The Possibility of Paraphrase Open Access

Abstract

It is often claimed that, in at least some areas of language use, the relation between form and content is such that any attempt at reformulation or paraphrase amounts to a distortion of the significance of the original wording. In this article, I set out to vindicate an undemanding yet nontrivial conception of paraphrase. According to the rhetorical relations account of textual cohesion proposed, the meaning specifications required by a collection of sentences in order to constitute a text pave the way for a kind of reformulation which is in solidarity with the possibility of paraphrase. I substantiate my approach with prosaic and poetic examples from Woolf and Dickinson, respectively.

I. INTRODUCTION

It is a commonplace of criticism that it is not possible to say the same thing in different words. Restatement, reformulation or paraphrase is thought to be impossible because, in at least some areas of language use, especially literature, content and form are inseparable, indistinguishable or even identical. This unity of form and content speaks against the possibility of paraphrase, insofar as a certain distinction of form and content is required in order for the restatement of the content in a different form to be feasible. Sometimes it is conceded that paraphrase is not strictly impossible, but emphasized that it is in any case irrelevant, insofar as it is primarily form, not content, which matters. According to many theorists (from Brooks through Gadamer to Lepore),¹ paraphrase is a distortion of the significance of the original wording, an uncalled-for misrepresentation of it, an inadequate response to its integrity or a heretical imposition upon it.

In this article I set out to vindicate an undemanding yet nontrivial conception of paraphrase. The kind of reformulation which I will be concerned with cannot be conceived of as an arbitrary imposition, insofar as it is grounded in meaning specifications which are required by a collection of sentences in order to constitute a text, whether it be prosaic or poetic. This kind of paraphrase is, I believe, an integral and important part of interpretation. Before presenting my own conception of paraphrase (section iv), I pave the way for it by looking at the assumptions made in traditional statements of the impossibility of reformulation (section ii) and also at the merits and limitations of some recent vindications of paraphrase (section iii).

II. THE IMPOSSIBILITY OF PARAPHRASE

Bradley is a typical proponent of the impossibility of paraphrase, saying that “in true poetry it is, in strictness, impossible to express the meaning in any but its own words, or to change the words without changing the meaning” (1926, 19). About “To be, or not to be—that is the question,” he then says:

You may say that this means the same as ‘What is just now occupying my attention is the comparative disadvantages of continuing to live or putting an end to myself.’ And for practical purposes—the purpose, for example, of a coroner—it does. But as the second version altogether misrepresents the speaker at that moment of his existence, while the first does represent him, how can they for any but a practical or logical purpose be said to have the same sense? Hamlet was well able to ‘unpack his heart with words,’ but he will not unpack it with our paraphrases. (Bradley 1926, 20)

As usual, the dismissal of paraphrase comes with some qualifications. The impossibility is not claimed to be complete: paraphrase manages more or less to preserve content, but spoils the effect. Schleiermacher says that paraphrase “can perhaps reproduce the contents of a work with some accuracy, but the impression made by the work must be dispensed with altogether” (2004, 48). Coleridge goes further by denying that even content is preserved. Yet after having declared untranslatableness to be “the infallible test of a blameless style,” he adds that he “include[s] in the meaning of a word not only its correspondent object, but likewise all the associations which it recalls” (Coleridge 2004, 536). Thus Coleridge seems to concede that the denotation may be preserved while the connotations are lost. There is thus hesitancy as to the exact extent of the devastation brought about by paraphrase. If paraphrase is not strictly impossible, because it captures some meaning, it is at least uninteresting or irrelevant, in so far as it passes by what is essential or of importance in the original wording.

Dismissals of paraphrase usually also come with qualifications as to the range of the impossibility. In effect, often the impossibility of paraphrase is mentioned in support of the distinctiveness of some specific area of language use. Typically it is said that in the case of metaphors, or poetry, or literature, it is not possible to say the same thing in different words. In statements such as Bradley’s it is implied that it is the specificity of poetical language which precludes the possibility of paraphrase. Yet when it comes to other areas of language use many thinkers are actually quite reluctant to accommodate paraphrase. Dummett questions the idea that theoretical discourse is paraphrasable when affirming that great philosophical writing “aims not to formulate theses detachable from their authors’ expression of them” (2010, 35). Goodman suggests that literal language is not different from metaphorical language when it comes to paraphrase, insofar as “paraphrase of many literal sentences also is exceedingly difficult” (1979, 176).

An important motivation of many theorists opposing paraphrase is the conviction that (great) literature exhibits a “special intimacy of form and content” (Kivy 1997, 116). Readers of literature should not, like ordinary hearers, attend bluntly to subject matter, but sensitively to “subject-realized-in-just-this-way” (Lamarque 2015, 29); in literary texts every detail serves some purpose, so that “meaning is bound up in the very words [employed]” (Cavell 2002, 78). Recourse to paraphrase is thought to be oblivious to this distinctiveness of literature. The connection of paraphrase and such inconsiderateness is facilitated by two specific assumptions. Paraphrase is taken to be the ambitious project of exhaustively capturing the meaning or significance of an entire work of art, or an elusive passage or a particularly complex formulation in a couple of lines or in plain prose. This assumption is questionable, since it is doubtful that anyone ever pretended that such a paraphrase was possible. One is tempted to point out that there is no reason for paraphrase not to be partial and approximative and capture just a bit of the import of the original and unique formulation. One reason why such a less ambitious paraphrase is not envisaged by deniers of paraphrase is certainly a second assumption usually made, according to which paraphrase amounts to replacement, so that whatever is not captured by the paraphrase is irremediably lost. This assumption is also quite peculiar, insofar as paraphrase could be thought of as accompanying the original formulation instead of taking its place. Deniers of paraphrase apparently conceive of paraphrase as translation rather than commentary. It is interesting to note that interlingual translation, which undeniably involves actual replacement—a translation is not good if one must have recourse to the original text in order to understand it—, even though it is frequently problematized, is not something which poetically sensitive people discourage other (poetically sensitive) people from doing.²

There is thus within the commonplace of the impossibility of paraphrase uncertainty as to the amount and the locus of the loss of meaning as well as assumptions about exhaustivity and replacement. If paraphrase did amount to exhaustive replacement, it would perhaps be impossible. But there seems to be no necessity to think of paraphrase in this way. Often when paraphrase is vehemently rejected, one has the impression that what is at issue is not really paraphrase per se, but rather the uniqueness or complexity of some particular formulation.

This paves the way for vindicating a relaxed conception of paraphrase. Yet one must of course avoid going too far in the opposite direction. For surely, if paraphrase is supposed to do no more than capture a little bit of the original formulation, then it will certainly be too trivial to be of any real interest at all. The challenge for anyone wanting to defend paraphrase is to show that there is such a thing as a partial and approximative reformulation which is neither uninteresting nor irrelevant, insofar as it actually contributes something essential to our understanding or appreciation of the original formulation, and which is in some sense called for by the original formulation itself and not opposed to the proper appreciation of the poetical.

III. RECENT VINDICATIONS OF PARAPHRASE

There are in the recent literature attempts at rehabilitation of paraphrase along the lines just proposed. Kivy argues that paraphrase preserves propositional meaning and to the extent that poems have semantic content at all, they are paraphrasable. Kivy’s target is for example Bradley claiming that in (true) poetry horse and steed do not have the same meaning. On this view, if in this couplet from Byron, “‘Bring forth the horse!’ The horse was brought;/ In truth he was a noble steed!” (Byron 2009, lines 358–359), horse and steed are transposed, so that we have “‘Bring forth the steed!’ The steed was brought;/ In truth he was a noble horse!” (Bradley 1926, 20), the meaning has changed. Kivy admits that if we take steed to mean “spirited horse,” the meaning of the two lines respectively certainly changes. Nevertheless he continues: “Has the meaning of the whole couplet changed? My claim is that it has not. Something has changed, of course. The aesthetic character of the couplet has changed. But that is just beside the point. Let us say its mood has changed. But again, that is beside the point. What hasn’t changed is its meaning” (Kivy 2011a, 168; compare Kivy 1997, 2011b and Jolley 2008).

It is not altogether clear what kind of meaning Kivy thinks paraphrase should capture. On the one hand, he insists on propositional content (Kivy 2011a, 171), which points in the direction of conventional meaning. On the other hand, he talks about what the author intends to convey and holistic meaning (Kivy 2011b, 372–373), which leaves open for more specifically contextual meaning. In his only concrete example, merely lexical meaning is at issue, which threatens to make his paraphrase appear too predictable and rather trivial. It would have been valuable to have some more substantial examples and to know how the meaning to be captured is to be established when it transcends dictionary meaning. Another vulnerability about Kivy’s approach is his taking for granted the relevance of propositional content. This is plausible enough, but nevertheless something which at least some deniers of paraphrase are inclined to question. Kivy also says that it is necessary to have recourse to paraphrase in order to appreciate the form/content unity, but this point is not further developed (2011b, 369).

Another vindication of paraphrase is due to Hills. After having insisted that the reading of poetic language certainly does not exclude the identification of propositional content (Hills 2008, 14), he advances the stimulating idea that paraphrase, far from being discouraged by the authors of allegedly unparaphrasable formulations, actually is practiced by the authors themselves.³ Accordingly, Hills speaks of authorized paraphrase, his idea being that since authors paraphrase their own language, paraphrase can be neither impossible nor improper. The author’s paraphrase sanctions, as it were, the reader’s paraphrase. One of his examples comes from Pascal: “Man is but a reed, the most feeble thing in nature; but he is a thinking reed. The entire universe need not arm itself to crush him. A vapor, a drop of water, suffices to kill him. But if the universe were to crush him, man would still be more noble than that which killed him, because he knows that he dies and the advantage which the universe has over him; the universe knows nothing of this” (Pascal 1958, 97). Hills comments: “‘Man is but a reed… but he is a thinking reed.’ That much is metaphor. The rest, I contend, is paraphrase—paraphrase from the horse’s mouth, as it were—Pascal’s own paraphrase of Pascal’s own metaphorical utterance” (Hills 2008, 21; compare Hills 1997).

Such an example is definitely neither predictable nor trivial. However, even if one is not hostile to paraphrase as such, one may nevertheless want justification for a particular paraphrase. Cavell, who has the unorthodox view that metaphors can be paraphrased, affirms that when Romeo says “Juliet is the sun,” he “means that Juliet is the warmth of his world; that his day begins with her; that only in her nourishment can he grow. And his declaration suggests that the moon, which other lovers use as an emblem of their love, is merely her reflected light, and dead in comparison; and so on” (Cavell 2002, 78–79). Bradley and others would presumably say that Cavell somehow or other misrepresents Romeo, but granting the possibility of an undemanding kind of paraphrase, one would still like to tell Cavell’s paraphrase from mere association.⁴ Hills says that he has paraphrased from the horse’s mouth, but it is unclear to what extent. One question is whether Pascal thinks of what follows after the initial metaphor as its paraphrase. Another question is whether, even if it were clear that he did, that would be sufficient for us to take it as paraphrase. It happens that authors pretend to put their thoughts in different words, though we doubt it is so.⁵

I think that these rehabilitation projects point in the right direction, but could be further elaborated. Paraphrase had better be more clearly substantial than Kivy’s example and Hills’s authorized paraphrase should be better justified. We would like to have clear reasons to say that paraphrase of a substantial kind occurs or that it is vital to have recourse to it. It will here be instructive to look at Lamarque’s response to Kivy. Lamarque argues for the form/content unity and says about paraphrase: “Bradley doesn’t deny the ‘practical’ usefulness of the Kivy-style paraphrase but merely insists that a fine-grained response to the line, doing justice to the dramatic context and speaker, cannot be satisfied with any expression but the line itself” (Lamarque 2010, 40). Lamarque believes that the unity of form and content is not discovered in poetry, but demanded by poetry when read as poetry. It is possible to paraphrase a piece of poetry, in the sense that nothing in the poetical text itself prevents us from doing so. However, the kind of reading strategy properly applied to poetry is precisely a kind of reading strategy for which there is unity of form and content. Let us exploit what Lamarque concedes, namely that there are reading strategies which do make a distinction between form and content. In order to vindicate the general possibility of paraphrase, we have only to find a reading strategy involving a distinction between form and content and which even proponents of form/content unity would admit is properly applied to any text.

IV. RHETORICAL RELATIONS AND MEANING ADJUSTMENTS

I believe that a modest kind of paraphrase could be vindicated from the consideration of the contextual specifications of meaning which result from textual cohesion. These are neither uninteresting, in so far as they are unpredictable from a lexical point of view, nor inessential, irrelevant, or arbitrary, because they take part of the meaning of the term in so far as the term is a component of a text.

IV.A. Cohesion and Rhetorical Relations

A text, by definition, is not an arbitrary collection of sentences, no more than a sentence is a mere list of terms. Let us call cohesion whatever it is that makes a text a text. Cohesion has been a rather neglected area of philosophy of language and even of mainstream linguistics. According to the most promising account, however, cohesion is achieved by so-called rhetorical relations (or coherence or discourse relations). A collection of sentences is a text if and only if each sentence is connected to at least one other sentence in the collection via a rhetorical relation. Among such relations we find Narration, Occasion, Background, Continuation, Elaboration, Explanation, Exemplification, Consequence, Parallel, Contrast, Question–Answer, Restatement, and so on. By telling us what the sentences of a text have to do with each other, rhetorical relations account for the fundamental level of text structure which is constitutive of a text as such.⁶

Rhetorical relations may be signaled by connectives, discourse particles, and other lexical terms, such as too, but, as a result, in other words. They are often enough suggested by the linguistic content of the sentences. Fundamentally, however, rhetorical relations are not a manifest feature of sentences, but assigned to the sentences by the reader. Assignment of rhetorical relations is what reading at a fundamental level amounts to; the text is constituted by the reader’s reception. The reader has a great amount of liberty in assigning rhetorical relations, the linguistic material often being underdetermined with respect to them. One and the same sentence may be connected to other sentences via several different, and even conflicting, rhetorical relations, which reflects the fact that a text typically can be read in various ways. Some readers want to assign the rhetorical relations intended by the author, and they use connectives and other cues as evidence of the author’s intention. Other readers may want to assign those relations which they think make best sense of the collection of sentences or which make best sense to themselves. In any case, anyone claiming to be reading a text has to assign rhetorical relations.

The theory of rhetorical relation assignment, that is which rhetorical relations are reasonably assigned to a given sequence of sentences, is a topic which I will not go into here. What is immediately relevant is that rhetorical relations bring about connection of sentences by imposing constraints on their content. Rhetorical relations come with requirements; the reader assigning a rhetorical relation has to show how it obtains. The rhetorical relation Explanation, for example, requires there to be a certain temporal order between the sentences which it connects, in so far as effects follow upon their causes. A reader taking a sentence following a sentence describing an event as describing its cause must consider temporal and textual order not to match. In most cases, the reader will have to complement compositional meaning by various sorts of specifications in order for the sentences to connect. Many contextual determinations of meaning such as disambiguation, reference resolution, temporal determinations, bridging reference, saturations, explicatures, implicatures, narrowing, broadening, enrichments, and ad hoc concepts discussed in the semantics and pragmatics literature may thus be regarded as a side effect of the establishment of textual cohesion. Instead of saying that these determinations depend on the context, the speaker’s intention, the discourse situation or considerations of what is interesting or relevant, we might simply say that they depend on cohesion. Such meaning specifications are necessary in order for sentences to be connected and thus required by the very constitution of texts.⁷ Here is a simple example to illustrate the approach:

• (1) a. He drank up his tea.

  • b. Some drops fell on his little pointed beard.

  • c. He took out his large silk handkerchief

  • d. and wiped his chin impatiently.

(Woolf 2008, 14)⁸

Many readers take such a sequence of sentences as a text without noticing the interpretive efforts which go into composing this heap of sentences and clauses into a text. Insofar as we read this particular collection of sentences as constituting a text, we take the sentences to stand in specific relations to each other, and we amplify the content of the compositional meaning of the sentences accordingly. Different relation assignments are certainly possible, but Occasion naturally comes to mind. An antecedent event serves as the occasion for a subsequent event. Certain specifications of meaning will follow. In order for his drinking tea to occasion the falling of the drops, the drops should be drops of tea. Furthermore, if the falling of the drops and his taking out the handkerchief occasion the wiping, he should have wiped away the drops of tea with the handkerchief. In (1b) we add “of tea,” and in (1d), we add “the drops of tea away” and “with the handkerchief.”⁹

These adjustments do not amount to free associations which the reader is disposed to make in order to have something more concrete, vivid, imaginable, intuitive, interesting, relevant, or truth-conditionally evaluable.¹⁰ Admittedly they do not belong to higher level or “adventurous” interpretation. However, they are tied to the constitution of the text as a text. Cohesion by means of rhetorical relations shows that there is a solidarity between contextual specifications of meaning and the constitution of a sequence of sentences into a text. In order to read a pair of sentences as a text, it is necessary to impose some rhetorical relation or other and to show how it obtains. Provided that the rhetorical relation has semantic impact and that the linguistic meaning of the sentences is underdetermined with respect to it, this assignment will result in various contextual specifications of meaning.

My claim is that the kind of meaning adjustments which we have just considered ground the possibility of paraphrase. In an important sense, when it comes to the meaning of one of the lines in the Woolf extract, the reader cannot be satisfied with the line itself. The contribution of the line is not exhausted by its compositional meaning. It would be inadequate to say that we make these meaning specifications animated by a coroner’s concern or for reasons of practical usefulness; we do so as readers doing justice to the lines as a text. Our taking sentences and clauses as forming texts motivates us to express their meaning in other words than the words used by the author. Here wiped his chin means the same as “wiped the drops of tea away from his chin”. In this sense, I think that paraphrase is not only possible, but demanded by the text.¹¹

So far we have seen that a lexical term, according to the requirements of rhetorical relations, must be taken to ascribe properties in addition to the property lexically expressed by it. Paraphrase, in order to be more fully interesting, demands, as it were, alternative words rather than merely additional words. We would have evidence for a more full-blooded kind of paraphrase authorized by the text, if we could show that the property ascribed is also qualitatively distinguished from the property expressed.¹²

IV.B. Rhetorical Relations in Prose

Consider the following extract:

• (2) a. “It’s been a long day; you look pale.”

  • b. “And you look tired too,” said Kitty.

(Woolf 2008, 77)

Let us assign Parallel to this pair of sentences, which is certainly suggested by Kitty’s use of too in (2b). Parallel requires the same predication to be made in the two sentences.¹³ This would evidently be the case if we had, for example, “Kitty looks pale. You look pale too.” In (2), however, it is not evident from the surface linguistic form of the sentences that same property predication occurs. There is neither repetition of lexical terms nor property identity at the level of the lexical meaning of the terms, since it is possible to look pale without looking tired and vice versa. Hence, it is required that we make meaning adjustments so that there be parallel elements in the properties ascribed by look pale and look tired. Accounting for how Parallel obtains in (2) may be done by taking the property ascribed by the predicate term look pale to be a novel property P* comprising the properties of looking pale and of looking tired. Our assignment of Parallel does not go through without our taking the term in (2a) to ascribe a property qualitatively different from the one it lexically expresses and constructed according to its correlation with the term in (2b). To state that Kitty looks pale and that any alternative wording of the predication is inappropriately coarse-grained would not do justice to the predication made. Parallel requires looking pale here to amount to looking tired and that tells us something important about the property ascribed in (2a). This departure from the original formulation is demanded by the sentences insofar as they compose a textual sequence.

Here is another pair of sentences:

• (3) a. Now that the daylight was extinguished and the air glowed yellow and crimson, she no longer felt absurd.

  • b. On the contrary, she felt appropriate.

(Woolf 2008, 172)

We may assign Contrast here, as suggested by on the contrary in (3b). (3a) and (3b) considered separately ascribe the properties of no longer feeling absurd and of feeling appropriate respectively. At the lexical level there is no strict contrast relation between the terms in question. Feeling absurd is not the opposite of feeling appropriate. It is possible to feel absurd and appropriate at the same time (for example, when complying with certain rules of etiquette). We must consequently adjust the content of the terms in (3) so that Contrast may be said to obtain. This relation requires there to be contrasting elements in the sentences and clauses connected. The property ascribed in (3a) by felt absurd is not the property P of feeling absurd, that is the property expressed by the term, but rather something like the property P* involving absurdity and inappropriateness, constructed in accordance with the requirements of the rhetorical relation. The same goes for felt appropriate in (3b). The property ascribed is the property P** involving appropriateness and nonabsurdity. The predication in (3a) amounts to her no longer feeling absurd and inappropriate and in (3b) to her now feeling appropriate and not absurd. If now we say that she felt appropriate means the same as “she did not feel absurd,” I think that would not be inappropriate.¹⁴

In (4), we have sentences which reasonably are connected via Continuation:

• (4) a. His small blue eyes looked round them as if to find fault; at the moment, there was no particular fault to find; but he was out of temper; they knew at once before he spoke that he was out of temper.

  • […]

  • b. His gloom relaxed a little.

(Woolf 2008, 12)

If his gloom in (4b) refers to the state of mind ascribed to him in (4a), which Continuation requires, then out of temper must be taken to involve the property of gloom, something which is not lexically supported; it is perfectly possible to be out of temper without being gloomy. What is stated in (4b), insofar as (4b) is connected to (4a) via Continuation, is that he is now less out of temper.

(5a) and (5b) may also be connected via Continuation:

• (5) a. “There aren’t any adventures for an old fogy like me,” said the Colonel surlily.

  • […]

  • b. Then he seemed to repent of his gruffness; he pondered for a moment.

(Woolf 2008, 14)

We must here take the surliness in (5a) to involve gruffness.

Here is an instance of explanation:

• (6) a. Eleanor took off her dress, and stood in her white petticoat washing herself, methodically but carefully,

  • b. since they were short of water.

(Woolf 2008, 187–188)

There are many ways of washing carefully and water-wasting ways of washing may be careful washings. However, if we take (6b) as presenting the reason for the manner of washing in (6a), washing carefully excludes washing with a large amount of water. Thus the property ascribed by washing carefully includes the property of washing with a small amount of water, whereas methodically apparently ascribes a property suggestive of washing with a large amount of water. If we say that she washed carefully here means (more or less) the same as “she washed with a small amount of water,” we would not change or distort the meaning of the original formulation, but simply tell what it (partly) amounts to in the context at hand. The aesthetic effect is certainly different, but we have said something essential for the meaning of the term as part of a text.

Let us last consider this Restatement:

• (7) a. “What an undertaking!” she said.

  • […]

  • b. “What a business!” said Eleanor again.

(Woolf 2008, 186–187)

The author’s again is evidence that she takes Eleanor to be making a restatement, and our rhetorical relation assignment may follow suit. Consequently, we take Eleanor to be saying the same thing at the two occasions, even though she uses different words.

IV.C. Rhetorical Relations in Poetry

We will now look at the work of rhetorical relations in poetry by considering some examples from Emily Dickinson. Let us first look at the first stanza from “How the Waters closed above Him”:

How the Waters closed above Him

We shall never know—

How He stretched His anguish to us

That—is covered too—

(Dickinson 1999, 400)

A rhetorical relation spontaneously assigned is Parallel between lines 2 and 4, as suggested by too in line 4. We must then have parallel elements in the two predications. One way to achieve this is by having we shall never know to ascribe the property of “we shall always ignore as a result of the water’s concealment of his body” and to let is covered, as it were, pick out the same property. Such a meaning adjustment includes certain contextual components in the meaning of the term while excluding other elements, such as any further disguising activity. If we now say that how he stretched his anguish to us is something which, like how the waters closed above him, we shall be ignorant about as a result of the water’s concealment, it is clear that this would represent a departure from the original wording and would not mean exactly the same thing as the original. This kind of paraphrase amounts to perhaps no more than a “school comprehension test” (White 1996, 201). But even so, it is a reformulation which is not arbitrarily imposed on the text, but which is demanded by the lines themselves, insofar as they are connected.

Let us next consider:

After great pain, a formal feeling comes—

The Nerves sit ceremonious, like Tombs—

The stiff Heart questions ‘was it He, that bore,’

And ‘Yesterday, or Centuries before’?

(Dickinson 1999, 170)

We may ask what a formal feeling in line 1 amounts to. Our taking lines 2 and 3 to be related to line 1 via Exemplification permits us to specify the quality of the feeling. Line 1 gives a general characterization of a situation which lines 2 and 3 provide specific instances of. If ceremoniousness and stiffness contribute to the feeling of line 1, we may likely understand the formality as involving tenseness. If we would qualify the feeling as not relaxed, we would certainly leave out certain aspects of the meaning of the line, but also capture other essential aspects of its meaning.

Consider now what going abroad amounts to in the following stanza:

The Soul that hath a Guest,

Doth seldom go abroad—

Diviner Crowd—at Home—

Obliterate the need—

(Dickinson 1999, 266)

It seems that go abroad in line 2 should be understood as involving the activity of socializing. This is because we relate lines 3 and 4 to lines 1 and 2 by Explanation. The reason for staying at home while having a guest is not the obligation to take care of the guest, but because the need of going abroad is satisfied. Accordingly, we read lines 1 and 2 as conveying that socializing by having a guest obliterates the need of socializing by going abroad. Lines 3 and 4 provide a reason for not going abroad in this specific sense while leaving open for the need of going abroad in other senses not obliterated by having a guest. Explanation thus requires us to take having a guest and going abroad as both being concerned with socializing. Additionally, we understand the need to stand for the property “the need to go abroad in order to socialize.”

Consider next smiling in this context:

My Soul—accused Me—And I quailed—

As Tongues of Diamond had reviled

All Else accused Me—and I smiled—

My soul—that Morning—was My friend—

(Dickinson 1999, 353)

We may know what it is to smile, but here, we must know what smiling implies as opposed to quailing. Lines 1 and 3 are related by Contrast, so that the acts of the soul’s accusing the poet and of all else’s accusing her respectively have contrasting effects. Perhaps quailing implies being discouraged and smiling implies being encouraged. There are alternative ways of showing how Contrast obtains. Quailing could imply a kind of threat and smiling merely an absence of threat. If we would say that all else accused the narrator—and she did not feel threatened, we would, it seems to me, unpack her heart in a way which is not the coroner’s concern.

Consider last this poem on the importance of selecting the right word:

Shall I take thee, the Poet said

To the propounded word?

Be stationed with the Candidates

Till I have finer tried—

The Poet searched Philology

And was about to ring

For the suspended Candidate

There came unsummoned in—

That portion of the Vision

The Word applied to fill

Not unto nomination

The Cherubim reveal—

(Dickinson 1999, 488)

The poem pictures the poet’s search for the right word, fit to evoke her vision. A candidate word is asked to wait until it has been established that there is no better word. When the poet has made up her mind that the candidate will do and is about to call upon it, the vision which the word was supposed to evoke unexpectedly presents itself. Lines 6–7 could be connected to lines 8–10 by Narration (a sequence of temporally contiguous distinct events) or by Occasion (one event being the enabling cause of another event). In any case, the lines are also related by Contrast (or, more specifically, Violated Expectation): at the moment the poet is about to ask the candidate word to come, it is the vision which comes. This has implications for our understanding of the terms ring for and come in unsummoned: we should think that the word is about to be summoned whereas the vision is not rung for, being in this sense unsummoned. In another sense though—ironically—the vision is hardly unsummoned. The job of the candidate word was after all to summon or evoke it. The vision appears as an intruder—the parataxis highlights the disarray—when all the philological anxiety was about making it liable to appear.

If this narrative serves as the exemplification of the moral of the last two lines, it should be noted that nomination must be understood equivocally. On one hand, setting out to find the mot juste is motivated by the idea that visions must be nominated or given name to in order to reveal themselves. On the other hand—ironically again—the revelation here occurs not thanks to any such nomination, but only in some conjunction or other with an about to be appointment of the candidate word. Whether we relate lines 6–7 and 8–10 by Narration or Occasion depends on whether we take the revelation to be an indirect reward for the poet’s effort or as wholly gratuitous. In any case, does not the poem speak for the possibility of paraphrase? The vision appears in the midst of alternatives and approximation, the right word was not paramount after all.¹⁵

V. CONCLUSION

Paraphrase, on the conception which I have been concerned with, does not render the whole meaning or the full effect of the original formulation, nor is it meant to replace it. It lays only claim to bringing out an implicit and essential aspect of the meaning of the original formulation as it appears in the text. It is my contention that this explicitation of meaning cannot be rejected as impossible or irrelevant, insofar as it is required by the very constitution of the text. This vindication elaborates on Kivy’s, insofar as his ideas concerning an unambitious propositional paraphrase are substantiated by (hopefully) less trivial examples and motivated not by truth conditions but rhetorical relations. It elaborates on Hills’s, insofar as this kind of reformulation originates with the necessity of cohesion; whereas, his vindication rested on the claim that the author engaged in paraphrase. To show that paraphrases like “Indeed it was a noble horse” or “What is just now occupying my attention is the comparative disadvantages of continuing to live or putting an end to myself” are not inappropriate or uninteresting, neither arbitrary associations nor the coroner’s concern only, it suffices to show how they result from the requirements of rhetorical relations.¹⁶

If we associate paraphrase with the kind of interpretive activity which our readings have exemplified, there is no reason to consider paraphrase as heretical. There is certainly a mutual dependence between form and content, especially perhaps in literary texts, but paraphrase need not do violence to the complexity of the original. There are alternative formulations which represent an adequate response to the original words. Reformulation may be in the service of fine-grained attention to the entanglements of unique content/form unities (compare Hulatt 2016, 58), even for those who believe that literary interpretation is foremost a matter of enhancement of aesthetic experience.¹⁷ Attending to subject-realized-in-just-this-way does not exclude careful explicitation of meaning. In contrast to the supposed irreplaceability of the original word, I have sought to display the solidarity of the original word with alternative wordings. A certain distinction between content and form has its place in the reading of any text, in so far as rhetorical relations having semantic impact are applied and linguistic meaning is underdetermined with respect to them.

Leighton says that paraphrase is “in the very nature of reading” (2009, 172). Rhetorical relations and their requirements permit us to tell more specifically how it is so. Perhaps this kind of paraphrase, pretending neither to exhaust the original formulation nor to dispense with it, but demanded by the intelligibility of the text, is not objectionable to anyone. But then, there was no reason to associate the notion of paraphrase with anything else and most of the traditional statements concerning the impossibility of paraphrase should somehow be restated. Most of what is interestingly said about literature is certainly not at the level of paraphrase, but there seems to be no reason for singling out what is interesting about literature by means of the impossibility of paraphrase.¹⁸

References

Footnotes