Download 5. the problem of implication and more Exams Physics in PDF only on Docsity!
5. THE PROBLEM OF IMPLICATION
The problem of implication is two-fold: how assertions are distinguished from implications and how much implied meaning is a necessary or a per- mitted part of semantical grammatical analysis. Difficult as these questions are to answer, they must be met head on or grammatical analysis remains shrouded in a mist of inconsistent intuition. It is certain that different reasonable answers are possible, but as a bare minimum we must strive to be consistent. And to be consistent is impossible without clearly formulated principles to serve as criteria of judgment. One of the basic inadequacies of transformational-generative grammar, for example, is the indiscriminate lumping together of assertions and implications in the realm of 'deep structure'. By this point in our analysis we should be clear enough on what constitutes assertions. The problem now is to determine what is definable as implied assertions and still distinct from actual assertions. To be of maximum use our conception of implication should exclude from consideration the infi- nitely great penumbra of suggestions made up of what individual readers at specific times and places can discern to their own satisfaction but that are not confirmable by any reasonable sampling of intelligent, disinterested readers. To construct various sorts of symbolic or iconic analyses or to draw out stream-of-consciousness connections is not what is meant here by impli- cation. Only agreed-upon interpretation can form the basis of grammatical and rhetorical analysis, and insofar as implications are agreed upon, they are as much a matter of convention as are the interpretations of assertions. Thus our notion of implication is perhaps closer to that of assertion than it is to the usual notion of suggestion. Indeed, one of our two basic categories of implication is "necessarily implied". Here the meaning is not strictly speaking asserted but must of necessity be assumed by virtue of what is asserted. One may disagree with the truth of a necessary implication, but one cannot rationally deny its necessarily implied character. There is, of course, more to implication than this, but by emphasizing this we are emphasizing the assertional basis of our conception of implication and discouraging the common attempt to find unlimited semantic significance in a composition. In characterizing the various manifestations of implication we might quite properly begin with a more
positive principle than "discouraging the common attempt to find unlimited semantic significance in a composition". But this is easier said than done. The easiest way to proceed, and probably the easiest to comprehend, is to begin with necessary implications. What is not necessarily implied is not necessarily agreed upon. Thus we will find very helpful an initial understanding about kinds of implication that no one could rationally deny.
An obvious sort of necessary implication is the syllogism of traditional Aristotelian logic. To assert that all men are mortal and that Socrates is a man necessarily implies that Socrates is mortal. And this is what we mean by necessarily implied: that asserting something precludes the legitimacy of denying something else. It does not mean that such an implication has actually been asserted; it only means that to assert the one and deny the other is inconsistent. Assertions are asserted; implications are not asserted. But an assertion can be just as inconsistent with an implication as it can be with another assertion. The syllogism, however, is not a prime example of what we mean here by necessarily implied because it involves a minimum of two separate assertions as the means of implying a third assertion. Perhaps we could refer to this as rhetorical implication and distinguish it from grammatical implication. Rhetorical implications are those that arise out of the relationships between assertions; grammatical implications arise out of single assertions. Before saying more about rhetorical implication, let us focus on simpler manifesta- tions — on single restrictive modifiers that, in addition to providing a necessary qualification, necessarily imply another assertion. The first kind of necessarily implied assertion results from the special kind of truth word being modified.
A. NECESSARILY IMPLIED
The that he is a government agent is irrelevant.
confess
There can be no dispute with these sentences as to what is subject, attribute, assertion modifier. Unlike the exclusive modifiers discussed above — where
The is (^) ( he was a government agent. }
i '
\
\
One cannot make these assertions and at the same time deny that he was a government agent. But in the following example, one can make the similar identity assertion and at the same time deny that he will not return:
One's report about what exists as a widespread fear does not necessarily include oneself among those who fear, nor does it equate what is feared will happen with what is known will happen. The second category of necessarily implied assertions results from compari- sons or contrasts. By no means all such assertions necessarily imply an addition- al assertion. For example, "Sam has more friends than Don" can be taken to imply that Don has friends, but there is nothing necessary in this interpreta- tion because the assertion is equally true if Don has no friends at all: "Sam has more friends than Don; indeed, Don really has no friends at all" is quite a reasonable sentence. But in the assertion "Sam has fewer friends than Don" we are obliged to infer that Don has friends because the comparison makes sense only insofar as Don provides the standard of comparison. If it makes sense to say that something is less or fewer than something else, it is because that something else is more or greater. And to have more or greater is necessarily to have. This difference between necessary and not necessary implication is not strictly speaking a matter of elliptical constructions. Neither of these two non-elliptical forms is quite idiomatic, but the second one still necessarily implies that Don has friends and the first one does not:
Sam has more friends than Don has friends.
Sam has fewer friends than Don has friends.
For the moment, this is all that needs to be said about comparison and contrast. The more common interpretation of such assertions is of the not- necessarily implied sort, and these will be discussed in the next section. The point to emphasize here is that reference to something in a comparison may
The widespread is l \ that 1 he will not return, i \ /
I
I / »
necessarily imply its existence but may not. What makes the difference is not so much a semantic understanding of the sorts of things being compared as it is a formal understanding of the sort of comparison being made. To be less than something else necessarily implies an assertion about the existence of the more:
Poverty has less influence than does affluence. (necessarily implies that affluence does have influence)
Children have fewer rights than do adults. (necessarily implies that adults do have rights)
To be more than something else implies, but not necessarily, an assertion about the existence of the less:
Affluence has greater influence than does poverty, (poverty may have no influence)
Adults have more rights than do children, (children may have no rights)
The third logical possibility for a comparison is that the things compared are equal, but for our purposes of deciding between necessarily and not necessarily implied, these are like the last pair of examples: they do not necessarily imply an additional assertion:
I am as rich as he is.
They are as quiet as a herd of elephants.
Someone may previously have claimed that he is rich or someone may have asked if they are quiet, but the two assertions of ours are equivocal on the existence of richness and quiet. All they commit us to is equality in relation to the pertinent standards. We would not be concerning ourselves at such length about the problem of implication if all that was involved was such obvious kinds of necessary implication. Implication is a problem because some aspects of assertions seem semantically essential yet neither are asserted nor necessarily implied — at least in the logical senses analyzed above. The most obvious manifestation of the problem is the reference to things without ever asserting their existence. In any ordinary, practical language use it seems to go without saying that we talk
use be considered as primarily a matter of referring. This, of course, is the familiar positivistic theory of meaning. At this late stage in the debate we can simply indicate our approval of the recent decline of such proposals. But quite aside from a fundamental disagreement with the theoretical validity of positivism's atomistic conception of meaning and of linguistic analysis, the goal we are seeking — to make possible the systematic analysis of extended discourse — would be impossible to reach if every assertion had to be reduced to a dozen or so sub-assertions before the main assertion could be related to the assertions coming before and after. Requiring the reduction of every sentence into its component assertions is quite a different matter from requiring the reduction of every assertion into its existential implications. Reference to something implies, in the absence of indications to the con- trary, its existence, but this implication is never necessary. A grammar book, full - as it necessarily is - of various kinds of sample assertions, would be almost meaningless, or at least certainly false, if we were obliged to infer that every reference to "my wife", "my husband", "my alcoholic father", "Albert A. Alweather", necessarily implied the existence of such. Sustained composi- tion would be an impossible business if we were required always to assert existence, and it would be almost as bad if we were required always to infer existence. We must thus conclude that, while we can usually assume that reference implies existence, this will by no means always be the proper inference in a given instance. The principle behind our refusal to make reference equivalent to necessary implication applies equally to restrictive modifiers as to what they modify. Whether the restrictive modifiers are mere single-word references to or whether they contain the full subject-predicate elements of an assertion, their restrictive function is the governing factor. Take for example the following pair of sentences:
My tent collapsed. The tent I slept in collapsed.
Each sentence contains just a single assertion, but the second one has a more detailed implied assertion by virtue of the subject-predicate form of the modifier: I slept in a tent. Restrictive modifiers (e. g. "My" and "I slept in") cannot be detached from what they modify and be made separate sentences without changing the meaning of the original assertions. "Lincoln was the president who freed the slaves" is not reducible to "Lincoln was a president" and "A president freed the slaves". "The tent I slept in collapsed" is not reducible to "A tent collapsed" and "I slept in a tent". But the question of necessary implication goes beyond this. It is probably safe to say that every restrictive modifier implies something
beyond its role of limiting or qualifying what is modified. But where more than one thing seems implied and different people have different formula- tions of what precisely is being implied, then there is no one necessary impli- cation that every rational person is logically obliged to recognize. In the sentence "My tent collapsed" there are several plausible candidates, and thus no certainty exists in accepting one as necessarily implied: I owned a tent. I rented a tent. I lived in a tent. I pitched a tent. I slept in a tent. Not unexpec- tedly, we find that the gain in economy made possible by pre-positional modification results in corresponding disadvantages. Not only is there a possible ambiguity in taking the modifier to be both restrictive and non- restrictive, there is a possible uncertainty in deciding what precisely is being implied. Admittedly, in many cases there is no problem. And in many others the differences between alternative interpretations will be of minor significance. But problems can and do arise, and the fact of the matter is that a full subject- predicate form of post-positional modification is a means of reducing con- fusion. There is an important difference between the pre-positional "My tent collapsed" and the post-positional "The tent I slept in collapsed", but this is a difference in precision of implication, not in different kinds of implication. In the latter the implication is precisely that I slept in a tent. However, there is nothing logically compelling about this implication. Different contexts could lend different existential interpretations to this restrictive modifier. Grammatically it is no different from
Greek tragedy was conceived on the day Zeus raped Leda.
One is free to infer that Zeus did in fact rape Leda, but he can still deny it without having to deny the whole sentence as a meaningful, appropriate assertion in a particular context. And this is the case irrespective of whether the relative pronoun is used. "Greek tragedy was conceived on the day that Zeus raped Leda", with its relative pronoun, has the same overall assertional meaning and the same not-necessary implication. And while on the subject of relative pronouns, we might note examples of when and where clauses. These result in restrictive assertion modifiers even when they include complete subject and predicate elements:
A new truck is parked outside. A red truck is parked outside. A Ford truck is parked outside. A pickup truck is parked outside.
Consequently, the four modifiers are restrictive, and as such they can give rise to implied assertions. In a given context we may or may not be justified in inferring from this sentence that other trucks are parked outside. Even out of context it is reasonable to assume that there are trucks in existence that are not new, not red, not Fords, not pickups. But there is nothing logically necessary in the rule of thumb that a distinction made implies the existence of an actual contrast. It is occasionally possible for multiple pre-positional modifiers to result in multiple explicit assertions; however, this requires some semantic means of making a precise distinction between multiple characteristics of one thing and multiple things being characterized:
The first example is not necessarily the same as "He has a large truck" plus "He has a red truck", because these two sentences could be referring to two different trucks. However, the second example does clearly and unequivocally refer to one man's one nose, regardless of whether there is one sentence with multiple modifiers or two sentences with single modifiers. Theoretically, we would expect that separable parallel modifiers could always be conjoined by and. In practice, this is not always idiomatically the case, although it frequent- ly is:
He has a large red truck.
Nansen's crossing of Greenland was a trip. difficult
/ long ^ Nansen's crossing of Greenland was a
long Nansen's crossing of Greenland was a and
long and difficult
What is involved in the interpretation of "A new red Ford pickup truck is parked outside" is not primarily a matter of implication beyond what is explicitly stated but a matter of recognizing the different elements of an explicit assertion. No assertion is semantically atomistic. "A truck is parked outside" is a composite of several unified ideas, and the addition of four restrictive modifiers to the subject does not change this. The problem of implication is not a matter of extracting atoms of ideas from an assertion but of constructing additional assertions on the basis of what is explicitly asserted. Grammatical implications are structures of restrictive modification that are interpreted as assertions even as their restrictive granfmatical function remains unchanged.
As a general rule we assume that a distinction made implies the need for it and thus that we can infer some kind of contrast. To refer to "my oldest brother" implies that I have more than one brother. Indeed, we can go so far as to claim that this is a necessary implication, because the superlative degree of comparison could not logically have been used if I had only one brother. If I refer to "my older brother", this too implies that I have more than one brother, but it may be that the comparative degree indicates only that I have one brother and that he is older than I. Thus we cannot term this a necessary implication. Necessary implication is not a matter of great probability but of the logical impossibility of making an assertion and denying the resulting implication. Words like fact and oldest have as part of their meaning this ready potential for indicating logical, either/or, necessity. However, most words are like new, red, Ford, pickup, and truck in having for the most part potential only for equivocal implications.
B. NOT NECESSARILY IMPLIED
We need to do little more than refer back to our discussion of comparison and contrast by way of noting that this type of construction never states more than one assertion, but it does imply more than one. Some of these are necessarily implied, as we have seen, but by no means all. "Sam has fewer friends than Don" is not contradicted if Sam has no friends at all. However, in addition to this problem of the more and the less is the problem touched upon briefly back in Chapter 2, where we discussed numerical characteristics.!
If we say that Sam is shorter than Don, does this assert or necessarily imply that Sam is short? If Sam is six feet tall and Don is seven feet tall, should we have said instead that Don is taller than Sam? This latter transformation would seem to preclude any objection to finding as a necessarily implied assertion that Don is tall.
Are we obliged to infer that money (among other things) is required? and that whatever else this session may be it is also a bull session? Would it (as a matter of context) ever make sense to say that more than money is required and then to deny that money is required? Is it (as a matter of logical or semantic analysis) ever possible to say that being more than something does not include as part of itself that which it is more than? Volunteers are called for on a stormy night to man a lifeboat; someone pulls out his wallet and offers to support the venture. He is curtly informed that more than money is required. Indeed, there is no substitute here for courage, strength, and endurance. Perhaps there is an element of irony in the answer; perhaps it is a deliberate understatement; perhaps the most precise answer would have been that money is not required. But ironic or not, such an answer would have been appropriate and comprehensible under the circumstances. However, we are not really obliged to invent fictional situations to test the question. The concept of more than can, of course, be interpreted in terms of more of the same: More than a couple of dollars is required to buy this camera. More than your half-hearted effort is required to gain a place on this team. But the concept can also be interpreted in terms of something else more important: More than good intentions is required to become a successful business man. More than an inside tip is required to play the horses. Many dollars are required to buy this camera. All-out effort is required to gain a place on this team. These are more-of-the-same comparisons. Ruthless- ness is required to become a successful business man. Money is required to play the horses. Because one can be a successful business man without good intentions and play the horses without inside tips, these are something-else- more-important comparisons. We are thus not able to say that such assertions necessarily imply that the specified criterion is included, even though in- sufficient. In the sentence about the bull session we might make this point more obvious than it is in its "not just" form by rewriting the sentence thus: This is not a mere bull session. A committee meeting that is drifting further and further from the subject at hand does not have to become a bull session for the chairman to make the comprehensible and appropriate statement that this is not just, or not a mere, bull session.
In the third example, we must take "rather than work for slave wages" as a limiting assertion modifier, because the sentence does not unequivocally assert that I will not work. Unlike the two previous sentences, this one has an assertion modifier; but like them, its meaning is that of a comparison. The speaker does not claim that he will not work at all nor that he will not work for slave wages. Rather, when the two alternatives are presented, he will choose the one over the other. However, the sentence leaves open the possi-
bility that he might be forced to work for slave wages and the possibility that he will be able to work for non-slave wages. It is probably fair to infer from the sentence that the speaker will not work for slave wages. We assume as a Constitutional right that "Neither slavery nor involuntary servitude, except as punishment for crime whereof the party shall have been duly convicted, shall exist within the United States, or any place subject to their jurisdiction". But there is nothing logically compelling about an inference based on such an assumption. Rather than often functions as a negative, but only by implication. In essence it is a comparison, like more and less-, but unlike them, it does not create a necessary implication. Where the phrase does occasionally occur with an explicit negation, it is redundant. In sentences like "He is twenty rather than twenty-one" and "He is a Swede rather than a Norwegian", nothing new is added by the "rather than" phrase. To say that someone is twenty means, among other things, that he is not twenty-one; to say that he is a Swede means, among other things, that he is not a Norwegian. The next general category of not necessarily implied implications is rhetorical questions. An explicit question is not, of course, in itself an asser- tion. It may contain within itself as an additive modifier a separable assertion:
Why do you, | j h e least qualified, claim the greatest reward?
"You [are] the least qualified" is unequivocally asserted; there is nothing inferential about it; it can be lifted out of the context without altering the nature of the question. There are, however, two possibilities for implications: You claim the greatest reward. You shouldn't claim the greatest reward. The first of these cannot be treated as a separate assertion because to lift it out of context would destroy the question. "Why do you claim the greatest reward" is not a composite of "You claim the greatest reward" and "Why do you claim? " The precise nature of the claim is lost this way. We can, however, interpret the question as necessarily implying that you claim the greatest reward. The difficulty with this interpretation is that our standard test for necessary implication is not really applicable to questions. We cannot ask whether asserting one thing and denying another is logically inconsistent, because there is not a stated assertion and an implied assertion but rather a question and an implied assertion. This gives rise to the frustrating kind of situation exemplified by the question, "Have you stopped beating your wife yet? " The answer called for is either "Yes" or "No". But if you have not in fact been beating your wife, how do you answer? The question clearly
Yet, while a reference like "The rape of Mother Nature" cannot be taken to imply existence in the way that "the rape of the Bangladesh women" can be, there are additional kinds of implicational possibilities not available to true/ false referential assertions. Metaphorical assertions imply, although of course not necessarily, important similarities between things from quite different realms of discourse. To speak of the natural, biological, non-cultural, non- technological processes of the Earth in terms of motherhood is to imply that this aspect of our world deserves the highest respect and consideration we are capable of, even as we are allowed great familiarity with the object of respect. This is the traditional metaphorical implication. To take this traditional metaphor as the basis of a new metaphor is to create a further realm of implication. If motherhood is deserving of the highest respect and consideration because it is the very source of our existence, growth, and development, then the ultimate violation of this source, its pollution, is deserving of the greatest condemnation. This kind of metaphorical statement implies a kind and intensity of value judgment that could hardly result from a reference, for example, to "the ecological imbalance". To speak of an imbalance seems to imply that it is possible to restore the balance, but to speak of a rape seems to imply that something irrevocable has happened, akin perhaps to gaining the knowledge of good and evil. Of necessity, metaphors always imply, but the individual implications are never necessary ones. We are a long way here from "I confess that I shot him". But what is equally true of the two examples of the two kinds of implication is that the meaning of each sentence is not completely interpreted until we have articulated the inevitable implications. The fact that most implications are a matter of disputable interpretation does not gainsay their presence. This chapter, coming as it does before the discussion of inter-assertional relations, is essentially concerned with grammatical rather than rhetorical implications. However, by way of introducing the discussion of the latter (the subject of the next chapter), we might profitably examine a special sort of inter-assertional implication that arises out of a standard grammatical pattern. This is the implied causality of 'participial' constructions. The use of 'participials', especially having and being, as a means of making an explicit additive assertion and an implicit causal relationship is common, but this convention slips easily into problem cases — where not only is the implication unclear but also where the distinction between restrictive and non-restrictive modification is unclear. First, let us make clear the sort of having and being constructions that are not at issue here. When the 'participial' form is the subject or the attribute, there is no additive modifier and no question of a causal implication:
Being a woman is a disadvantage when applying to medical schools. Having a headache is no excuse. Having children is having a life-time commitment.
As additive modifiers, have and be phrases are set off by punctuation, although their usual position is before rather than after the thing modified:
Having failed her exam, she was ineligible.
She was ineligible. She had failed her exam.
His daughter, ^having married a bum^j is living on welfare.
His daughter is living on welfare. His daughter has married a bum.
jjieing a woman,J she was ineligible.
She was ineligible. She was a woman.
Written as explicit rather than implicit causals, these sentences readily lend themselves to parallel structure:
failed her exam She and therefore was ineligible.
married a bum His daughter and as a result is living on welfare.
- Being weak with hunger, food was finally served.
The boat did not eat our lunch; it did not argue in vain all evening; the teacher was not the one unable to write on the assigned t o p i c \ f o o d was not weak with hunger. The counterpart of this problem is obvious syntactical parallelism but ambiguity of implication. The use of having and being in additive modification clearly establishes a prior condition, and the semantic convention seems fairly well established here that post hoc, ergo propter hoc : to come after is to be the result of. There seems to be no other special rationale for the pre-positional 'participial' form. But sometimes such constructions seem simply to be post hoc, and the reader looks in vain for a because/therefore relationship where nothing but and can be justified:
Having left the theater, they walked toward the delicatessen. They left the theater and walked toward the delicatessen.
Packed as full as practicable by automatic machines, it contains the full weight indicated. It was packed as full as practicable by automatic machines and contains the full weight indicated.
At this point our topic of 'participial' modifiers begins to expand beyond having and being, but the problem is familiar: deciding whether we are confronted with one assertion or two. There is no problem with the two parallel sentences above: they are composed of pairs of assertions. But with the modificational counterparts a case can be made for "having left the theater" and "packed as full as practicable by automatic machines" as being either restrictive or non-restrictive. Partly this kind of problem results from our insistence on imposing an either/or conception of analysis on modification, but partly it results from the latent ambiguity in this kind of sentence structure. As long as the 'participial' phrase uses just having or being and as long as there is a reasonable causal interpretation, then we are dealing with non- restrictive modification. The proof of such an interpretation is the converting of the modificational structure into a parallel structure and the providing of an explicit causal conjunction (e. g. because or therefore). An alternative possibility to parallel structure as a more adequate formula- tion of some of these 'participials' is as restrictive modification. The phrase may simply be the answer to questions of when, where, or under what circumstances:
Returning from lunch, I was accosted by four men.
Hanging from the gallows, his corpse is a warning to all traitors.
Swimming beyond the reef, he was attacked by a shark.
"When were you accosted? " "Where is his corpse? " "Under what circum- stances was he attacked? " Answers to these questions provide restrictive assertion modifiers. But is there, in addition to this kind of interpretation, a causal implication in these sentences? On the one hand, it seems to make little sense to say that I returned from lunch and therefore was accosted by four men. On the other hand, it is not unreasonable to say that there is some kind of causal relationship between the corpse hanging from the gallows and the reaction of traitors who see it. Yet the second sentence is not about what the hanging corpse causes but what it is. We are not told that traitors are taking warning but only that the corpse is a warning. No more than the first sentence can the second one be restructured as a causal relationship without altering the meaning. The third sentence is more obviously causal than the first one (being attacked by a shark seems more likely to accompany swimming beyond the reef than being accosted by four men is likely to accompany returning from lunch). But unlike the second one, it is a matter of mere accompaniment rather than of essential connection. This causal ambiguity arises because of a tendency in English to treat this sort of pre-positional modification as a deliberate deviation for causal purposes from the more usual post-positional structure:
I was accosted by four men when returning from lunch.
His corpse hanging from the gallows is a warning to all traitors.
He was attacked by a shark while swimming beyond the reef.
These restrictive assertion modifiers clearly and unambiguously answer questions of when, where, and under what circumstances and raise no causal implications. The journalese tendency topile on unintegrated information at the beginning of a sentence results in more than "a stylistic monstrosity" (as one commentator labels it); it often results in causal implications that are not clearly in accord with the explicit meaning. It would be fortunate if all of our difficulties with these 'participial' con- structions could be attributed to the innate ambiguity of the form, or at least to the illegitimacy of attempting causal implications except with having and
