Wittgenstein ’ s unique “ Great Analysis ” : a consequence of the construal of propositional sense as truth-conditions

The goal of this paper is to clarify an idea proposed by Wittgenstein in the Tractatus that every proposition has “one and only one complete analysis” (TLP, 3.25), the “Great Analysis” as we are going to call it. This new construal will not be modeled on Russell’s theory of descriptions. We are going to argue that Wittgenstein’s approach in the Tractatus deviates from that of Russell’s in some important details, although it was obviously inspired by it. Our working hypothesis will be that Wittgenstein’s main motivation for this alternative way of conceiving the process of analysis was an elucidation of propositional sense exclusively as truth-conditions and nothing more. According to our interpretation his proposal of a “Great Analysis” was due to a radical observance of that construal of sense, a precept which took him much further and leads him to a much more radical philosophical position than his philosophical interlocutors, Frege and Russell. In order to accomplishing this goal we will show how this process of analysis of propositions could be carried on. Our second task will be to present an explanation of how it should end and an evaluation of all the consequences of its adoption.


Introduction
The goal of this paper is to clarify an idea proposed by Wittgenstein in the Tractatus that every proposition has "one and only one complete analysis" (TLP, 3.25), the "Great Analysis" as we are going to call it.This new construal will not be modeled on Russell's theory of descriptions.We are going to argue that Wittgenstein's approach in the Tractatus deviates from that of Russell's in some important details, although it was obviously inspired by it.Our working hypothesis will be that Wittgenstein's main motivation for this alternative way of conceiving the process of analysis was an elucidation of propositional sense exclusively as truth-conditions and nothing more.According to our interpretation his proposal of a "Great Analysis" was due to a radical observance of that construal of sense, a precept which took him much further and leads him to a much more radical philosophical position than his philosophical interlocutors, Frege and Russell.In order to accomplishing this goal we will show how this process of analysis of propositions could be carried on.Our second task will be to present an explanation of how it should end and an evaluation of all the consequences of its adoption.
The present proposal was partly inspired by Porto (202) on how to interpret the logical analysis in the Tractatus.I want to thank Professor Ghisoni for his comments and suggestions on an early draft of this paper.I want to thank also my students Bruna, Igor e Paulo, for their questions and debates.Before presenting a summary of our argumentation, one important point has to be emphasized: our goal is not to solve all the various difficulties involved in the Tractatus' idea of analysis.Quite on the contrary, our proposal is rather critical, especially concerning the possibility of advocating the logical independence of elementary propositions and also in respect to providing a final explanation of the nature of "simple objects".Nevertheless we think we do have some crucial remarks about both these two radical demands, logical independence of elementary propositions and simple objects.So, as the reader will realize in our concluding observations, our alternative way to read the Tractatus' analysis has a critical purpose: pointing to an inevitable internal relation between elementary propositions which, as we all know, Wittgenstein came to realize only latter.We can call that "a residual elementary semantical holism", for it will spring from the presence of unavoidable predicates at this more fundamental level.
In our exposition, we will first explain why we choose this name, "Great Analysis", and how it emphasizes some differences between Wittgenstein's proposals and that of Russell.We then move on to the process of analysis itself, considering its beginning and its motivation.The fourth section introduces Quine's paper "Identity, Ostention and Hypostasis".We will try to explore in detail how his nominalist approach deals with singular terms like "this river" and why it involves a translational procedure.After presenting Quine's nominalism we are going to explain why his approach is illuminating to our discussion of Wittgenstein's "Great Analysis".The fifth section will be about the connection between the two philosophers, i.e., about the construal of propositional sense as truth-conditions and how this construal approximates their view on several matters.In the sixth section we are finally going to use Quine's nominalism and his view of bodies as complexes, that is to say, as extended process in space-time, to explain the method that might have been what Wittgenstein had in mind in order to implement his analysis.In this section we also discuss James Griffin's ideas about Wittgenstein's logical atomism and compare it to our own proposal concerning the propositional level.Up to that point we will be dealing only with propositions and considering the elementary ones as the end of the process of analysis.
The elementary propositions are not the end of our Great Analysis's whole story though, for there are some consequences of what we are proposing also to the elementary sub-propositional level.So, the seventh and final section of our paper will be about this elementary sub-propositional level, which includes the genuine names and their ontological counter-part, the simple objects.This section will be an attempt to do three things: to comment on these Wittgenstein's unique "great analysis" volume 18 número 1 2014 sub-propositional elements; to criticize Griffin's construal of "simple objects", which, following Hertz, he took to be "material points" and finally to present an alternative proposal of our own, which better fits Wittgenstein's "Great Analysis", at least according to our own construal.This alternative proposal has similarities to David Hyder's suggestion in important points.Our aim in adopting this alternative construal is only to emphasize certain very important consequences of our elucidation of the "Great Analysis": () objects cannot be taken "to b e in the logical space", for they have to be considered only as theoretical devices for solving certain further problems regarding elementary propositions; (2) if we adopt a construal of propositional sense exclusively as truth-conditions, either we will have to accept only names within the internal structure of elementary propositions -"… subject-predicate propositions cannot be elementary" (GRIFFIN,964,p. 57), or else we will have to adopt some other holist approach to explaining the meaning of propositions, as Quine as well as Wittgenstein himself latter did.

An initial characterization.
In this first preliminary section, we are going to characterize in general terms our main topic: Wittgenstein's "Great Analysis", comparing it, albeit very briefly, with the analysis suggested by Russell.Differently from some quite well known interpretations -such as Anscombe (97, p. 7/4), Max Black's review of Griffin's Book (966), Stenius (960,p. 65/39/206), Hacker (989,p. 30), Frascolla (2007, p. 58/42/45/25) and in Brazil, Cuter (2009, p. -3) -our proposal will not involve viewing Russell's "Theory of Descriptions" as the model for Wittgenstein's Analysis, although we don't deny some similarities between them.We are not alone in this alternative construal, though.Griffin's book (964, provides a whole battery of arguments against this "Russellian descriptivist approach", which we are going to review in sections 6 and 7. Raymond Bradley (992) also highlights a series of important differences between Wittgenstein and Russell's proposals.Bradley emphatically asserts that: […] something is wrong with the commonly held view that Wittgenstein subscribed to Russell's Theory of Descriptions and took it as his model for propositional analysis.[… Wittgenstein] discards Russell's Theory of Descriptions and proposes a different sort of analysis, one which leads ultimately to propositions consisting of concatenations of simple names.(BRADLEY,992,p. ) araceli rosich soares Velloso volume 18 número 1 2014 Bradley's quote mentions the first important difference we are going to propose between Russell's Theory of Description and the Analysis suggested by Wittgenstein in the Tractatus: Wittgenstein's analysis leads to "propositions consisting of concatenations of simple names".So, instead of supposing that Wittgenstein's Great Analysis is modeled in Russell's "Theory of Descriptions" conception of elementary propositions, we will picture it as projecting a very different ending.This alternative proposal has also a different initial motivation, as we are going to argue, the full-blooded adoption of a more radical conception of propositional sense as given by its truth-conditions.
The sense of a proposition is its agreement and disagreement with possibilities of existence and non-existence of state of affairs.(TLP, 4.2) To understand a proposition means to know what is the case if it is true.(TLP, 4.024) Thus, at the end of the analytical process "propositions' sense" must be for Wittgenstein simply, and only, "situations which could make them true" and situations which could make them false.As we shall see in the next sections, this conception has serious and interesting consequences for the logical form of elementary propositions: they should contain only names and no predicates of any sort.So, to put in a nutshell, the main difference between Wittgenstein's Analysis and Russell's is that according to the first one we cannot have unsaturated elements in the elementary propositions, while according to the second, elementary propositions do contain those unsaturated elements, the predicates operating within the various definite descriptions.
Another difference, which is also emphasized by Bradley (992,p. ), is that for Wittgenstein we only have acquaintance with complexes and not with simple object, but for Russell quite the opposite takes place: we do not have acquaintance with complexes, but only with simple particulars.We are especially interested in this difference because those "complexes" are going to be the starting point of the "Great Analysis".

A. Why "The Great Analysis".
By calling the process of analysis presented by Wittgenstein in the Tractatus the "Great Analysis" we are trying to capture Wittgenstein's idea that "a proposition has one and only one Wittgenstein's unique "great analysis" volume 18 número 1 2014 complete analysis" (TLP,3.25).We are also implying that the process we are talking about is a very special kind of analytical process, for it is by far a much more complex and involved procedure, which establishes a correlation between language and reality in a different way from that of Russell's.
A.1 A more complex and involved procedure.
The "Great Analysis" has to go much further and deeper then the other two proposals.In fact, one might even say it really begins where those other processes of analysis are ending: with ordinary grammatically singular propositions.The "superficially singular propositions" we are talking about are ordinary propositions.They contain only singular terms plus their predicates (monadic or relational), with no quantificational modifiers or plural endings.In other words, plain sentences like "Mary is tall", "Mary loves John", "That is bad", "He likes ice-cream" or "John's father is tall".Our claim will be that Wittgenstein seemed to have strong qualms about considering these sentences, the ones with a subject-predicate structure -meaning a nominal argument and a predicative part -as basic propositions.He thinks that at the elementary level this two-component structure is inadequate in reflecting the right logical form of reality.We are going to argue that for him the process of analysis should go further and further until we find the hidden logical form of the genuine propositions.These propositions will have a much more complex logical multiplicity, diverging completely from the ordinary subject-predicate structure of the so-called "singular ones".
We think that the rejection of a subject-predicate structure at the elementary level comes directly from Wittgenstein's full-adoption of the construal of propositional sense as truth-conditions.Our claim is ultimately founded on Wittgenstein's metaphysical view of the world as something completely determined up to its finest details.
The world must be what it is, it must be definite.Or in other words, what vacillates is our determinations, not the world.(NB,7.6.5,p. 62) This precept leads him to the subsequent idea that the world is made of occurrences, states of affairs which "are the case", each of which is completely particular and determined in araceli rosich soares Velloso all its minute details.Thus he concludes that a completely analyzed proposition should be able to offer a picture of those particular events 2 .If we now return to the ordinary "singular propositions", we will realize that they are not themselves completely singular, for the situations that are described by them have a lot of "generality" still hidden in them. 3What we mean is that instead of a single truth-condition, they have a plurality of truth-conditions.It is in response to this "hidden generality" that, instead of ending the process of analysis by dividing the singular propositions into nominal parts and a predicative part, or in predicates, variables and quantifiers plus identity, like Russell, Wittgenstein suggests us to extend our procedure further.The process should endure until we find really elementary propositions, the ones in which all this "grammatical generality" has been made explicit.The new elementary propositions will have to be more complex than the unanalyzed ones, so that they can cope with the "manifold of spatial and temporal objects, colors and sounds of reality" (SRLF, p.3).The alternative logical structure of these new elementary propositions will be the one which in fact connects them, each one independently of all the others, to only one particular atomic state of affairs that could make it true.For Wittgenstein thus, the "Great Analysis" was a precondition to the central semantic relation of truth making.

A.2
The new frontier between sense and lack of sense.
The obstinate elongation of the ordinary logical process of analysis we are proposing here is a direct consequence of a very particular notion of sense, for not only will we consider as senseful only those sentences which have truth-conditions, but we will have to exclude those which could only have one of them: either always the "true" or always the "false", i.e., the necessary propositions.
In order for a proposition to be capable of being true it must also be capable of being false.(NB,5.6.5,p. 55) 2 We are taking about "events" as situations which could actually occur in the world, i.e., states of affairs.3 We are not going to deal here with Wittgenstein treatment of quantifiers and variables.Our concern is exclusively with the kind of generality "contained in general terms", i.e. predicative expressions, as well as in disguised or apparent singular terms.We will come back to the topic further on.
Wittgenstein's unique "great analysis" volume 18 número 1 2014 Here we see Wittgenstein asserting clearly the construal of propositional sense as truthconditions and its main consequence: the complete representability of sense in the obtainment, or not, of situations that would make the proposition true and also into the ones that would make it false.Once this precept is adopted, a lot of radical results have to be faced, as Wittgenstein realized.According to this radical construal, the old positivist purpose of getting rid of metaphysical propositions as senseless will have to be much more radicalized.We will not only have to throw away as senseless Metaphysical Propositions , but also other more "crucial" ones, for any proposition that could not be both: true and false as well, will not describe reality and so has to be considered without proper sense.In accordance with this radical criterion Wittgenstein was forced to reject logical propositions as senseless and mathematical propositions as pseudo-propositions. 4   This drastic application of the construal of propositional sense as truth-conditions has as a result the drawing of a new frontier between what belongs and what doesn't belong to language.Sense is the capacity to describe reality, or to establish what situations would make the sentence true and to exclude de situations that would make it false.Furthermore the concept of reality within the Tractatus involves the idea of contingency, i.e. reality is the result of drawing a frontier in the logical space between those states of affairs which happen to occur (the world) and those that happen not to occur (the rest of "reality").So, language will involve only those senseful propositions which describe possible states of affairs.Everything else we might want to convey through language will have to be somehow "shown in its logical form".

4
In TLP 4.46 and 4.46 Wittgenstein says that tautologies and contradictions lack sense, but are not nonsensical.But what about pseudo-propositions? Should they be classified as sinnloss or Unsinn?In TLP 4.272 he uses the expression "unsinnige Scheinsätze", nonsense pseudo-propositions, to characterize a propositions like "there are 2 objects such that …".This last aphorism suggests a reading of pseudo-propositions, including the mathematical ones, as nonsenses and not as simply lacking sense.One possible explanation is that according to the bipolarity principle tautologies and contradictions divide the logical space in only one single space and are a truth function of the elementary ones, whereas pseudo-propositions are those which seem to be describing a situation, but are incapable of segregating the logical space into situations that turn them true and situations that turn them false.Pseudo-propositions are not truth functions of elementary ones, either.So, although they seem to have sense, they do not indeed have it according to the bipolarity principle and so according to the construal of propositional sense we have adopted.That is why they would be Unsinn and not sinnlos.But the matter is still controversial.

The beginning
A. Where do we start?
Our starting point should be the ordinary singular propositions.They are bipartite, as we have already said, and their "nominal parts" -"grammatically singular terms" -usually refer to ordinary objects5 of our day-by-day life, say, bodies.Wittgenstein frequently calls those denoted entities "complexes" implicitly questioning their status as particulars.These "complexes" will constitute our main interest in this section.From the exegetical point of view, then, our investigation initiates in the preparatory Notebooks 1914-1916, since it is there that one can find the widest and most detailed discussion concerning this topic.
Let us then turn our attention to those ordinary objects, watches say, to use Wittgenstein's own example.In the following passage we find him wondering how much an expression like "this watch" is really capable of fixing its reference.After all, so he says, the watch in front of us can be pointed at by a gesture of ostension and named.
It is in fact clear that I can correlate a name with this watch just as it lies here ticking in front of me, and that this name will have reference outside any proposition in the very sense I have always given that word […].(NB 5.6.5,p. 60) Nevertheless, after conceding that a term like "this watch" is used in ordinary language as a name for an object, Wittgenstein reevaluates the question of how much an expression like "this watch" "[…] is really capable of performing its duty of determining its referent and thus fulfills all the conditions for being a name of a 'simple object'" (NB, 6.6.5, p. 60).His question in the subsequent quote enunciates the very point we want to explore: In order to know the syntactical treatment of a name, must I know the composition of its reference?If so, then the whole composition is already expressed [in a hidden way] even in the unanalyzed proposition.[my emphasis] (NB, 6.6.5, p.6) volume 18 número 1 2014 One of the difficulties Wittgenstein is facing here seems to be that he is already worried about the inevitable ambiguity present in the method of ostensive naming.The problem here is that if we want to establish the truth-conditions of any ordinary singular proposition, we will have first to establish the event, or state of affairs, it projects.We could use an ostention act to do that, but when we point to something what we get is at best some ill-defined extended space-temporal part of the world.Wittgenstein's concern seems to be that all the minute parts of what we are pointing at are already there and so in order to determine our very subject matter we would have to delimitate it in a thoroughly rigorous matter.In the next two passages Wittgenstein is again dealing with the question whether we can treat ordinary bodies as objects ("things") or as complexes: Can we regard a part of space as a thing?In a certain sense we obviously always do this when we talk of spatial things.(NB,9.5.5,p. 46) Let us assume that every spatial object consists of infinitely many points, […].(NB,7.6.5,p. 62) In that case, then, what we mean by "complex objects do not exist" is: It must be clear in the proposition how the object is composed, so far as it is possible for us to speak of its complexity at all.(NB,7.6.5,p. 63) In those passages, his option seems to be that of treating what we normally call a spatial object as a complex.He is also making an observation about the usage of the word "complex": it implies that the thing in question does have parts and so it doesn't make sense to talk about complexes, unless we can talk about their parts too.The next point in establishing the need for a complete "Great Analysis" can be read in the next quote.Here Wittgenstein is linking the idea that the world is made of elements with the idea that the world is determined.
[…] we feel that the WORLD must consist of elements.And it appears as if that were identical with the proposition that the world must be what it is, it must be definite.Or in other words, what vacillates is our determinations, not the world.[my emphasis] (NB, 7.6.5,p. 62) So, our language could vacillates and be imprecise, but the world cannot.And more: the elements Wittgenstein was talking about are not the complexes, for complexes do have constitut- Wittgenstein is clearly using the word "name" in two ways here.The first use is in unanalyzed propositions.In those propositions a "name" stands for a complex and in this case they will have to "disappear on further analysis".Hereafter we will use "ordinary names" to refer to this first usage.The second use is in the completely analyzed propositions, the elementary ones.For this second employment we will apply the expression "genuine names" meaning "the proper ones", the ones that refer to really simple objects.
After differentiating those two usages of the word "name", for reasons of clarity we will have to differentiate also and very sharply two distinct propositional levels: a top level of ordinary propositions, containing ordinary names and ordinary predicates, and a bottom level of elementary propositions that should contain only genuine names.The bottom level is composed only by elementary propositions and the important difference between these propositions and the ones of the other levels is that only the first ones have their sense completely determined.
[…] When a propositional element signifies a complex, this can be seen from indeterminateness in the propositions in which it occurs.In such cases we know that the proposition leaves something undetermined.[…] (TLP, 3.24) Wittgenstein's use of the expression "determined" and "undetermined" in this passage seems to indicate that those we called ordinary propositions have an undetermined and ambiguous sense.Ambiguous in the sense that it is not exactly established which situations would make them true, and which ones would make them false.So, to pass from the first level of propositions -the ordinary ones, with undetermined senses -to the second foundational one, we will have to analyze away terms like "this Watch".This will be done by replacing the proposition Wittgenstein's unique "great analysis" where it occurs by a disjunction of conjunctions of propositions that describe the sub-events that compose it instead: the occurrence of the current, of the hands, of the numbers, etc.If we go ahead with this replacing strategy, we will be clarifying the original sense.With this sort of "generality disclosure process" we will also be revealing the hidden logical form of the proposition, as Wittgenstein himself says referring to Russell's project: It is not humanly possible to gather immediately from it [Everyday language] what the logic of language is.Language disguises thought.(TLP, 4.022) […] It was Russell who performed the service of showing that the apparent logical form of a proposition need not be its real one.(TLP, 4.003) An important point though is worth emphasizing about this "disclosure strategy": the initial entity to be analyzed is a whole proposition and the next stage of analysis will still be composed by further propositions.Wittgenstein stresses this point in the Tractatus: Every statement about complexes can be resolved into a statement about their constituents and into the propositions that describes the complexes completely.[my emphasis] 6 (TLP, 2.020) According to this passage then our starting point is a proposition that describes the complex of events that constitute, say, the watch's lying there, ticking and the subsequent steps are replacements of this proposition by "statements about their constituents and into the propositions that describes the complex completely".Therefore, we have to substitute the initial proposition for other propositions that completely describe a cluster of particular occurrences -the occurrences of each watch's parts and their connections -which could incarnate one of the (several) states of affairs projected by the ordinary assertion.Thus the intermediate stages of our process will have to contain propositions which differ in meaning from the original, ordinary ones.They will have to involve different expressions and describe the original "watch event" in ever growing details, but always preserving its original sense.6 This passage is already in "Notes on Logic" (NB, Appendix I, p. 93).
araceli rosich soares Velloso volume 18 número 1 2014 The analysed proposition mentions more than the unanalyzed.Analysis makes the proposition more complicated than it was, but it cannot and must not make it more complicated than its meaning was from the first.When the proposition is just as complex as its reference, then it is completely analysed.But the reference of our propositions is not infinitely complicated.(NB,9.5.5,p. 46) As already said, if we go all the way "down" with this replacement process we will find a bottom level of unanalyzable propositions, the elementary ones, which contain only genuine names.Last but not least, it is this complete final description that would be for Wittgenstein the most important consequence of the construal of propositional sense as truth-conditions.

B. Why performing this unrelenting process.
At first sight it is not easy to see why we should ever have to postulate this laborious process of analysis, even if we conceive it as being possible only in principle.If all we want is to fix the sense of our propositions for the purpose of communication, all this discussion about "completeness of sense" seems entirely dispensable.Max Black, in his review of Griffin's book, complains along similar lines: What reason is there to think that what we mean by talking about a broom must really be something very complicated about material atoms?Very little, so far as I can see.(BLACK,966,p. 375) Even Wittgenstein himself appears to consider this objection now and then, when he emphasizes the complete logical correctness of day-to-day propositions: Obviously propositions are possible which contain no simple signs, i.e. no signs which have an immediate reference.And these are really propositions making sense, nor do the definitions of their component parts have to be attached to them.(NB,9.5.5,p. 46) In fact, all the propositions of our everyday language, just as they stand, are in perfect logical order.(TLP, 5.5563) Wittgenstein's unique "great analysis" volume 18 número 1 2014 What Wittgenstein seems to be saying here is that we don't need the "Great Analysis" to be able to understand others, or to attach some sense to our propositions, or even to establish their truth, say, for the purpose of a trial.We don't need such "Great Analysis" to communicate.
What do we need that analysis for, then?For Wittgenstein we do need it if we want to explain the relation between language and reality.An important aspect of this relation comes from the pictorial theory of language.According to it language is capable of representing reality only if both have the same logical multiplicity, i.e. if they are isomorphic.
The theory of logical portrayal by means of language says -quite generally: In order for it to be possible that a proposition should be true or false -agree with reality or not -for this to be possible something in the proposition must be identical with reality.[Cf.2.8.] (NB, 20.0.4,p. 5) The completely analyzed proposition must image its reference.(NB,25.0.4,p. 8) Wittgenstein is saying here that we need to complete the process of analysis in order to find propositions which "image their reference" completely.That is to say, we have to find propositions which are "identical with reality" in order to cope with the philosophical problem of how a proposition can perform its duty of having sense by being a picture of reality.But sense is equated with truth-conditions.Therefore to picture reality is to picture the proposition's truth-conditions in all their complexity.And it is this precept about sense which will explain in turn the propositions' capacity to be a picture of anything at all.In the following fragments we have support for this claim: the "Great Analysis" as a means for the complete determination of truth-conditions and so of sense.
A proposition must restrict reality to two alternatives: yes or no.In order to do that, it must describe reality completely.(TLP, 4.023 ) Does such a complete analysis exist?And if not: then what is the task of philosophy?!!? (NB,3.9.4,p.2) The answer proposed by Wittgenstein at the time of the Tractatus is clearly the following: yes, we need a complete analysis to solve the Philosophical problem of establishing how lan-araceli rosich soares Velloso guage can represent reality.However, propositions represent reality by displaying the situation that would make them true and the situations that would make them false.Furthermore reality has a much more complex structure than the simple subject-predicate one of our ordinary propositions.That being so we will have to have a method for searching the really meaningful elementary propositions, the ones which appear in the genuinely lower linguistic level.The lowest level of our language thus, the one which would "touch reality", should contain propositions of an entirely different logical form, a structure which would completely match the logical multiplicity of reality.The "Great Analysis" would thus offer us something crucial: a perfect match between our representations and reality so that each elementary proposition could be isomorphic to a single possible state of affairs.This state of affairs will be the proposition's possible truth maker and this minimal atomic state of affair would fix its sense -its (singular) truth-condition -once and for all.
Now, if we construe language pictorially in this way, then some precautions will have to be taken.As long as there remains some residual "hidden" generality in the apparently "grammatically" singular terms of our ordinary propositions, their senses (i.e., their truth-conditions) will have to be still considered undetermined 7 and to solve this indetermination we will have to analyze away its grammatical generality substituting them by really singular propositions, the elementary ones.(TLP,3.24)To recapitulate what we've seen so far, the "Great Analysis" is really a process of eliminating the indeterminateness of ordinary "singular terms", as well as of the "normal" general ones by making explicit its general content.The goal is to end up only with genuine names.The joint concatenation of these genuine names will constitute the elementary propositions and the resultant structure will picture the structure of their possible true makers.
The process of eliminating ambiguous terms from the ordinary propositions though would need some kind of strategy for its implementation.In the Tractatus however Wittgenstein was not completely clear about how this elimination would be effected.One of the difficulties is that it seems to involve the recourse to definitions taken as rules of translations, but Wittgenstein never explained in detail from where these definitions would come from.In the next sections we are going to suggest a method that might have been what Wittgenstein had in mind.To accomplish this task, we are going to trace a parallel between Wittgenstein's arguments and the arguments of another philosopher of the analytical tradition, one who proposes a similar account of language in terms of singularity, despite their different approaches.We are referring here to Quine in his article "Identity, Ostension and Hypostasis".

A parallel with Quine's nominalism
A. The example of the river Caÿster.
In his paper "Identity, Ostension and Hypostasis" Quine suggests that we consider names of ordinary objects, those that we use in our day-by-day life, as names with divided reference.For Quine, those "names" will refer to a vast physical area, but also to a vast extension of time.The example given by the American philosopher is the river Caÿster in Lydia around 400 B.C. Quine suggests that we could consider this river as an enormous "scattered spatial-temporal thing" (996a, p. 73) that covers a vast area of land and a still more extended length of time: So why not view 'red' quite on a par with 'Caÿster', as naming a single concrete object extended in space and time?(QUINE, 996a, p. 69).
Quine suggests then that we could move on to a more detailed description of the river instead of this all-inclusive one.For doing this we would have to divide up the time-extended river "thing" into parts, so that each part would be considered as a "space-time slice of the molecules of water that compose the river along all its existence".Instead of saying thus, "This is the river Caÿster", we would say, "This is a momentary stage of the River Caÿster, in Lydia, around 400 B.C" and "That is another momentary stage of the Caÿster two days later".For him the treatment of all those slices as one and the same river is only possible if we look at them all together as a single "river process".
The same thing we've done with the river could be done with other physical bodies.Let us take individuals human beings, for example.We can always think of them as extended processes through time and space.To accomplish the goal of analyzing a physical body in this way araceli rosich soares Velloso we only need to add to our language a new predicate: "[…] is a space-time slice of so-and-so", or in Quine's nominalist approach, a name "The b space-time slice of so-and-so".In this new enlarged language, "The River Caÿster" will be a name, substitutable by a set of others more specific expressions, "The n space-time slice of the Caÿster", for n varying between and a million, for instance.
Let us now see how this enlargement of our language by means of a translational procedure could be viewed as a way of analyzing away our ordinary "apparent singular terms", the ordinary names for complexes, in the way proposed by Wittgenstein in the Tractatus.

B. Definitions as translational procedures.
Now it is time to fulfill our promise of explaining what exactly are the definitions needed in Wittgenstein's "Great Analysis".We will propose approximating Quine's translation technique to the analytical process suggested by Wittgenstein.What the Austrian philosopher takes to be an elucidation of the structure of the original "complex" is also a translation of the old unanalyzed proposition by means of definitions to a conjunction of new propositions.That is to say: we are going to think about the tractarian definitions of ordinary names as rules of translation.The following quote from the Tractatus seems to make a very similar point: Definitions are rules for translating from one language into another.Any correct sign-language must be translatable into any other in accordance with such rules: it is this that they all have in common.(TLP,3.343)As we observed earlier, ordinary names denote complexes and a complex is an event, an extended process that can occur in the world.We also said that, if they are complexes, they should be made of smaller parts.The conclusion is that, if they are possible states of affairs, or events, their parts will be sub-events or states of affairs of a shorter space-temporal dimension.Now according to Quine, new names enlarge our language and allow us to give more detailed explanations of the real nature of things.Of course, by "explanation" here we understand "propositional explanations".Combining thus their suggestions into one single approach, we can think of Wittgenstein's definitions as explanations that gradually specify shorter and shorter events.
Wittgenstein's unique "great analysis" volume 18 número 1 2014 All this talk about definitions seems to be slightly misplaced in face of our initial proposal.We've suggested that the important point in Wittgenstein's construal of "propositional sense" is that at the elementary level the propositions' sense should not depend on anything else except their truth-conditions.We kept nevertheless talking about definitions as a way to explain the sense of unanalyzed propositions.Our answer to this apparent inconsistency is that the appeal to definitions is justified only because they are just dispensable links, bridges that are there merely to bring us down to a more elementary level where we are not going to need them anymore.The image Wittgenstein seems to have in mind is that of a descending chain of definitions bringing us to the level of elementary propositions (TLP, 3.26).
If it's true that every defined sign signifies via definitions then presumably the chain of definitions must sometime have an end.(NB,9.5.5,p. 46) According to this passage, definitions will be used just to provide an elucidation of what the "sense" of these unanalyzed propositions should imply in term of truth-conditions at the lower levels.Nonetheless, sooner or later we will find the bottom strata where they are going to be sent away.At that point, the truth-conditions, and thus their sense, has to be completely explained independently of any definition.That is why we have called this kind of explanation an "Atomistic Explanation", for it is an attempt to find the most singular possible description of the world, one which is completely independent of anything else except possible states of affairs.Wittgenstein is apparently imagining what we called "Atomistic Explanation" as the termination of a process, a set of elementary propositions, which describe each minimal possible event or state of affairs.
Once having settled the connection between Quine's nominalist approach and Wittgenstein's definitions, is now time to move on to the most difficult part of our investigation: analyze and cope with the consequences of adopting such radical approach.Thus in the next section we will basically deal with the question of how we should construe propositional sense exclusively as truth-conditions and the implications involved in that.
araceli rosich soares Velloso Tractatus's second proposition states in a nutshell the lemma that brings the two philosophers together: ".The world is the totality of facts, not of things".Albeit underestimated, two ideas are very important in this famous quotation.The first one is that facts and their linguistic correlates, the propositions, have to be considered as the minimal units of what can be talked about.The second idea, a consequence of the first, is that one cannot talk about an isolated object, because "the world is not made of things", but only of facts.
Let us explain this point better.According to our interpretation, to emphasize "." is to adopt a construal of propositional sense as truth-conditions, for facts are precisely the truthmakers of propositions.In accordance with this approach the minimal unit of sense must be a proposition.But a second important point hidden behind this idea of a "minimal propositional unit of sense" is the strong appeal made by Wittgenstein to the context principle.
An expression has meaning only in a proposition.(TLP,3.34)Let us first analyze Wittgenstein's employment of the word "expression" [Ausdruck] in this passage.According to his explanation, this word means any "part of a proposition that characterizes its sense, everything essential to their sense that propositions can have in common with one another."(TLP, 3.3).What we have here is an observation about any significant propositional part, including the ordinary names we've been talking about.For the moment we just want to make an initial preliminary point about the source of this first claim regarding the meaning of expressions, more specifically, regarding ordinary names.Our suggestion is that this claim was inspired by a descriptivist conception of reference present in the work of both, Frege, and Russell as well.It means that we cannot refer to "things" outside a propositional context.For Frege: When we have thus admitted objects without restriction as arguments and values of functions, the question arises what it is that we are here calling an object.I regard a regular definition as impossible, since we have here something too simple to admit of logical analy-Wittgenstein's unique "great analysis" volume 18 número 1 2014 sis.
[…] Here I can only say briefly: an object is anything that is not a function, so that an expression for it does not contain any empty place.(FC,p. 32) According to Frege's approach, then, objects don't have any other identity criteria except being whatever could satisfy a function.For him there is no way one could ever talk about objects in abstract from the functions that they could satisfy, i.e. their properties.The central point is that it is only through properties that the intended referent can be identified.Furthermore the reverse is also true: the criteria for identifying an object is introducing it via definitions, i.e. giving their essential and characteristic properties.This kind of mediated naming relation could be viewed alternatively not as a singular term naming, but as a disguised proposition describing a possible state of affairs as well.This is Russell approach in "On Denoting" (956).For him, when we apply an expression like "this watch", we would be really saying: "There is something that exist, is unique, and has the property of being a watch".Following Russell, the only way to name an ordinary object, a complex, is to make it the value of a variable and pinpoint it through a description of the properties which that and only that object could have, i.e. the properties that are capable of identifying it.
Let us compare now these two views.One common aspect stands out clearly: the descriptivist approach to the ordinary naming relation.On this approach, ordinary objects are only the support of properties and are given by them.Furthermore, since it is only in the context of propositions that we can attribute properties to them, our minimal semantic unit must be a proposition, taken together of course with its respective truth-condition and the definitions of its components expressions.
One point is still untouched though and demanding clarification in this whole discourse about a descriptive approach.In this section we had been considering the minimal unit of sense as propositions, but what we have at the end of the process of analysis are just the elementary ones.Now, if definitions were whatever establishes the sense of our ordinary names at the top level, and we don't have them anymore at the bottom level, we are left with no alternative to endow sub-propositional expressions with a meaning but through a direct designative relation to simple objects.We are then left with the problem of how to apply the method of the context principle at this level without appealing to definitions, i.e., without giving the objects by their properties.
araceli rosich soares Velloso volume 18 número 1 2014 A direct answer to this question is essential to our argumentation at this point.Let us then try to anticipate something of our discussion, which will take place in last two sections of our paper.If one accepts our construal of the two propositional levels, there are in the Tractatus two aphorisms that, taken together, seem to offer an alternative formulation of the context principle that could make sense also at the bottom level.In (TLP, 4.23) Wittgenstein says, "It is only in the nexus of an elementary proposition that a name occurs in a proposition.".He says also, "only propositions have sense; only in the nexus of a proposition does a name have meaning [Bedeutung]."(TLP, 3.3) We want to emphasize two features of the first aphorism: the explicit reference to "elementary propositions" and the substitution of "an expression has meaning" in: An expression has meaning only in a proposition.(TLP,3.34)by "a name occurs".Right, in 3.3 we find the words "have meaning" again.But, if we contrast the attribution of meaning to names with the attribution of sense that can be done only to propositions, we could conclude that names cannot ever have that second "attribute", "sense".Hence, according to our suggestion, the main difference between this and the first traditional formulation of the context principle is that it is only in the context of an elementary proposition and in the interconnection with other names that a genuine name has reference.
We'll return to that point later.Here we only want to stress the idea that the behavior of genuine names at the bottom level is completely different from that of ordinary names.The main important result of our arguments in this section is that the connection between the "construal of propositional sense as truth-conditions" and a "descriptivist conception of reference" works only at the top level, where one can still employ definitions.As we conclude in section 4.b, definitions are only dispensable links, bridges which are there merely to bring us down to the bottom level where we wouldn't need and could not even employ them anymore.So names don't describe, they only refer.
In view of this radical scenario, we now have to analyze the threat of failing to get completely rid of definitions, including among them any kind of term which involves an implicit mention to any other term.For example, a term like "is red" implies "is a color and is applicable to concrete things", "is a dog" implies "is an animal, is not a bird, etc." and so on.Our claim Wittgenstein's unique "great analysis" volume 18 número 1 2014 is that any kind of grammatical generality brings with it a serious risk: the risk of loosing the logical independence of elementary propositions and the following failure of our initial claim that propositional sense should be construed at this bottom level exclusively as a singular truthcondition8 for each disjunctional compound.

B. The need for independence of elementary propositions.
Let us now make a brief remark about the construal of propositional sense as truth-conditions at the bottom level of elementary propositions.We have concluded that at this level we cannot employ any definitional device to understand sense, but just truth-conditions.Therefore it seems that at that level only the atomic state of affairs figured by its isomorphic elementary proposition is relevant for the complete determination of the propositional sense.Thus, if on the top level we have the meaning of sub-propositional expressions fixed by definitions, at the bottom level we cannot have symbols that signify via other signs.The genuine names will be "primitives", for they cannot be "anatomized" as Wittgenstein remarks (TLP, 3.26), and will have to depend exclusively on their denotation for their significance.
As we've concluded earlier, even singular terms could establish inferential connections via definitions, because they are really general "names" disguised as singular, as the old ordinary names were for Wittgenstein and Quine.Therefore we are forced to analyze away all ordinary names together with other general terms in order to get rid of the definitional connection and so satisfy both the independence principle and the goal of having only one truth-condition for each elementary proposition.
The logical independence of elementary propositions is a very important principle in the Tractatus, although still a quite enigmatic one.A lot of questions crop up, but the most important one is the following: is it at all possible to have a complete independence of sense at this level of language?Or, equivalently: is it possible at all to define sense exclusively as truth-conditions?It is now time to explain in detail how this "process of analysis" will be carried out and this is where Quine's proposal sets in.We will use the translational method proposed by him to get rid of those ambiguous terms, the outright general ones and the "deceptive singulars" as well.This is the process that we've named the "Great Analysis".Our analysis will begin with a pretense "singular proposition" and will finish when we reach the elementary ones.
Let us consider a fictional example for illustration: the proposition "John is sitting down".As a general description of all the possible events in John's life this proposition is still ambiguous.It could be the description of several "John-is-sitting-down" events at different time and places.So, before beginning our analysis we will have to incorporate these required specifications.After doing that, the next step will be to replace the proposition describing the event of "John sitting down at his favorite chair in 9 of October 204 at 0:00 am" by a disjunction of conjunctions, each conjunction describing a position that John could have adopted in his act of sitting down.
In order to carry out our project of narrowing down time and place specifications, we will have to separately consider each one of the conjunctions.We have to consider the description of the possible event of John being sited in each one of the multiple possible determined positions, for example.Then we will have to treat this description as a collection of more detailed descriptions of smaller sub-events, which involves, each of them, an extended area of spacetime.So, according to our definition of a "person sitting down", we will have first to describe, for each possible sub-event, say, the location of his head, the way his legs are crossed, his arms rested in the arm of the chair and so on.Those more circumscribed events could in turn be further described in terms of evermore-specific predicates, one for each of their compositional sub-sub-events, all the way down to, say, the collection of atoms that could constitute them all.The guiding idea of this process is to find the bricks that would compose "the sitting down of John" in a specific time and place for each alternative position.After doing that we will be able to recompose in a completely determined way the sense of the proposition describing the initial event of "John sitting down at a specific time and place".
Wittgenstein's unique "great analysis" Thus, if we carry on this process for each of the conjunctions of our initial long disjunction, we will have at the end all the possible elementary propositions, which represent each one of the states of affairs that could have entered in this construction.In the next section we will present a quite similar construal of that analytical process proposed by James Griffin, although with a somewhat divergent concluding stage.

B. James Griffin position about the method.
In his book Wittgenstein's Logical Atomism, James Griffin proposes a new construal of Wittgenstein's process of analysis in the Tractatus.Griffin's interpretation is based on his lectures on Heinrich Hertz's book "The Principles of Mechanics presented in a new form".He also mentions as inspiration for his construal a seminar he attended conducted by Pears and Mc-Guiness (GRIFFIN, 964, p. 34).Griffin begins his construal by bluntly stating that he believes Hertz's influence on Wittgenstein's Tractatus was even stronger than Frege's or Russell's: The important influence on the form, I think, was Hertz's.In The Principles of Mechanics Hertz gives the characteristics which any language for the description of the world (from the mechanical point of view) must possess.Wittgenstein models so much of the Tractatus on this work that there is point to thinking of the Tractatus as The Principles of all Natural Sciences.(GRIFFIN,964,p. 5) In this quote Griffin appeals to Wittgenstein's proposed previous title for the Tractatus as an indication of the importance Hertz's work had for Wittgenstein's book.In Griffin's view, just as in ours, the analysis has to go further then the level of ordinary subject-predicate singular propositions.We have to maintain it until we find "genuine elementary propositions".He also emphasizes a very important point quite congenial to our own interpretation: their logical form cannot be the one of subject-predicate.
The present account of analysis shows that objects are always particulars, that names can only be names of particulars, and that subject-predicate propositions cannot be elementary.(GRIFFIN,964,p. 57) araceli rosich soares Velloso volume 18 número 1 2014 As we have discussed earlier the reason is that they cannot contain, neither general terms, nor merely apparently "singular" ones.Instead they should be exclusively composed by genuine names.
His new construal of the "Great Analysis" also criticizes interpretations which employ Russell's definite descriptions as a model for the whole enterprise (GRIFFIN,964,.Likewise he believes, as we do, that analysis has to be carried through by a substitution procedure, which, via definitions of the terms involved, disclosed the real form of the propositions, analyzing thus away all the hidden generality.As support for his new conception of analysis he cites a piece of exegesis which Wittgenstein himself wrote in the Philosophical Investigations: When I say: 'My broom is in the corner', -is this really a statement about the broomstick and the brush?Well, it could at any rate be replaced by a statement giving the position of the stick and the position of the brush.And this statement is surely a further analysed form of the first one.-But why do I call it 'further analysed'?-Well if the broom is there, that surely means that the stick and the brush must be there, and in a particular relation to one another; and this was as it were hidden in the sense of the first sentence, and is expressed in the analysed sentence.[My emphasis] (IF, § 60) According to Griffin Wittgenstein in this quote is suggesting that what he meant in the Tractatus by "an analysis of the original proposition" is a replacement of it by other statements describing the sub-events in question and the fact that they are connected.Griffin then goes on saying, just as we did above, that the properties we normally attribute to ordinary objects are going to be analyzed away and replaced by structural ones: the configurations of the molecules that compose the event analyzed in space and time.
A consequence of this interpretation of analysis is that we are committed to saying that all names are names of particulars.The expression 'the broom' becomes 'the brush in a certain relation to the stick', and analysis will go on in this way until eventually we speak only of particulars and their relations.Where are any general terms now?Evidently all get analysed away.But note that on the fully analysed level instead of speaking of 'the broom' what we now speak of are 'these particulars in this configuration and these other particulars in that configuration, &c'.So, it looks as if configurations take over the role of general terms.[My emphasis] (GRIFFIN, 964, p.52) Wittgenstein's unique "great analysis" volume 18 número 1 2014 According to Griffin then, Wittgenstein is proposing a thoroughly physicalist and atomistic description of reality so that our ordinary facts will be reduced to aggregates of small objects and their configuration in space and time.Unlike Griffin, we would like to avoid taking sides on the physicalist versus phenomenalist readings of the Tractatus.Our discussion concerns only the analysis of events into sub-events without specifying their ultimate ontological status.Griffin also adopts the desiderata that all the descriptive properties should be analyzed away and in their turn we will have only the internal configuration of elementary propositions.This time we agree with him completely.
Wittgenstein's thesis is a universal one: all properties disappear on analysis.Perhaps Wittgenstein intends his remark on the incompatibility of colors to apply equally widely.That is, it may be part of his argument that for any property there is at least one other to which it stands in internal relation, which pair of properties logically excludes one another.964,p. 79) As we said before, according to our proposal any kind of grammatical generality brings with it a serious risk: the risk of loosing the logical independence of elementary propositions and the following failure of our initial claim that propositional sense should be construed at this bottom level exclusively as a singular truth-condition 9 for each disjunctional compound.
The thesis that all properties should be eliminated from elementary propositions is a very bold one though and it goes squarely against a more traditional way of interpreting the Tractatus.According to this thesis at the elementary level we should not have general terms, but only genuine names.The strangeness of this thesis arises from thinking about a proposition as projecting an event without attributing a property to anything.Nevertheless we agree with Griffin that this claim is essential in order to get a better understanding of what goes on at the bottom level.We agree also with the reason he gave: the threat of losing the logical independence of elementary propositions and the insight that the color's problem is not the only menace, since other properties could cause the same difficulties.9 We have used the singular form "truth-condition" on purpose to emphasize the singularity of the truth maker, i.e. the state of affair we are searching for.
araceli rosich soares Velloso volume 18 número 1 2014 All of Griffin's proposals were severely criticized by Max Black's review of Griffin's book.Black sees Griffin's interpretation as "felicitous", "illuminating", but "incoherent", and probably as not being what Wittgenstein had in mind (BLACK,966,p. 376).He criticizes mainly two aspects: Griffin's new conception of analysis and Griffin's proposal regarding the logical form of elementary propositions.Concerning analysis, Black complains that: This conception of analysis, new or not, runs into insuperable difficulties when coupled with the view (misguided, I think) that the final analysis must contain only names of particulars.(BLACK,966,p. 375) Black protests that we cannot "understand the use of definite descriptions, as we do, when unacquainted with their references".In this quote he is clearly embracing the more common view, which is that Wittgenstein's analysis should be modeled on Russell's Theory of description and the descriptivist construal it recommends.But we think this is a wrong way to go about it.Here there is a very cogent argument against Black's complain made, not by Griffin, but by Raymond Bradley: Wittgenstein and Russell disagree totally, about which objects are known by acquaintance and which are known by description.Russell holds that complex objects such as Socrates are logical constructs out of the sense-data of which alone we have knowledge by acquaintance and hence concludes that these complex objects can be known, if at all, only by description.Wittgenstein, however, insists that it is the other way around, complex objects being the things with which we are acquainted, simple ones those which we can know only by description.[...] (BRADLEY, 992, p. 4) The point of Bradley's argument is that nothing forces us to presuppose that for Wittgenstein we do have acquaintance with simple objects.Moreover it is very reasonable to suppose instead that for the Austrian philosopher we only have acquaintance with complexes.This fits very well with both Griffin's and our approach.
Another criticism given by Black in his review of Griffin's book was about the final structure of elementary propositions.He argues that: Wittgenstein's unique "great analysis" volume 18 número 1 2014 Opportunistic nominalism of this sort turns out to be self-defeating anyhow.For the target of analysis, as Griffin conceives it, will be an elementary statement whose sense is that the named objects are configured in a certain way.Unfortunately, Griffin does not tell us how he understands 'configured' or 'configuration ' […] He also holds that different facts can be configured in the same way (p.52) and that configurations sometimes vary from one fact to another.So a mode of arrangement is plainly general, and not unique; hence full symbolization of an elementary proposition should, on Griffin's principles, call for symbolizing the definite configuration in question.(BLACK,966,p. 375) This time we think that Blacks arguments against Griffin's interpretation are more substantial.Instead of simply reacting to Griffin's non-orthodoxical way of reading the Tractatus, he is demanding for more details about just how to understand the notion of "configuration".He also wants to know how this so-called "configurations" should be incorporated into each particular elementary proposition.Be that as it may the suggestion of seeing the structure of elementary propositions as portraying its logical form is a very central in the Tractatus.
Up to this point we have only discussed the analytical process that happens at the top level.As we've anticipated above we think that things go differently at the sub-propositional level.

A. Simple objects come into the scene.
As we've seen, the accomplishment of this exhaustive process of analysis involves viewing "propositional sense" as "translatability into the basic elementary level".This means that as long as we encounter (disguised) general terms as propositional components we still need to go further down with our process of disclosing the sense via definitions of these "ordinary names".This process should go on until we find really elementary propositions that could be matched by atomic states of affairs, the particular possible events.
The point we want to discuss now is the internal structure of those elementary propositions.They are still complexes and their composition is a very strange one: a structured concatenation of genuine names.An elementary proposition consists of names.It is a nexus, a concatenation, of names.(TLP,4.22)When the sense of the proposition is completely expressed in the proposition itself, the proposition is always divided into its simple components -no further division is possible and an apparent one is superfluous -and these are objects in the original sense.(NB, 7.6.5,p. 63) The trivial fact that a completely analyzed proposition contains just as many names as there are things contained in its reference; […].(NB, 2.0.4,p. ) The first thing to be observed is that completeness of sense is identified in the two last passages with completion of analysis and also with the obtainment of the right multiplicity at the elementary level.The other striking point is that the impossibility of going further is achieved by the complete replacement of all the internal propositional positions by names, by "genuine names".Thus, the reason why we cannot go any further with this process, as we already anticipated in section 5.a, is given by the fact that these sub-propositional elements are themselves senseless (they only signify) and so are irreplaceable via definitions by others more primitive.
Names cannot be anatomized by means of definitions.(Nor can any sign that has a meaning independently and on its own.)(TLP,3.26)So, this whole translational process, which involves attributing descriptive sense to ordinary names via definitions, works only until we reach the bottom level.At that level we have merely genuine names and, according to Wittgenstein, their very nature is that they cannot be further analyzed.In the passage above we have to read "names" as genuine names, of course, for they are the only ones which are but label devices.Latter, in the Philosophical Investigations, Wittgenstein state this point very clearly: What lies behind the idea that names really signify simples?[…] there is no definition of the primary elements out of which everything is composed; for everything that exists in its own right can only be named, no other determination is possible, […] it is impossible to give an account of any primary element; for it, nothing is possible but the bare name; its name is all it has.[…] Both Russell's 'individuals' and my 'objects' (Tractatus Logico-Philosophicus) were such primary elements.(IF,§ 46) Wittgenstein's unique "great analysis" This is a second very important conclusion that Wittgenstein extracts from his construal of propositional sense as truth-conditions, for not only () just propositions which describe states of affairs can express a sense, but also (2) genuine names simply cannot have them.Now, if elementary propositions are merely a group of names, the question that imposes itself on us is: how could they ever have sense?
Our answer could be found in the idea of considering a proposition as an articulated sequence of names.
A proposition is not a blend of words.-(Just as a theme in music is not a blend of notes.) A proposition is articulate.(TLP,3.4)This answer is totally compatible with Griffin's proposal of a configuration that will take the place of our older predicates (GRIFFIN, 964, p. 52).
The "Great analysis" has an end at the elementary level with structured elementary propositions thus.Nevertheless we are still thinking of circumscribing smaller sub-propositional parts, the genuine names.This is a strange extension of our analytical process though, for we seem to be postulating two minimal units: the minimal unit of propositional sense and the minimal unit of linguistic complexity.

B. The minimal unit of linguistic complexity.
After reaching the level of elementary propositions we still have problems.In section 5-a we have concluded that from the point of view of sense the end of the analytical process should be that of the elementary propositions.We've concluded also that the process using the translational method would go on only up to them and their ontological counterpart, the atomic states of affairs.All through this large "first stage of analysis" we could still talk about the sense of any proposition coming from the composition of the sense of its parts.However, after reaching the end of propositional analysis and finding the units of sense, the elementary propositions, we still need to go ahead, because an altogether new notion of unit comes into play: the genuine names.Except that this time we have a purely referential unit.Only then we will finally find the araceli rosich soares Velloso volume 18 número 1 2014 ultimate terminus of complexity: a sub-propositional level structured by a sequence of genuine names.This other "kind" of simplicity is determined by a property of genuine names: they cannot be further analyzed, for they do not consist in a combination of anything and also do not describe anything either, they are mere label devices.Nevertheless they are still required and Wittgenstein himself says that the completeness of sense demand them.
The requirement that simple signs be possible is the requirement that sense be determinate.
(TLP, 3.23) However now we face a really big difficulty: if they are not a minimal unit of sense, but only a minimal referential unit, we have to explain what kind of entity they denote.Furthermore we also have to explain what it means for them to be combinable, and finally, what is their role in elementary propositions.Let us then explore a little more this ultimate end point in an attempt to give an interpretation which copes better with all these features.
An initial crucial issue is that genuine names have reference, but not sense.So, they cannot perform their duty of referring to their denotation, the simple objects, by means of descriptions.Wittgenstein himself clarify this point stating that this is a relation without intermediaries, a kind of arbitrary relation akin to what is contemporarily known as direct reference.
[…] This part of the representation (the assignment of names) must take place by means of arbitrary stipulations.Every proposition must accordingly contain features with arbitrarily determined references.(NB,22.0.4,p. 7) That arbitrary correlation of sign and thing signified which is a condition of the possibility of the propositions, […] (NB, 29.2.4,p. 25) [TLP,5.523.]The directness is implied by the fact that genuine names do their job without the use of properties to "harpoon their denotations".This proposed "direct" naming relation associates an empty sign, a proxy, to each denoted simple object.According to Wittgenstein it is a one to one correlation.
In the proposition the name goes proxy for the object.(NB,29.2.4,p. 37) [TLP,3.22.]One name stands for one thing, another for another thing, […] (TLP, 4.03) Wittgenstein's unique "great analysis" volume 18 número 1 2014 Now, keeping in mind that genuine names are incapable of describing, we are forced to conclude that their denotata are not describable, i.e., they are not the bearer of properties.If they are not describable, they cannot be a minimal occurrence in the logical space, for these minimal occurrences must be something which could be describable by a proposition, i.e., a state of affairs.This conclusion gives rise to another difficult question though: what else could they be?How could these denotata be something and at the same time not belong to the logical space?The next step will be an attempt to clarify the role of genuine names in elementary propositions and the nature of their denotata, the simple objects.

C. The role and nature of simple objects.
Before trying to answer the questions above, a preliminary consideration has to be made.Simple objects seem to have been postulated by Wittgenstein simply to solve two related problems: the isomorphism between propositions and reality and the problem of false propositions.Wittgenstein himself explains the first role.He says that it is through genuine names that a proposition could be isomorphic to a state of affairs.They are like "feelers that touch reality" via its simple objects and so mirror the structure of the possible states of affairs into that of the proposition.
That is how a picture is attached to reality; it reaches right out to it.(TLP, 2.5) These correlations are, as it were, the feelers of the picture's elements, with which the picture touches reality.(TLP,2.55)The second problem is the famous platonic problem about the sense of false propositions.The central problem of Plato in his dialog Sophist is to attribute sense to false propositions.If sense comes from the relation between propositions and their truth makers, the facts, then in case that particular fact didn't obtain, the proposition will simply lose its sense, together with its truth-value.
To solve this ancient problem Plato proposed as we know that every proposition must be a complex."Being false" would then simply be a combination of a nominal-term and a predicate-term which fails to correspond to a situation in the world.The failure is due to the fact that araceli rosich soares Velloso volume 18 número 1 2014 in the world the corresponding "things" are not combined in the way presented by our proposition.Or, in a tractarian terminology, a false proposition is just a possible combination of genuine names which is isomorphic to a state of affairs that doesn't obtain 0 .The possibility of non-actual states of affairs, though, could not be based on something contingent, thus Wittgenstein had to postulate atemporal simple objects to be the substance of this and other possible worlds as well (TLP, 2.02-2.023).
Our conclusion in this section is that simple objects can be seen as an answer to the old "problem of non-being".Possible but not-actual states of affairs would be then a kind of "shadow of reality", to use Wittgenstein's later terminology.They appear in the logical space, for they are combinations that contingently happen not to be the case.It is the combinatorial potentiality of simple objects that projects those shadows and so provides sense to false propositions.Our initial answer to the question about the nature of simple objects is then to consider them merely as combinatorial possibilities.In order to do this though let us first investigate Griffin's suggestion and then propose another solution that better fits our view of the "Great Analysis" and its distinctive two levels.

D. The nature of Simple Objects according to Griffin.
In his book Griffin proposes to construe "simple objects", following Hertz, as "material points": Then, Wittgenstein says that the world is the sum of facts, not of objects.If material points are objects, then it is true that one can characterize the world only by saying that objects have some configuration.The world consists of watch-wheels, watches, books &c.These are not objects, but objects plus configuration.In other words, they are facts: the facts that such and such objects stand in such and such relations to one another.[My emphasis] ( GRIFFIN,964,p. 32) We agree in part with what Griffin says in this passage.The part which we agree with is 0 This short explanation is not enough to clarify how falsity works at the level of elementary propositions.This topic deserves a separate discussion which will be accomplished in a future paper.
Wittgenstein's unique "great analysis" volume 18 número 1 2014 that ordinary objects are not really objects at all, but complex situations that could occur in the world.We also agree that "there are no objects, but [only] objects plus configuration", although we do distinguish simple objects from ordinary ones.And we have already said that we assume, just like Griffin, that one of Wittgenstein's main problems was with the subject-predicate form at the elementary level."[…] subject-predicate propositions cannot be elementary."(GRIFFIN,964,p. 57).But there are also some aspects of Griffin's position with which we sharply disagree as well.
Our first objection is that Griffin should have gone a step further concerning the importance attributed to the statement that the "world is the sum of facts, not of things" adding to it another tractarian idea: Just as we are quite unable to imagine spatial objects outside space or temporal objects outside time, so too there is no object that we can imagine excluded from the possibility of combining with others.The other point about Griffin's position that differs from our own is a consequence of the first.Right from the beginning our argument was that simple objects couldn't be composed: "Objects make up the substance of the world.That is why they cannot be composite."(TLP, 2.02) Consequently, if the world is made of facts, not of things, we could not correlate genuine names with material points and neither with material points plus some configuration.The definition provided by Hertz himself is very clear about this topic: material points are extended in space, have mass and, most important of all, are composed of other things, the material particles.
A finite or infinitely small mass, conceived as being contained in an infinitely small space, is called a material point.

A material point therefore consists of any number of material particles connected with each other.
It is always permissible to regard a system of material points as being composed of an infinite number of material particles.(HERTZ,956,p. 46) araceli rosich soares Velloso We think that the main difficulty for the interpreters here is to give the appropriate relevance to the fact that Wittgenstein's simple objects are just not events!Being only combinatorial possibilities, they cannot have the kind of "autonomous status" which a possible event is supposed to have.More than that, although simple objects cannot be described by propositions, the occurrence of material points can, because they occur in the physical world.According to Hertz they even have properties, such as "mass": The sum of the masses of the separate points is, by §4, the mass of the system.(HERTZ,956, Instead of construing material points as simple objects, Griffin could have construe these "points" as atomic states of affairs.An atomic state of affairs is what could correspond to a minimal possible event, i.e., the possible occurrence of a particular atom in the physical space-time (or else a color patch in the phenomenological space according to a phenomenological interpretation).But then, what could these strange entities be, these "simple objects"?
E. An Alternative Proposal.
In this section we are going to suggest another line of interpretation of what is going on at the bottom of this whole process of analysis.This alternative interpretation was suggested by Hyder in his book The Mechanics of Meaning and it is much more in tune with Wittgenstein's remarks on the atemporality, inalterability and subsistence of simple objects in every "imagined" world (TLP, 2.02-22).
Hyder's suggestion consists in treating genuine names as proper names of "the coordinates of the manifold in question" (HYDER, 2002, p. 2).These coordinate's names will determine a system of spatial and temporal possibilities.Hyder even adds to these spatial and temporal coordinates further "sensorial spaces", one for each sense faculty, and claims that they should likewise be considered as "a priori manifolds" (HYDER, 2002, p. 22).Wittgenstein himself ends up making a very similar proposal in his later article "Some remarks on Logical Forms".In this paper he talked about replacing the apparently propositional logical-form of subject-predicate by another, more complex one.If, now, we try to get an actual analysis, we find logical forms which have little similarity with the norms of ordinary language [subject-predicate propositions].We meet with the forms of space and time with the whole manifold of spatial and temporal objects, as colours, sounds, etc., etc., with their gradation, continuous transitions, and combinations in various propositions, all of which cannot seize by our ordinary means of expression.(SRLF,p. 3) In this quote Wittgenstein mentions the notion of "manifold" to account for the logical form of completely analyzed propositions: time, space, but also sound and color manifolds.And in the very next passage he proposed the idea of representing these manifolds by systems of coordinate-values and a chosen unit.
We can describe the shape and position of every patch of colour in our visual field by means of statements of numbers which have their significance relative to the system of co-ordinates and the unit chosen.(SRLF, p. 3) The notion of "manifold" was considered by Hyder (2002, p. 23-24 and chapter 6) as central to understand the nature of the tractarian notion of "Logical Space".According to him, the idea of seeing the Logical Space as a space of "manifold of possible occurrences" comes from the influence of Hertz' works upon Wittgenstein, and Hertz by his turn was Helmholtz's student.The proposal is that this notion of "manifold" comes indirectly from Helmholtz' works on perception.Hyder's main idea is that, taken together, all these manifolds, including colors, sounds, etc., will constitute the logical form of all possible representation: "So logic is, in a sense, a general theory of manifolds."(HYDER, 2002, p. 24).
The expansion proposed by Hyder involves as we saw treating the tractarian names as proper names of coordinates.In order to do that we only have to see them as forming groups of coordinate-values' names which taken together become capable of describing an event that is time-extended, spatially localizable, colored, and so on.The main requisite is that these subpropositional elements should have a "completely unsaturated nature", so that they can only be considered as names in the context of an elementary proposition.If we erase all of them but The coordinate names need not be numerical, i.e. sequentially generated out of an initial unit.They could be understood instead as being just "proper names" of the various dimensions of a manifold.one from a sequence, we will stay with an unsaturated expression, analogous to Frege's functions, a single coordinate-value name, which alone doesn't mean anything at all 2 .Hence genuine names will be unsaturated expressions which name coordinate-values and their combination would give the address of an atomic event localizable in space/time/color/sound manifold structures.An elementary proposition will thus be a sequence of coordinate-value names, like: [x, y, z], for example.
In the quote bellow Hyder suggests the very same distinction we are suggesting here between ordinary objects and the simple ones: The objects that Wittgenstein's names denote are characteristic features of classes of facts.Such objects […] have little in common with the things of our experience.They agree better with the conception of objects involved in the Helmholtz/Riemann definition of a manifold: the objects are determinations of the variables in a dimensional concept.They are like coordinate-values.And what corresponds to a full determination of the concept is an element in the manifold: a location, in other words, and not a coordinate-value.(HYDER, 2002, p. 5) We agree with Hyder that this is a reasonable option to explain the sub-propositional structure of elementary propositions.One gain of such interpretation would be that it offers a plausible image of how things are going to be at the bottom level, when we've eliminated all "general" terms.Another is that simple objects will be completely "foreign to the logical space".They will not occur in it, for they will only be values which determine manifolds possibilities.Whichever we construe as an "occurrence in the logical space" will have to be treated as a state of affairs, not as an object.Some problems still remain of course.As we said at the beginning, one of our working hypotheses is that the principle of logical independence of elementary propositions is unsustainable once we assume the construal of propositional sense as truth-conditions.The difficulty does not involve just the problem of color exclusion.Other pairs of empirical predicates, which could appear at this basic level, like sounds, would suffer from the same predicament.Further, if we take numbers as names of coordinates, we are going to have also problems like the one we 2 Cuter (2009, p. 7) for example talks about an unsaturated character of tractarian names.
Wittgenstein's unique "great analysis" volume 18 número 1 2014 had for the unsaturated terms, because ordinal numbers are connected by a successor relation, an internal relation (SRLF,993,p. 33).Coordinates could be construed as proper names instead, but then we will still have an asymmetry between "nominal parts" and "predicative ones", for we could not have the same position with two different colors (or other sensory qualities) and we would thus be back to the color exclusion problem.

Conclusion
Summing up, our hypothesis was that the construal of propositional sense as truth-conditions forced Wittgenstein to face several very important consequences at the bottom level of elementary propositions.To begin with, we explored a very different conception of "analysis", diverging significantly from the interpretation which considers Wittgenstein's analysis as completely modeled on Russell's Theory of Description.At the top propositional level the main difference is that according to our interpretation we have to analyze away all general terms and all "apparently singular ones".At that top level thus the goal is to eliminate all the hidden generality of our "ordinary terms".At the bottom level of elementary propositions there is another kind of difference, a difference concerning logical structure.The structure of elementary propositions should be distinct from the traditional subject-predicate one, for they are composed only by label devices, genuine names in a certain arrangement.We concluded too that those genuine names, differently from the ordinary ones, should have to refer in a direct way to simple objects, since no mediating descriptions could be available at this level.Finally, concerning the nature of simple objects, we suggested an alternative interpretation, similar to Hyder's account, in an attempt to explain better what we'll ultimately encounter at the end of the "Great Analysis".
We are not suggesting we've "saved" the Tractatus tough, and neither that we've proposed the only correct way to solve all its various difficulties.Our proposal was more of a critical nature.We've suggested that instead of looking at the Tractatus as a "project to be saved", we could look at it as an inquiry about the problems involved in the universal applicability of the subject-predicate structure and the idea of truth-conditions.It is as if Wittgenstein were forcing us to conclude that the understanding of "propositional sense" as truth-conditions implies that we would have to construe the process of analysis as a unique process.This process would gives us a complete determination of sense at the bottom level and, to reach this goal, we would be araceli rosich soares Velloso obliged to let go the thesis that the atomic base of our language should be structured in a subject-predicate form.
For Quine, as well as for the latter Wittgenstein, an alternative to this conclusion, which would also maintain some elements of the construal of propositional sense as truth-conditions, would be to accept some form of semantical holism and to abandon the idea that every sentence has an empirical content of its own, completely translatable into its truth-conditions.Quine's arguments proceeds in two different directions: to show that reductionism in all forms doesn't work and to create examples of reinterpretations that show an insurmountable indetermination in the traditional subject-predicate schema (QUINE, 960, §2).
One interesting idea which approximates Quine's arguments to those of Wittgenstein is to think about the Tractatus as an option for the other way out of the dilemma produced by the construal of propositional sense as truth-conditions.That is to say, we could give up the idea of finding an empirical content for each atomic proposition.Or we could keep trying to somehow translate "the significant discourse, statement by statement, into a sense-datum language", as Quine said Carnap has tried once (QUINE,996b,p. 39).Rather, in the Tractatus, Wittgenstein chose a third path: to abandon the desiderata of a subject-predicate schema for the propositions of the bottom level.This move allow him to keep the reductionist approach together with the idea of a unique complete analysis and maintain the search for multiple structured elementary propositions which could be isomorphic to the complex and structured manifold of experience.As Griffin puts it: for Wittgenstein in the Tractatus "… subject-predicate propositions cannot be elementary."(GRIFFIN,964,p. 57).Of course this "tour de force" finally shows some insurmountable difficulties steaming exactly from the same source: the inevitable presence of unanalyzable predicates that bar the logical independence of elementary propositions.
Wittgenstein's unique "great analysis" volume 18 número 1 2014 abstract our goal in this paper is to propose a new way of understanding just how the famous process of analysis proposed by Wittgenstein in the Tractatus could actually proceed.the guiding line of our presentation will be to assume that in this famous work Wittgenstein adopts as a procedure for elucidating the sense of our representational devices -the propositions -a search for a complete and exclusive determination of its truth-conditions.We will show that it was due to this motivation that Wittgenstein has proposed de analytical process that we will call here "the great analysis".the process of analysis proposed by him is a pretty radical one though and implies a lot of different challenges which has to be faced.our strategy will be to offer first an explanation of how this process could be executed and, then, try to extract its various consequences.our main interest will be to show how this radical criterion for the elucidation of propositional sense produces deadlocks and leads to a series of very important philosophical consequences.the most important ones are included in the following list: a strict frontier between senseful and senseless propositions; a view of analysis as a process of "analyzing away" all hidden generality contained in general terms, as well as in grammatically apparent "singular terms"; an alternative way of understanding genuine names and their ontological correlates, the simple objects, which fits better our approach; and finally, the most important one, the conclusion that we cannot have both: the presence of unsaturated parts in the elementary propositions and the construal of sense as truth-conditions.

Recebido em março de 2015
Aprovado em maio de 2015 araceli rosich soares Velloso […] If I can imagine objects combined in states of affairs, I cannot imagine them excluded from the possibility of such combinations.(TLP, 2.02) So, this conclusion seems to indicate the thorough inconceivability of uncombined simple objects.But as we are going to see next, Griffin's position does imply this possibility.