Monday, January 14, 2013

Anti-introductionism. Treatise No 8.



Summary.

The subject of this treatise is to discuss Zeno’s paradoxes with runners on the basis of immediate considerations.
Some philosophers understand the concept of limit of a sum-function as being identical with the concept of a sum of infinitely many numbers, as far as such a sum exists. It proves, however, that the solution of these paradoxes is found in (1) the way the story of the race between Achilles and the tortoise is told, which can be rendered recursively without a stop. Moreover, the right understanding of the concept of infinity proves to dissolve an alleged paradox concerning (2) how it is possible to have passed infinitely many points (in a finite time). This corresponds to the difference between respectively 1) being in the present and 2) considering a past-construction.
A later treatise deals with the discussion in Wesley C. Salmon’s anthology Zeno’s Paradoxes ([Salmon 70]).


1. Achilles and the tortoise.

The paradoxes with runners of Zeno of Elea can obviously be solved by mathematical formalism, e.g. that of the race between Achilles and the tortoise. However, the question is whether they deal with the mathematical world or with the physical world, i.e. the phenomenal, in which not all theoretical concepts have a counterpart. The fact that the theories of natural science can be used to predict the phenomena is a different matter, as this does not demand an analogy between the mathematical and the physical world, but just their usability or being empirically adequate, such as Bas van Fraassen argues in The Scientific Image ([Fraassen 80], p. 12, 18).

1.1 A description of the paradox.

In the popularized version of one of Zeno’s paradoxes, Achilles and the tortoise, Achilles has to catch up with a tortoise that has a lead. It must be observed that the paradox consists of two parts: 1) our common sense view of the race, which is based on memory impressions of finished processes, and 2) Zeno’s account of it.
According to the implicit reasoning in (2) Zeno’s account of the run, Achilles will never catch up the tortoise. Nevertheless, we will assert (1) that we know that Achilles will catch up with the tortoise. This is a paradox concerning our conception of time. Of course, Zeno has a point here.
Re 1: Zeno’s reasoning can be recounted in this way: Let us assume that Achilles runs G times as fast as the tortoise, and that the leap is F units of length. According to principles that are in accordance to our experience, it can be figured out that that Achilles has cached up with the tortoise when he has run the distance F G/(G - 1).
Re 2: When Achilles has run F units of length, the tortoise has run F/G. When Achilles has cached up with this lead, the tortoise has run F/G2. When Achilles has run F/G2, the tortoise has run F/G3. Etc. After each of these parts of the run, the tortoise still has a lead (that certainly gets smaller and smaller, but never zero).
This means two things: 1) Each time one of these stages has been finished, Achilles has yet to traverse a stage. I.e. that Achilles is always in the same situation. 2) In a situation where Achilles has caught up with the tortoise, he must have run through infinitely many stages. This is another aspect of the paradox above. The first one can be comprised in:

1.1.1 A recursive description of the race.

After each of the described parts of the race, the same situation is present, except that the tortoise has a lead that is 1/G times what it was at the start of the run.
We can call the length of the stretch to the starting point of the tortoise d, and relation between Achilles’ velocity and that of the tortoise g. Then, the race between Achilles and the tortoise can be described by an account, B(d).
B(d): “Achilles catches up with the tortoise that has a leap d”.
This account has the following recursive form where the expression R(s) means, “Achilles runs the distance d”:
B(d) -> R(d) + B(d/g).
If the initial leap was F and the relation between the velocities g = G, the whole account can be expressed as
B(F).
If this account is regarded as a pure mathematical description, the situation after each step is quite the same as at the start of the race, for there are no absolute lengths on the line in geometry. Thus, according to this mathematic description, no changes are present that imply that the race will stop.
Thus, from the description above we get:
B(F)    -> R(F) + B(F/G).
B(F/G)  -> R(F) + R(F/G) + B(F/G2).
B(F/G2) -> R(F) + R(F/G) + R(F/G2) + B(F/G3).
B(F/G3) -> R(F) + R(F/G) + R(F/G2) + R(F/G3) + B(F/G4).
This recursive description can be compared to a recursive description that contains a not-recursive case that will even be satisfied and thus finishes the description.
Let “Run(N)” mean: The runner runs the remaining N/M parts of the stretch, and let R(M) mean: ”The runner runs 1/M of the whole track”. Then, the run can then be described by these recursive definitions:
Run(0).
Run(N) -> R(M) + Run(N-1).
Then the run can be described in this way:
Run(M).
Example. M=3:
Run(3) -> R(M) + Run(3-1).
Run(2) -> R(M) + Run(2-1).
Run(1) -> R(M) + Run(1-1).
Run(0).

1.2 Mathematical proposals for solution.

This subsection looks at two mathematical proposals for solution based on mathematical formalism for so-called summation of infinite many values. It proves that none of these attempts at solving the paradox succeeds. The first attempt, which consists in a simple computation, is circular, while the other cannot make good that which takes place in the description of the race between Achilles and the tortoise will lead to a termination.

1.2.1 Summation of infinite many values.

We can attempt to let the paradox deal with a summation S of all the steps until Achilles has caught up with the tortoise according to the common sense view[1]:
S = F ∑i=1 G-i+1
From this, we get:
S = F +F ∑i=2 G-i+1
S G = F G + F ∑i=2 G-i+2
S G = F G + F ∑i=1 G-i+1
By this, we get the following computation of the summation:
S G = F G + S
S = F G/(G – 1)
However, this calculus with a thought summation of infinitely many numbers presupposes that the summation can be carried out, i.e. that the paradox is solved. Thus, understood as a solution of the paradox, this attempt is circular. The circularity is committed in the line in which the summation is substituted by S, as this substitution presupposes that the summations understood as a summation of infinitely many numbers can be carried out, and that it has semantic meaning.

1.2.2 Mathematical limit.

In mathematics, the summation symbol including its parameters denotes the limit of Sn for n going towards , i.e. for n increasing without limit. Thus, it does not denote a summation of infinitely many numbers in any literal way.
More specifically, a function f(x) is said to go towards the limit b for x going towards, if
"εÎR+: $hÎN: "x: x > h => 0 < |f(x) - b| ≤ ε.
This definition can be applied to
Sn = F ∑ni=0 G-i.
This means that the following must be proved:
"εÎR+: $hÎN: "n: n > h => 0 < |Sn - S| ≤ ε, where Sn and S can be substituted by the expressions in the above subsection, after which the correctness can be confirmed.
However, this mathematical formalism does not solve the problem, since as just mentioned it does not deal with any sum, but with something else, for the definition above just means this:
A limit of a sum of n numbers for n going towards infinity is not a sum of infinitely many numbers in some literal sense, but is a value to which this sum of n numbers can get arbitrarily close without coming farther away for any higher n.
Our talk about infinitely many elements in a certain set does not mean that the set contains a number of elements we can call “infinite” just as when we talk about there being 1017 elements in a set, but that the set of these elements has no end, i.e. that we can continue pointing out new elements in the set. Here we have an ordered set of infinitely many intervals on a piece of the line, which means that we can continue pointing out new, still smaller intervals, even succeeding ones in the ordered set. The proof above is not in contradiction to Zeno’s paradox, but rather copies it.
Finally, it must be observed that the above definition only means that Sn comes relatively close to S, but not absolutely closer, as there are no absolute distances in geometry. It is even so that the division of the line from Sn+1 to S has the same shape as the division from Sn to S. So much the more the runner’s situation is not changed.

1.3 An unfinished account.

The following attempt at a solution is inspired by the fact that if Achilles velocity is v, the sum of the first n intervals of time is Sn/v. Number n of these time intervals has the duration F/(Gn-1 v). This lead to that solution of the paradox about Achilles and the tortoise that after each of its steps the above account of it has only concerned a certain space of time and not the whole run.[2] Thus, it only deals with what takes place within a sum of time intervals each of which has become still smaller until some moment before and arbitrarily close to the point where we would expect that Achilles would catch up with the tortoise:
It is thus not the race that never finishes but the very account of the race that is put in such a way that it never finishes, and therefore it self-evidently does not deal with the whole fiction about the race. What is the case is therefore that Achilles does not catch up with the tortoise within the space of time that the account deal is supposed to deal with. The reason to this is that the account has the above-described recursive form.
We can imagine that that we have just attended the race between Achilles and the tortoise and seen Achilles catch up with the tortoise. Nevertheless, we can subsequently start describing the race in the following way that can never end:
The tortoise started with a lead of the length F. Achilles ran G times as fast as the tortoise. At first, Achilles reached the place from which the tortoise started. At that time, the tortoise had moved a stretch of the length F/G. Thereafter Achilles reached the place where the tortoise was after the first stage, etc.
As mentioned, this account can never end for the reason that it is told in such a way that the sum of these intervals of time at no point exceeds the duration of the race.
However, it is also told in way that makes clear that Achilles’ situation is in principle is unchanged for each stage.
Does this mean that the paradox is solved? It can be put forward that the account just points out a number of points one after one on the time axis to which this fiction about past can be attached. This takes place according to a definite algorithm defined by the account. It is not paradoxical that, just as we cannot finish many other algorithms, we cannot finish this algorithm either. In fact, we can both put forward and combine many algorithms to division of a piece of line or a track into intervals, even with more than one point of accumulation. These can of course be passed. It can also be put forward that they just make up an intellectual construction that as such has nothing with the race to do. However, this way of looking at the paradox is to ignore Zeno’s message.

1.4 Scientific limitations.

After this going through, scientific considerations concerning the paradox are strictly speaking irrelevant. However, this shall not prevent us from looking at a scientific based proposal for a solution of the problem the account obviously implies, viz. that Achilles only gets arbitrarily close to the tortoise, but never catches up with it. The proposal is based on the idea that the mathematical distances can be insignificant according to scientific theories. Thus, it could be asserted that when, for the one or the other reason, the difference in position cannot be measured, Achilles and the tortoise have reached the same position on the racing track. According to this, the race would simply be finished when the distance between the two competitors and less than a certain magnitude. On its own premises, however, this does not explain the fact that Achilles not only catches up with the tortoise, but also runs past the tortoise, i.e. also with a length of half a point.
Below it is explained that this proposal for a solution does not fall in with Zeno’s premises, but just rejects the paradox. This will also be discussed in a later treatise.
Zeno’s account cannot fully deal with the physical world, for it is a question, which two physical points on the tortoise and Achilles have to be compared during the race, especially as they are moving. If we also involve Achilles’ steps and movements, when we talk about minimal distances, the discussion becomes boundlessly complex. Thus, we must regard the race as dealing with two points A and B that are moving along a line. Mixing physical contingencies into the paradox is not to look at its fundamental character. On the other hand, the mathematical definitions have to be correctly understood and their terms not to be literally understood.

1.5. Zeno’s dichotomy paradox.

In The Presocratic Philosophers, it is advanced about “Zeno’s arguments about motion ...”:
“The first asserts the non-existence of motion on the ground that that which is in locomotion must arrive at the half-way stage before it arrives at the goal…” ([Kirk 1983] p. 270)
Because of its short form, this argument it can be interpreted in two ways. (Cf. [Kirk 1983] p. 270.)

1.5.1 The right recursive dichotomy-paradox.

One of the interpretations reminds of the paradox about Achilles and the tortoise, and is right recursive like this paradox:
To reach his goal a runner must first run halfway to the goal, then the half of the rest, etc.
Let us presuppose that the goal is the point 1 on the x-axis, and define the statement R(s).
R(s): “The runner runs a distance of the length s”.
We can then define a recursive account C(d) about how the runner runs from the point d to the point 1:
C(d) -> R((1-d)/2) + C((1+d)/2).
The whole account can thus expressed by:
C(0).
The first four stages have the following form:
1: C(0)   -> R(1/2) + C(1/2).
2: C(1/2) -> R(1/4) + C(3/4).
3: C(3/4) ->  R(1/8) + C(7/8).
4: C(7/8) -> R(1/16) + C(15/16).
This gives the following composite result:
1-4: C(0) -> R(1/2) + R(1/4) + R(1/8) + R(1/16) + C(15/16).
The account is analogous to that about Achilles and the tortoise. Thus, after each stage, the situation in this account is the same as before the stage. For the remaining account still deals with the fact that the runner is running a stretch that can described in the same way as before the stage except that its length is here halved for each stage. Thus, we are again dealing with an account that due to its form never finishes. Moreover, it must be observed that just as it was the case with the paradox about Achilles and the tortoise, after each stage no real change has occurred, as there are no absolute distances in the mathematical universe.
We can imagine that that we have just attended the runner run a distance of 1 length unit. Nevertheless, we can start describing the run in the following retrospective way that never reaches an end:
The runner started from point 0. At first, he reached the middle of the distance. Then he reached the middle of the remaining distance, etc.
As just mentioned, this account will never end for the reason that it is told in such a way that the sum of these intervals of time at no point exceeds the duration of the run and especially that the runner’s situation remains unchanged in principle.

1.5.2 The left recursive dichotomy paradox.

The other interpretation of the paradox is left recursive and can be rendered in this way:
To reach a point P halfway before his goal, a runner must first run to a point halfway before the point P, etc.
Let us presuppose that the goal is the point 1 at the x-axis, and define the statement R(s):
R(s): “The runner runs a distance of the length s”.
We can then define a recursive account C(d) about how the runner runs from the point 0 to the point d:
D(d) -> D(d/2) + R(d/2).
By means of this, the whole account can be expressed by:
D(1).
The first four stages have the following form:
1: D(1)   -> D(1/2) + R(1/2).
2: D(1/2) -> D(1/4) + R(1/4).
3: D(1/4) -> D(1/8) + R(1/8).
4: D(1/8) -> D(1/16) + R(1/16).
This gives the following composite result:
1-4: D(1) -> D(1/16) + R(1/16) + R(1/8)+ R(1/4)+ R(1/2).
By this, it has been illustrated that the account D about the whole race has to be finished before the account of runs of the single stages can begin. This has already been established by the left recursivity. Therefore, the account D will never explicitly mention or concern the start of the run in any of its stages. This version is retrospective, and regarded in this way, it is analogous to the first version.
We can again imagine that that we have just attended the runner run a distance of 1 length unit. Nevertheless, we can start describing the run in the following way that never will end:
The runner started from point 0. Before he reached the middle of the distance, he had to run half way to this point. Before he reached the latter point, he had to run half way to this point, etc.

1.7 Aristotle’s arguments.

Aristotle asserts that length and time can be called “infinite” as regards both divisibility and their extremities, and concludes from this:
“So while a thing in a finite time cannot come in contact with things quantitatively infinite, it can come in contact with things infinite in respect of divisibility: for in this sense the time itself is also infinite: and so we find that the time occupied by the passage over the infinite is not a finite but an infinite time, and the contact with the infinites is made by means of moments not finite but infinite in number. (After Gaye)” ([Kirk 70], p. 270)
Aristotle naturally has a point in saying that a piece of line can be both 1) infinitely long and 2) infinitely divisible, i.e. also if its length is finite, such as Zeno demonstrates. Moreover, it is clear that about a runner’s passage of a line piece we can say the same about the division of the time the run lasts as about the division of the line piece traversed. For the passage of the line piece and the passage of time goes together. This means, however, that with his above pointing out Aristotle does not solve the paradox, but only shows two sides of it.
Thus, we can again look at the situation in which a runner has reached his goal. This is a paradox, as it is impossible to him according to Zeno’s account. For in order to arrive at it, the runner first must run half-way, and to arrive at that point he must run half-way of this distance, etc. However, if the runner’s velocity was h, he reached the halfway after the time (1/2)/h, he reached 1/4 of the distance at the time (1/4)/h, etc.
Obviously, there is no end to the related problems we can think over in connection with the three discussed paradoxes. According to The Presocratic Philosophers, from Aristotle’s’ discussion the following left-recursive reasoning can be deduced:
“(1) To reach his goal a runner must touch infinitely many points ordered in the sequence 1/2, 1/4, 1/8, ...
(2) It is impossible to touch infinitely many points in a finite time.
So
(3) the runner cannot reach his goal.” ([Kirk 70], p. 270)
According to Aristotle’s argumentation above n, it is possible to pass infinitely many points in a finite time, wherefore (2) is false. As this argument, however, deals with a space of time regarded as a whole, it does not deal with the runner’s situation, which is in the now. Therefore, the demonstration that (2) s false does not solve the paradox, but only illustrates one aspect of it.
Aristotle puts forward that this argument presupposes that the single intervals only exist potentially. However, Aristotle wants to discuss this condition:
”... when someone asks the question whether it is possible to traverse infinite things - either in time or in distance - we must reply that in a way it is but in a way it is not. For if they exist actually it is not possible, but if potentially, it is; for someone in continuous movement has traversed infinite things incidentally,...” ([Kirk 70], p. 271)
It is true that this applies to e.g. the retrospective description. The question is what it would mean that the intervals should exist actually, i.e. as intervals. The answer is that this is the case for the single interval when it is pointed out, which it is in the now. It is exactly intervals pointed out we have to do with in the recursive interpretation of Zeno’s description. It is true that according to this recursive account the run has no beginning, just as it has no ending according to the right recursive account.
It must be observed, that it cannot be the case that the intervals exist all of them actually as something different from existing potentially; for we cannot point them all out at one time.

1.7 Summarizing considerations.

As it is simpler to comment on the right recursive account C above, a concluding remark shall deal with this account. We can imagine that we are in a situation in which a runner has reached the goal, and we can then ask how the runner has been able to pass infinitely many distances of the lengths (½)n. The fact that there are infinitely many distances of this kind before the point 1 just means that for each such distance we can point at, we can point at one distance more. The fact that the runner has run infinity many of this kind of distances before the point 1, therefore just means that for each distance of this kind the runner has run, we can point at yet a distance the runner has run further along the running track. To each n corresponds the interval from 1-(½)n-1 to 1-(½)n. In this way, the runner passes infinitely many points within a finite time: there is no end of the set of points he touches in a finite time.
It can be noted that Zeno’s account of the run  deal with the present time, i.e. the runner’s situation, while the retrospective explanations deal with past fictions. We cannot explain the flight of time. We can only explain how it is to be in the present time and look at a past-construction, as here, and at a future-construction.
As already touched on we can certainly imagine and that a person objects: “A little while ago you were talking to me. That was both real and present at that time. Now it is past time. This proves that time is passing.”[3] However, this only proves the fact that we have memory impressions, and that the person in question is placing them in a past-construction. Naturally, an unsolved philosophic problem cannot be solved by insufficient explanations. It must rather remain unsolved for the moment, only illuminated.


Literature.

[Fraassen 80]      Bas C. van Fraassen: The Scientific Image
Oxford University Press (Oxford 1980)
[Kirk 83]          G.S. Kirk, J.E. Raven), M. Schofield: The Presocratic Philosophers, 2. ed.,
(Cambridge 1983)
[Salmon 70]        Wesley C. Salmon, ed.: Zeno’s Paradoxes
Bobbs-Merrill (Indianapolis & New York, 1970)




[1] I have seen this argumentation somewhere, but have not been able to find it again.
[2] Here I owe to refer to Aristotle who discusses this subject in ([Kirk 70], p. 270).The subject id treated in the subsection “Summarizing considerations”.
[3] Cf. the subsection, “The present and past time and the future of the past” in Treatise No 1.

No comments:

Post a Comment