirm the premise ‘water is H2O’ is true and get the
conclusion ‘all water is H2O’. In this example, all premise is true, but is it all water is H2O?
Consider the predicate of ‘water’, when we use the predicate of ‘water’ in daily it usually refers to
some clear liquid or the liquid that can be drink, which include river, lake and sea. But for the sea,
the main component of sea water is not H2O, is NaCl (Sodium chloride). Therefore, it seems
that not all water is H2O, although all premise is true.

Someone may respond that as water is H2O is a scientific fact. We need to distinguish the daily
languages of water and the scientific sense of water, and what we conclude is all scientific sense
of water is H2O. But it is nonsense. First, how can we distinguish the daily languages of water and
the scientific sense of water. It will go back to what Goodman tried to say, there’s no language-
independent way to well distinguish predicates. Second, if we can distinguish the daily languages
of water and the scientific sense of water, then what the heck is scientific sense of water? I think
someone may say scientific sense of water is the main component of liquids is water molecules.
But this is circularity, because the ‘scientific water’ is liquids with water molecules. Is means
before verifying that all water is H2O, it already assumes that water is H2O.
However, the example of ‘water’ is unvalue inductive arguments. According to what we use the
predicate of ‘water’ in daily it usually refers to river, lake and sea. So, when we verify that all
water is H2O, during the verification process we will find sea or some kind of lake (Dead Sea) we
call ‘water’ is not H2O. This will lead to some premise ‘water (sea) is H2O, water (Dead Sea) is
H2O’ is false during the verification process. An unvalue inductive arguments is not what I want
although it seems pointed out we can’t distinguish the daily languages of water and the scientific
sense of water something like Goodmen’ view. But again, in this paper is not to arguing how to
help Goodmen’ view. This paper is based on I already agree with Goodmen’ view, and try to give
new example to replace the problematic ‘grue’ example. Therefore, our work must continue, to
find a valuable inductive argument, which is all premises is true, and the conclusion is contingently
true to replace ‘grue’ example.

Example of ‘Half-life’
Half-life refers to the process of decay of radioactive elements. The half-life of an element is the
time it takes for its radioactive nuclei to decay to only half their original number. Bismuth-209
(Bi) is the radioactive elements, but is only have extremely weak radioactivity, Bismuth-209 half-
life is as long as 1.9×1019 year, it more than a billion times the age of the universe. According to
Bismuth-209 stability, we can well confirm radioactive nuclei of Bismuth-209 is around 209
radioactive nuclei. Therefore, we build an inductive argument:

In this inductive argument, all premises are true. Because so far, or even in the future (nearly
infinite time of future), Bismuth-209 as an extremely stable radioactive element must have around
209 radioactive nuclei. Thus, we can conclude ‘all Bismuth-209 is around 209 radioactive nuclei’.
But at the same time, this conclusion is contingently true, as we can imagine in the infinitely distant
future Bismuth-209 will complete its first half-life, although the length of its half-life is negligible.
In the example of ‘Bismuth-209 Half-life’, it works perfectly as a replacement for the ‘grue’
example because it avoids all problems with the grue example. First, Bismuth-209 is an existing
predicate, it avoids the controversial predicate of ‘grue’. Second, this is a value inductive argument.
Third, from the argument, all premises are true, and the conclusion is contingently true.

Return to original
In this part, I'll give a very normal example, but this example may raise other philosophical
questions, these questions may require detailed explanations. But in this paper, I want to focus on
replacing problematic ‘grue’ example. Therefore, I will only get an example and some basic ideas
behind this example.
The idea behind this example is maybe Goodman and we think too much. We overcomplicate the
problem of inductive argumentation. Oe the other hand, we simply don’t need to invent or create
more complex examples to prove the problem of inductive argumentation. Actually we only
suggest the inductive argument like ‘Emerald is green’ and conclude that ‘all emeralds are green’
is good enough

Based on color science and color anti-realism view, none of us can objectively define color. For
example, if we are asking someone what is the unique green from the following figure, different
people will point out different colors.

Back to the example, normally we think emeralds are green, so generally we would agree with the
conclusion ‘all emeralds are green’. But different people may have different senses of green. Then
the conclusion ‘all emeralds are green’ is only subjectively correct. But the problem is, we cannot
reduce the conclusion deductively to be objective and formal like logical positivism. Because we
don’t have an objective definition of green. Therefore, ‘all emeralds are green’ is contingently true,
as it is only in subjective term.

conclusion
To conclude, Goodmen’ grue example may have some problem, it led unable to support his view.
Therefore, to make his point valid again, it is necessary to propose new examples to replace ‘grue’
example which is example of ‘Half-life’. On the other hand, we may be too complicated, common
examples are enough support Goodmen’ view, but this may involve other areas of philosophy or
science, we need to discuss further.