аЯрЁБс>ўџ bўџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ§џџџўџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџRoot EntryџџџџџџџџџџџџўџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџRoot Entryџџџџџџџџ КѕаcData
џџџџџџџџџџџџ1TableџџџџZObjectPoolџџџџџџџџ Кѕа Кѕаџџџџџџџџ§џџџўџџџ
!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYўџџџ[\]^_`aўџџџўџџџўџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџWordDocumentџџџџџџџџџџџџ&ЈџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџьЅСG Поbjbjйй &ЈьГьГчGz
џџџџџџ](<PPPP8$ЌkЖд02$!єV<<фф<<<<<,ШзЦbt'Нє SEQ CHAPTER \h \r 1Comments on Frederike Moltmann s Variable Objects and Truth-Making
I have long admired Frederike Moltmann s work at the interface of linguistics and
philosophy; and I have always been especially impressed by the way in which she has attempted to break free of the stranglehold of the possible worlds approach by showing how the diversified ontology of objects commonly associated with traditional metaphysics provides a much better tool for the investigation of natural language than the simple and stylized ontology of Montague grammar. The present paper is a characteristically rich, original and thought-provoking contribution to the subject; and I am afraid that I can do no more than pick my way through one or two of the many issues that she raises.
Moltmann is interested in phrases such as the book John needs to write . Many philosophers would wish to deny that such phrases are genuinely referential. To say that the book John needs to write must have great impact is, for them, just a roundabout way of saying it is necessary that the book John writes has great impact ; and, under this paraphrase, any inclination we might have had to take phrase the book John needs to write to be referential will disappear. Many linguists, perhaps more attuned to the demands of compositionality, will grant that the phrase is referential but take it to refer to an individual concept, in the tradition of Montague semantics - a function which, in this case, will take each possible world in which John writes the book he needs to write into that very book.
Moltmann is not satisfied with either view. She thinks that there is clear linguistic evidence that such phrases are referential (p. 2) and she also thinks- quite apart from other, more technical, difficulties - that the individual concepts will not have the required low level features, such as being of a certain length or being the subject of tropes (p. 15).
But if these phrases refer and if they do not refer to individual concepts, or other objects of higher type, then to what do they refer? Let us call the kind of object to which such a phrase refers an I-object. Other examples include the dollar he owes or the screw that is missing or the the wife he hopes to marry. What, then, are I-objects?
Moltmann takes them to be what she calls variable objects , which are like the variable embodiments of Fine [1999] but which, instead of being manifested , or embodied , at different times in different worlds, are manifested in different situations. Thus consider the situation in which the need for John to write a book is exactly satisfied (i.e. the situation is wholly relevant to the satisfaction of the need). Then in this situation, John will write one book and the I-object, the book John needs to write, will be manifested, in the situation, as that book.
We previously considered a related proposal concerning individual concepts, under which we took the book John needs to write as the function taking each world in which the need to write a book is satisfied into the unique book in that world by which the need is satisfied. Of course, we no longer wish to take the I-object to be identical to the function; and we should also allow the world to include a reference to a time. But in that case, we could then take the I-object to be the corresponding variable embodiment - one which is manifested, at each world and time, by the value of the function at that world and time.
But as Moltmann points out this alternative account is subject to a uniqueness problem (pp. 1607). Perhaps John goes into overdrive and, at a given time, produces two books, each of which will satisfy the need to write a book. How, then, at that world and time, is the book John needs to write manifested? Clearly, it cannot be taken to be one book as opposed to the other. But nor can it be taken to be both, since it is a single book that satisfies the need, not a pair of books.
Moltmann s conception of an I-object as a variable object, indexed to situations rather than worlds, appears to avoid this difficulty. For in the one situation, in which John writes the one book, the book John needs to write will be manifested as that book and, in the other situation, in which John writes the other book, the book John needs to write will be manifested as the other book.
However, it seems to me that Moltmann s account is subject to similar difficulties. For one thing, she assumes that in any exact satisfier of the need there will only one object of the required sort by which the need is satisfied. This is perhaps plausible in the given case but there is no reason to think that the assumption will hold in general. For consider now the paper John needs to write in conjunction with a related paper. This is, of course, a paper which must be written in conjunction with a related paper; and so any situation in which the need is satisfied will be one in which John writes two papers, not just one. But then, just as before, there is no good reason to think that the I-object is manifested in that situation as either one of the papers, to the exclusion of the other, or as both taken together.
A related difficulty arises from the need to fuse different situations. Go back to the original case in which each exact satisfier of John s need to write a book will involve just one book and consider again the case in which John simultaneously writes two books which satisfy the need. Then in one situation s1 the I-object, the book John needs to write, is manifested by the one book and, in another situation s2, it is manifested by the other books. Now I take it that the two situations are compatible (since otherwise we would not be allowing for the possibility that the need might be simultaneously satisfied by two books); and so there is a possible situation s = s1 " s2 in which John writes two books which satisfy the need. But how then is the I-object manifested in s?
You might think there is no need to say. But consider the counterfactual if John has written all the books he wanted to write then he would have written the book he needs to write and let us suppose that the books he wants to write include both of the books that would satisfy the need. We therefore need to evaluate the consequent within a situation which includes both s1 and s2; and so we are again forced to make an invidious choice between one of the books or the two together.
The same problem does not arise on the possible worlds approach since, when we evaluate a statement concerning an individual concept at a world, we can take the statement to be about the individual picked out by that individual concept at the world. But under the truthmaker approach, we may be forced, as we have seen, to evaluate a statement concerning a variable object at a situation at which the variable object is not defined or can only sensibly be taken to be multiply defined. Thus we see that, under the truthmaker approach, the uniqueness problem is not so much avoided as postponed.
A more serious difficulty still concerns indiscernible I-objects. Suppose that there are two missing screws, of exactly the same type, from the kit I recently bought or that there are two dollars that I owe you from a bet. There are then two I-objects in each case, the two missing screws or the two dollars; and they are absolutely indiscernible from one another, with neither of the two I-objects possessing a qualitative feature not possessed by the other.
This would then rule out any treatment of the I-objects in question as individual concepts or as variable objects. For the underlying space of possible worlds or of possible situations might be Leibnizian and not admit of any indiscernibles. But this would then mean that any two individual concepts or any two variable objects will be discernible from one another.
We should also note that the existence of indiscernible I-objects will give rise to indeterminacy. Suppose I replace one of the missing screws. Then which one have I replaced? Clearly,there is no determinate answer. And similarly, henI pay back one of the dollars I owe, there is no determinate answer as to which of the two dollars it is.
Again, there is nothing in the previous accounts of I-objects as individual concepts or variable embodiments or as variable objects which could account for such indeterminacy. Clearly, a much more radical account of I-objects is required if we are to deal with difficulties of this sort. Fortunately, I think a solution is close at hand. This will involve our taking more seriously the theory of arbitrary objects in Fine [1985] as a basis for defining I-objects; and so let me say a little more about what this theory is and how it contrasts with the theory of variable embodiments.
Arbitrary objects have values, perhaps in much the way that variable embodiments have manifestations. However, arbitrary objects, in contrast to variable embodiments, may stand or fail to stand in dependency relations. Thus an arbitrary real x will take all reals as its value independently of all other arbitrary objects. On the other hand, the variable y = x2 will depend upon the variable x and will take as its value the square of the value taken by x and z = 2y will depend upon y (and indirectly upon x) and take as it values twice the value taken by y and hence twice the square of the value taken by x.
Three features of the theory (or, rather, of one development of it) are of special note. First, arbitrary objects may be multi-valued. This is clear for independent arbitrary objects. The arbitrary real x, for example, takes each real number as a value. But it may also be true of dependent A-objects. Suppose, for example, that y = Б"x. Then for positive x, y takes two values, for x = 0 it takes one value, and for negative x it takes no values. Second, A-objects may be mutually dependent. Suppose I take x1 and x2 to be two distinct reals. Then each is dependent upon the other, subject to the requirement that the value of one should always be distinct from the value of the other. Mutual dependence may also obtain among A-objects that depend upon another A-object. Thus given an arbitrary nonnegative real x we may take y1 and y2 to be the distinct square roots of x. Thus when x = 4, y1 and y2 may take the respective values 2 and -2 and they may also take the respective values -2 and 2. Finally, arbitrary objects may be indiscernible from one another. Two arbitrary (and independent) reals x1 and x2 are absolutely indiscernible; there is nothing to distinguish them but their mere identity. But the dependent reals y1 and y2 above are also indiscernible from one another.
The theory of arbitrary objects and variable embodiments were developed independently of one another and for entirely different reasons. But it is perhaps not too much of a stretch to regard variable embodiments as a special case of arbitrary objects. Suppose that we have a variable embodiment which manifests itself at different world-time pairs. We may then let i be an index variable which takes different world-time pairs (w, t) as values and we may let the variable embodiment itself be the arbitrary object which depends upon i and whose value, for any given world-time pair (w, t) as value for i, is the manifestation of the variable embodiment at (w,t).
There are certain difficulties that will need to be worked out if the assimilation is to be regarded as satisfactory. One concerns uniqueness. There are many different index variables i which will take the world-pairs (w, t) as values and we do not want each of them to lead to a different variable embodiment. Perhaps one might take care of this difficulty by supposing that for any range of values - all natural numbers, say - there will be many indiscernible independent arbitrary objects (each an arbitrary number) which take the numbers as it values but that there will also be a unique arbitrary object (the arbitrary number) which takes the individuals as values, only distinguished in kind from the others by the fact that it is unique. The unique index on which each variable embodiment depends will then be the index i, as opposed to an index i. There is also the question of how we deal with the modal dimension of variable embodiments within the framework of arbitrary objects. One way is to include worlds among the values of the arbitrary objects, as we did above. But an alternative, which I prefer, is to develop a modal theory of arbitrary objects, in which we talk about the values they might have without making any explicit mention of possible worlds. Another question concerns the mereological status of variable embodiments. We would like y0 to constitute the variable embodiment y at the index i0 whenever y0 is the manifestation of y at i0. This suggests, more generally, that if y is an arbitrary object depending upon i, with y taking the value y0 for a given value i0 of i, then y0 should be taken to constitute y when i takes the value (or is constituted) by i0.
In any case, let us suppose that there is some way of assimilating variable embodiments to arbitrary objects. This then opens the way towards using the resources of the theory of arbitrary objects to solve our previous difficulties. One difficulty concerned the fact that, in any given possible world, there may not be a unique book that satisfies John s need to write a book. But on the arbitrary object approach, we can treat the book that John needs to write as a multi-valued arbitrary object which, for a given world-time pair, can take many different books as values.
This may not look like much of a solution. For we can, in the same way, take an individual concept to be multi-valued and take the corresponding variable embodiment to be multi-manifested. But the apparatus of individual concepts, within the framework of Montague semantics, and the metaphysical apparatus of variable embodiments gives us no way of dealing with the features of an individual concept or variable embodiment at an index at which it is multi-valued or multi-manifested. But the corresponding problem already has a solution within the theory of arbitrary objects. For according to the Principle of Generic Attribution, a statement about arbitrary objects will be true when it is true for all its values. Thus the statement that the arbitrary number is odd or even will be true, while the statement that it is odd (or the statement that it is even) will be neither true nor false. It is because we think of arbitrary objects as generic, as arbitrarily standing for its values, that we have no difficulty in attributing features of individual objects to them.
We also have no difficulty with indiscernible I-objects. Consider again the two missing screws, call them screw one and screw two. Each will be an A-object depending upon the arbitrary index i. In any situation in which both screws are replaced by s1 and s2, say, screw one and screw two will take the respective values s1 and s2 and also take the respective values s1 and s2, while in any situation in only one of the screws is replaced by s, screw one will take the value s while screw two is undefined and screw two will take the value s while screw one is undefined. In this way, the indiscernibility of the two missing screws is respected and their characteristic features preserved.
Finally, there is no difficulty over indeterminacy. Suppose that only one of screws is replaced with s. Then the statement that s is identical to screw one or identical to screw two (i.e. that one of the screws is replaced) will be true since under any assignment of values to screw one and screw two since one of the values will be s. However, the statements that s is identical to screw one and that s is identical to screw two will be neither truth nor false. Thus the indeterminacy of statements about I-objects will simply fall out of the indeterminacy of generic statements in general.
There are two further questions we have not discussed. One concerns the compositional analysis of phrases such as the book John needs to write . How can the desired denotation of this phrase be derived on the basis of generally acceptable compositional principles? I am not sure the A-object approach to I-objects has any advantage over other approaches in this regard. But nor, it seems to me, is at any special disadvantage.
The other question concerns whether I-objects should be indexed to worlds (or world-time pairs) or to situations. There is good reason, it seems to me, to think of needs and the like in terms of the situations in which they are exactly satisfied rather than in terms of the worlds in which they are inexactly satisfied. Suppose the department has a requirement of one book for tenure while the university has a requirement of one book, with or without another book. Then the department requirement can only be satisfied by writing a book whereas the university requirement can also be satisfied by writing two books. Since the satisfaction conditions for the two I-objects are different, so are the I-objects; and since the I-objects are satisfied in the same world, the difference in the two requirements cannot be captured under the world-based approach.
Consider now the book one needs to write to satisfy the department requirement and the book one needs to write to satisfy the university requirement. The situation-based approach would predict that these two I-objects are not the same, since the book one needs to write to satisfy the university requirement would be defined in a situation in which one writes two books (in which case it could take either book as value), whereas the book one needs to write to satisfy the department requirement would only be defined in situations in which one writes one book. However, it is not clear to me that there is any difference in the two I-objects or that there is any need to postulate a difference. And if this is so, then it means that the finer grain of individuation provided by the situation-based approach is doing no useful work in the present context.
References
Fine K., [1998] Cantorian Abstraction: A Reconstruction and Defense
Fine K. [1999] Things and Their Parts
Moltmann incidentally remarks like other INP s, identificational INP s require the definite determiner the . But I do not think this need be so for the other INP s. Thus I can replace a missing screw or pay back a dollar that I owe.
I also refer to arbitrary objects as variable objects and, in certain ways, this is my preferred terminology. It is therefore unfortunate that Moltmann uses the term variable object for the situational counterpart of variable embodiments. But for present purposes, let me here go along with her terminology.
I have also made use of uniquely individuated arbitrary objects in Fine [1998].
*,Иъњ.RZамBz(|!Д!Р$ј$О%і%,)Ў)://\//0
00
2222222 2ш2ъ2т5ф5ц5№5ђ5є56=8=\H^HBKDK&L(L.L0L2LpLrLьLюLјLњLMM(M*MVMXMКMМM N"NЪOЬOЪPЬPжPиPљїёьцьцьцьцьцьцьцьцьцьцьцьцрьцьцрьцьцрьцрьцьцрьцрьиьиьцьцьцвьцьцьцьцьцьцьцьцьцьцьцmH CJH*0JUmH CJmH CJH*mH CJ6mH CJmH CJ>*UCJmH USиPQQQQ8Q:QQQ6R8R:RFRHRJR TЂTМTОTРTЪTЬTЮTUU2U4UBUDUFUPURUTUшVъVьVіVјVњVцWшWъWєWіWјW>[B[\\]]__aЌafbbddddj@jBjjjjjрjтjфjHnn&tftДxЦxаxшx`ybyжyиyкyфyцyшyіy
zz&zfzhzjztzvzxzТzФzЦzаzвzдzX{Z{^{p{{{Њ{М{р{ђ{||,|>|~~T~V~x~~И~Р~~Ќю№02Tfz|А:МЦћѕћѕяћѕяћѕћѕяћѕћѕћѕяћѕћѕћѕћѕћчћѕяћѕяћѕћѕћѕяћѕяћѕяћѕяћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћѕћmH CJ56mH CJH*mH CJ6mH CJ[ЦH~мркђ|~Юа ІFH~ІЈЊ РТФЦЪЬЮољєљєљєљєюєюєшхсхсхсхшхшхпнпн<UmH 6mH
UmH H*mH CJ>*mH CJmH CJ6!ЖИ^
xTd2#H&Ъ,ђ2Ц6t;?BДDZI*N\X]шjpoмwN}јњњђюююююююююююююююююююююююю1$1$а1$$ј\
жик№ђ~ЮЈРТЦШЪЮаћћћћћііћћђђђћйббйб1$/.1$$dЄ#$&$+D:D-D1$1$$1$авджикмођююхююю1$dЄ1$1$dЄ/.$Аа/ Ар=!А "А #$ %АА Аа
[(@ёџ(NormalCJmH <A@ђџЁ<Default Paragraph Font>@>
Footnote Text 1$аmH H*8&@Ђ8Footnote ReferenceH*.")0oJэ(y|oJЈџџџџ
иPФiЦоMNOPјаоQRSоoJџџCorel CorporationoJ@GTimes New Roman5Symbol3&Arial#аот1чот1чч!#Рўџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџџ