Re: 我想去埔心牧場烤肉
※ 引述《bobWerther (騙人，羊又來了)》之銘言：
: 可不可以請學長解釋一下何為possible world....
很慚愧
我在這方面亦是外行
若要做個較周延的解釋
恐怕我得翻箱倒櫃去把以前唸的那些資料找出
若您在學校
可以去請教楊金穆老師
或是博士班的蒲世豪學長
(就是那個頭髮留很長，常在研究生電腦室的那位)
而我記得好像劉福增老師在他的《語言哲學》那本書中也做過相關討論
以下我將Routledge的Encyclopedia of Philosophy中Possible worlds條目
的內文貼出，供您做參考
Possible worlds
The concept of Possible worlds arises most naturally in the study of
possibility and necessity. It is relatively uncontroversial that grass
might have been red, or (to put the point another way) that there is
a possible world in which grass is red. Though we do not normally take
such talk of possible worlds literally, doing so has a surprisingly large
number of benefits. Possible worlds enable us to analyse and help us
understand a wide range of problematic and difficult concepts. Modality
and modal logic, counterfactuals, propositions and properties are
just some of the concepts illuminated by possible worlds.
Yet, for all this, possible worlds may raise more problems than they
solve. What kinds of things are possible worlds? Are they merely our
creations or do they exist independently of us? Are they concrete
objects, like the actual world, containing flesh and blood people
living in alternative realities, or are they abstract objects, like
numbers, unlocated in space and time and with no causal powers?
Indeed, since possible worlds are not the kind of thing we can ever
visit, how could we even know that such things exist? These are but
some of the difficult questions which must be faced by anyone who
wishes to use possible worlds.
1 Pros and cons
Although known to Leibniz, it was not until the middle of the
twentieth century that possible worlds came to occupy the attention
of a significant number of philosophers and logicians. While this
was initially due to the discovery that possible worlds permitted the
formulation of a model-theoretic semantics for modal logic, it
gradually became clear that possible worlds could be used to shed
light on many other notions which had been regarded as problematic:
intensional logic, counterfactuals, propositions and properties are
but some of the areas illuminated by possible worlds. Indeed, some
have even argued that the existence of possible worlds is a tenet
of our ordinary common-sense beliefs. It is uncontroversially true that
there are many ways the world could have been - but what are these
many ways but possible worlds under another name?
However, although possible worlds may provide solutions to old
problems, they raise new difficulties of their own. Since possible
worlds are not the kind of thing with which we can causally interact,
how do we know that there are such things, and how have we come to
possess the kind of modal knowledge we credit ourselves with? Then
there is the problem of transworld identity: how one and the same
object can exist at different possible worlds. The problem arises
for the following reason: not only is it possible that there might
have been people who were twelve feet tall, Quine himself might have
been twelve feet tall. In possible worlds terms, this means that
there is a possible world at which Quine is twelve feet tall,
apparently making Quine exist in more than one world.
When it comes to questions about their metaphysical nature - what
kinds of things possible worlds are - there is very little agreement.
It is agreed that possible worlds must be capable of representing
many different ways the world could have been, and there can be no
possible worlds in which grass is both green and red; but whether
possible worlds are abstract or concrete, whether they exist
independently of us or are our own creation, whether they have
structure or are simples - are all contentious issues.
2 Extreme realism
For the extreme realist, all possible worlds are on a par: there is
no distinction in kind between the actual and the possible. All that
marks off the actual world from the infinitely many merely possible
worlds is that the actual world is the world we happen to inhabit.
Just as the actual world contains flesh and blood human beings, so
some merely possible worlds contain flesh and blood human beings.
Just as the actual world is made up of concrete spatio-temporally
extended objects, so other possible worlds are made up of concrete
spatiotemporally extended objects.
At first sight it appears that the extreme realist must solve the
problem of transworld identity by having one and the same object
literally appearing in more than one world. But there is an ingenious
alternative. Quine might exist at a possible world w not because
he is a part of w, but because w contains some other person (call
him 莜wine?) who resembles Quine. If Kwine and Quine share certain
properties, if they lead similar lives and have similar parents
then Kwine is said to be Quine旧 counterpart. It is possible for
Quine to be twelve feet tall if there exists some world which contains
a counterpart of Quine, and that counterpart is twelve feet tall.
Extreme modal realists use neither modal nor intensional concepts
in their theory. Accordingly, they can use possible worlds to analyse
non-circularly the concepts of necessity, of counterfactuals, of
propositions and of properties. Since the number of different concepts
taken as primitive is few, an advantage of this theory is that it is
ideologically parsimonious.
Yet for all its strengths, extreme modal realism has few proponents.
Although ideologically parsimonious, the theory is as ontologically
unparsimonious as possible: for any kind of object there could be,
be it a centaur, ghost or unicorn, there is some world which contains
such an object. Moreover, the very idea that there is an infinite
number of concrete possible worlds, all on an equal footing, simply
defies belief. The view that there really are non-actual Centaurs,
ghosts and unicorns goes against our strongly held common-sense
beliefs.
3 Combinatorialism
The combinatorialist takes possible worlds as recombinations of
the fundamental elements of the actual world. For any fundamental
property F and any atomic object a, the combinatorialist maintains
that it is possible that a instantiate F. Any collection of such
states of affairs counts as a possible world.
The restriction to atomic objects and fundamental properties is
needed to rule out impossible recombinations such as an object's
being simultaneously both round and square, or red and green.
It is a crucial part of the combinatorialist's idea that such a
restriction succeed in ruling out all impossibilities. But there
are reasons to think it does not. For there may be fundamental
properties which are nevertheless mutually exclusive. For example,
having a mass of one gram and having a mass of two grams may both
be fundamental properties - but nothing can be simultaneously both
one gram and two grams.
Not only are there recombinations which are not possible, there
are possibilites which are not recombinations. For it is possible
that there be things which do not actually exist, and possible that
there have been different fundamental properties not instantiated
in the actual world. Such possibilities cannot be constructed by
recombining only actually existing objects and properties.
4 Possible worlds as novels
On this view, worlds are nothing more than consistent works of
fiction which represent different possible states of affairs in
the same way novels do. It is possible that grass is red if there
is some consistent novel which contains the sentence 蠇rass is red?,
or some collection of sentences which entail that grass is red.
Since novels can contain names for actually existing objects, any
novel containing the sentence 葈uine is twelve feet tall? represents
Quine as being twelve feet tall- thus dealing with the problem of
transworld identity.
But though this theory neither inflates our ontology nor defies
belief (there are few who doubt the existence of novels!), as it
stands it simply fails to do justice to the facts of modality. For
not any book can count as a possible world. For instance, any
novel containing the two sentences' Grass is red? and 荛rass is
green? fails to represent a way in which the world could have been.
We could avoid this problem by identifying possible worlds with
those novels which could have been true - but there is a drawback.
If the concept of possibility is being used in saying what possible
worlds are, the concept of possibility cannot be analysed in terms
of possible worlds on pain of circularity.
Not only are there novels which do not represent possible worlds,
there are some possible worlds which are not represented by any
novel. For there are infinitely many different lengths an object
could be. Accordingly, there are an infinite number of different
possible worlds. Yet there is only a finite number of novels.
Moreover, there are merely possible worlds which contain as much
complexity and detail as the actual world around us, yet we
mortals have no hope of writing a novel describing such a world
in such fine detail. Some possible worlds appear to elude our
descriptive capabilities.
If we are to have any hope of reducing possible worlds to novels,
then our novels cannot be the kinds of things we actually read or
write. Instead, our novels must be identified with sets of sentences
in some idealized abstract language, a language which, though not
capable of being written or spoken, is capable of representing
many more possible worlds. One way to implement this idea is to
construct a language the terms and predicates of which are actual
objects and properties and the sentences of which are set-theoretical
constructions thereof. On this scheme, the ordered pair (F,a) is
interpreted as saying that a is F. Possible worlds can then be
identified with consistent sets of sentences.
Though such a powerful language enables one to represent an infinite
number of different possibilities, it is still a moot point whether
all the possible worlds can be represented. Moreover, the increase
in power comes at a price. For possible worlds are now identified
with mathematical objects: sets of ordered n-tuples. Unlike novels, t
he existence of mathematical objects is a contentious issue.
Finally, some fear that this theory does not provide an objective
theory of modality. For sentences have the meanings they do in
virtue of the way we interpret them: had there been no humans,
there would have been no meaningful sentences. Accordingly, possible
worlds owe their existence to us. This is difficult to square with
the view that whether or not something is possible is a matter which
is independent of human thoughts and beliefs. 'Possibly P' cannot
be equivalent to 'There is a possible world in which P' if the
former is objectively true but the latter holds only subjectively.
5 Moderate realism
Instead of taking possible worlds as sets of sentences, perhaps
possible worlds should be constructed out of the meanings of
sentences. Perhaps possible worlds should be taken as consistent
sets of propositions. Obviously, such entities must be understood
as existing independently of our own thought and language if worlds
are not to owe their existence to us. (Some variants of this theory
replace propositions with states of affairs or abstract properties.)
This theory faces some familiar problems. As above, worlds are
abstract. As above, since it is only those sets which could have
been true which form the possible worlds, the theory cannot analyse
possibility without circularity. Of yet greater concern is the fact
that this theory takes the notion of proposition as primitive.
Admitting such things as a new fundamental kind of object is
ontologically costly, and some find the idea that there is an
infinite number of abstract representations 裵ut there? every bit
as ludicrous as extreme realism. Moreover, the notion of proposition
had come under a lot of fire, and one of the more attractive features
of possible worlds was that they promised to throw light upon these
peculiar creatures. Yet how can this promise be fulfilled by a
theory which itself takes this problematic notion as primitive?
--
※ 發信站: 批踢踢實業坊(ptt.csie.ntu.edu.tw)
◆ From: 218.187.92.85
推
推 203.203.131.8 07/17, , 1F
推 203.203.131.8 07/17, 1F
討論串 (同標題文章)
本文引述了以下文章的的內容：
完整討論串 (本文為第 8 之 8 篇)：