Back to home page
Back to "Occasional Papers"

Reuven Tsur

Translation Theory, Cultural Context and Intertextuality
Some Methods and Constraints in Translation Studies

The ensuing discussion is focussed on an attempt to explore some limits of legitimacy in translation studies. Trivial studies meticulously observe these limits; being non-trivial entails some adventurous thinking which, in turn, not infrequently ends up in transgressing these limits. Eduardo Crisafulli's paper (1993) offers some such adventurous thinking. The present paper assumes that poetry translation is an art; and that a work of art presents an "elegant solution to a problem". Translation studies have a unique status among other kinds of literary study in an important respect: with reference to a translated poem, much more can be known about how the problem to be solved arose than with reference to an original poem. The author of an original poem may start out with whatever he wishes; any word, any prosodic structure, any idea contents, any figure of speech, any emotion or attitude, and he may mold all this and much more into any kind of poem. The translator of a poem begins with a source poem, in which very much of the data are very well defined; and one may quite safely assume that unless otherwise indicated, the translator's intention has been to offer in the target language a poem that is as close to the source poem as his verbal material would allow. This process is regulated, among other things, by the translator's poetics and norms of translation; the former, in turn, may be inferred from his original and translated poetry; the latter from his general practice in his translations and possibly also from the practice of the group of translators to which he belongs. I, therefore, have a strong inclination to accept Crisafulli's assumption "that a translator's choices are meaningful and can be accounted for. Such an assertion presupposes that the translator's choices show internal consistency" (ibid., 194). And if I disagree with him, it is only because I object to the way in which he puts this assumption into practice.

Crisafulli's article compares two English translations of a few lines from Canto XXXIII of Dante's La Divina Commedia. The translations are those of Charles Singleton (1970) and Seamus Heaney (1979). Singleton's version is part of his complete prose translation of Dante's poem, whereas Seamus Heaney's translation only covers lines 1-90 of Canto XXXIII and is found in the poem "Ugolino" in the collection Field Work.1

I shall distinguish (following such analytic philosophers as Beardsley, Weitz and Margolis) between two kinds of critical activities: description and interpretation (or explanation). Description consists in pointing out the elements in a text and the relationships between them. Interpretation points out possible relationships between the text and some context outside it. In the present work we shall be concerned with three kinds of such external contexts: the world represented in Dante's text, and what Crisafulli calls "the Irish situational context"; toward the end of this paper I shall try a third context: the translator's mind. Though, as Morris Weitz would argue, even some of the "facts" of a text are attributed ones, one may perform the description of a text with considerable confidence and accuracy. Interpretation, on the other hand, has the logical status of a hypothesis; and the constraints to be observed in the application of hypotheses should apply to interpretation too. Otherwise we get a scholarly situation in which "anything goes". It will be in strict conformity with this statement that before an interpretation may be ventured, a huge amount of descriptive analysis must be performed. Concerning descriptive statements we should insist on an additional distinction: the truth (or accuracy) of the descriptive statement in light of the evidence available in the text; and the possible significance of such a description. I shall argue that Crisafulli does not adhere rigorously enough to the constraints to be observed in the application of hypotheses, does not perform a sufficient amount of descriptive analysis before proceeding to interpretation, and even where his descriptive analysis is accurate, he sometimes attributes unwarranted significance to the elements described.

Crisafulli sets out to show "that the Irish situational context explains why Heaney's version differs from Singleton's. This hypothesis favours a theory of translation which takes into account cultural elements" (ibid, 196-197). As I said earlier, I accept Crisafulli's assumption that a translator's choices are meaningful and can be accounted for, but I object to the way in which he puts this assumption into practice. In fact, I shall argue that when relating Heaney's translation choices to the Irish situational context, Crisafulli takes for granted things without sufficient basis or justification; at the same time, he also relies on information that is not sufficiently unique to allow his argument exceed triviality (this would be triviality in a different sense from that mentioned in the first paragraph). Consider, for instance, the phrase "the spiritual strength of the Roman Catholic community, typified by Ugolino's resistance in the tower" (ibid., 200). "The Irish situational context" is complex enough to warrant the attribution of spiritual strength to the Roman Catholic community, or desperateness, or senseless brutality, or a wide range of other possible attitudes according to the needs of the argument. Thus, Crisafulli offers here a hypothesis to relate between two unrelated sets of data; but one of these two sets at least is merely assumed by him for the sake of the hypothesis: that the spiritual strength of the Roman Catholic community is more relevant to the Irish situational context than, e.g., its desperateness. Thus, whatever may have happened to Ugolino in the tower, the Irish situational context may offer some feature that would explain it. At the same time, "the spiritual strength" could be attributed to almost any community in the world, if there is no specification of what specific kind of "spiritual strength" one has in mind. As I shall argue later, concerning "Ugolino's resistance in the tower" too Crisafulli takes for granted more than is warranted by his data. But even if both assumptions would be warranted, Crisafulli gives no evidence whatever to justify his use of "typified by". It would appear that he is offering here an assumption of that which he set out to prove in the first place.

To be sure, literary interpretion may abstract, quite legitimately, from its texts such general abstractions as "spiritual strength", even if not mentioned explicitly; here the case merely seems to be that the present text does not sufficiently warrant it. What is much less permissible in literary interpretation is the attribution of (unmentioned) specific details to the context, unless there are very very good reasons for it. Some of Crisafulli's arguments concern precisely such unmentioned specific details: "it may be that Heaney does not believe that Protestants are responsible for the conflict with the Roman Catholic community. In the same way, the spiritual strength of the Roman Catholic community, typified by Ugolino's resistance in the tower, is no justification for betrayals of any kind" (ibid., 200).

There are two cardinal points which I find very problematic in his thesis. First, very little has been said about the "Irish context" that would apply more readily to an Irish translator of Dante than to the translator of any other poetry, of any other nationality. Second, I agree with the assumption that suggestions of the translator's general cultural context should be looked for in those aspects of the translation which cannot be accounted for by a linguistic analysis of the source text and the target text. However, in Heaney's instance, the "cultural context" analysis must be preceded by an additional stage: an analysis of the constraints of versification.

It is difficult to tell how to handle this stage systematically, since there are ten thousand solutions for any translation problem. My own solution to the general problem (which I would, certainly, not force upon anybody) would be something like this. Everything that enters the target text: idea contents, figurative language, metre, rhyme, the rules of syntax, word order, stylistic norms, etc., presents the translator with a problem, for which he must offer what some theorists following Dewey called "an elegant solution". "An elegant solution" would be one that satisfies to a maximum degree all the afore-said norms, violating them as little as possible. When one of them must be violated for some reason (e.g., word order, in order to manipulate a word into the rhyme, or to satisfy the requirements of metre, or both) the violation must be justifiable by some other norm(s) (e.g., by the requirements of thematic emphasis, or figurative language, or parallelism, or the specific norms of, e.g., the poet's modernistic poetics). This would hold true of all competent poetry. As far as translation is concerned, there is one more very important constraint: the requirement to convey as much as possible of the source-text's various aspects. Accordingly, the "elegant solution" in a translated poem would result in the "overdetermination" of every word in the poetic text.

Thus, for instance, in his example 2, Crisafulli points out that the archbishop who betrayed Ugolino is called "Ruggeri" in Dante's text, as well as in Singleton's translation; in Heaney's he is called "Roger" (ibid., 196). "As an Irish poet, Heaney uses 'Roger' because this name has associations which are meaningful to an Irish audience: it is a Protestant rather than a Catholic name. And since 'Ruggieri' is an archbishop there are additional 'religious' overtones. I suggest, then, that the Irish situational context explains why Heaney's version differs from Singleton's" (ibid., 196). This is Crisafulli's shortest, but perhaps most convincing example. Still, very little can be gained, I think, attributing a Protestant name to Ugolino's bishop. And, at least, one should also take into account the prosodic fact that Roger (unlike Ruggeri) is a bisyllabic word with the stress on the penultimate syllable. Before ascribing the recourse to this name to some idea related to Protestantism, one should also check, to what use is put the extra space gained by using the foreshortened form of the private name, which is admittedly odd here.

I can say more specific things about Crisafulli's example 3. Here the Italian text with its two translations runs as follows:

poscia, più che il dolor, poté il digiuno. (Canto XXXIII. 75)

Then fasting did more than grief had done. (Singleton, 353)

Then hunger killed where grief had only wounded. (Heaney, 63)

The annotators say that the Italian original (as well as Singleton's version) has two possible meanings: that Ugolino died of hunger rather than of grief; or that Ugolino fed upon his sons to prolong his own life. This line is not ambiguous as Crisafulli would say, but rather vague: it renders the reality represented in the poem ontologically incomplete. The annotations supply two hypothetical extralinguistic contexts to complete this reality. Later we shall examine what evidence can be found in favour of these interpretations.

At any rate, Crisafulli says that in Heaney's version all suggestion that Ugolino fed on his children is eliminated. What can be said in favour of Heaney's "elegant solution" to his translation problem? First, that Singleton's prose translation sounds extremely dull with its repeated pro-verb (did, done), which conveys absolutely no specific semantic information. To avoid this, Heaney may have had recourse to ten thousand different kinds of solutions. We do not know what the other nine thousand nine hundred ninety nine solutions were. However, we can tell some of the merits of this one solution: first, its pair of verbs loads the line with a heavy load of semantic information. By the same token, it introduces a parallelism between the two verbs, which points up both their common and their opposing semantic ingredients. All this makes the verse line semantically more powerful.

Now how should we account for hunger in Heaney's version as opposed to Singleton's fasting? From the phonetic point of view, Heaney establishes here an exceptionally sonorous sound pattern, by introducing the sonorants l, n, r, w wherever he can, and hunger is more sonorous than fasting; that is why he introduces the adverb only, and the verbs killed and wounded too contribute to this sonority. Prosodically, the verse line conforms with the iambic pentameter pattern, and ends with a (rhyme) word whose penultimate syllable is stressed; this would conform with one of the basic prosodic features of the source text, i.e., that all its rhymes are "feminine" endings (this is quite meaningless in the present instance, since from the other quotations one may gather that Heaney relinquished in his translation all claims for this prosodic feature). Some such analysis should, I believe, precede any attempt to look for explanations in the "Irish context".

Now what does Crisafulli find in the Irish context? "I suggest that Heaney endows Ugolino with spiritual strength by obliterating any hint that he might have fed on the bodies of his children in a desperate attempt to survive starvation. / This example is consistent with the hypothesis that the Irish context is essential to an understanding of Heaney's text". Why should the Irish context require the suppression of any hint that Ugolino might have fed on the bodies of his children in a desperate attempt to survive starvation? On the contrary, rather. Irish history is full with masses starving to death, and desperate attempts to survive starvation. Nothing can serve as a stronger poetic expression of such desperateness, than a father feeding on his own children's bodies. At least in one author's case, precisely the afore-said "Irish context" made Swift propose his "A MODEST PROPOSAL FOR PREVENTING THE CHILDREN OF POOR PEOPLE IN IRELAND FROM BEING A BURTHEN TO THEIR PARENTS OR COUNTRY, AND FOR MAKING THEM BENEFICIAL TO THE PUBLIC" the essence of which being to cook and eat them. Had Heaney suppressed the "hunger killed" meaning and foregrounded the cannibal meaning, Crisafulli would have had as good a case, or even a better one, to suggest that Heaney's translation reflects the Irish context. And he could even invoke intertextuality, appealing to Swift. Intertextuality is admirably suited to help us prove whatever we wish. But on the whole, I believe that the immediate impact of Heaney's version, as revealed by, e.g., our foregoing semantic, prosodic and phonetic analysis, should satisfy us in both cases, and obviate the need to invoke the Irish context.

As we have seen, the annotator's suggest two alternative completions for the extralinguistic context of this verse line. Which one is more plausible? If the cannibalism interpretation is more plausible, then we would be justified in looking for reasons for Heaney's preferring the other one. If, however, the hunger-rather-than-grief-killed interpretation is more plausible, then Heaney would have made exactly the choice suggested by the source text, and we ought not to look for special reasons for his preference.

I have also consulted Babits's full Hungarian translation of the Divine Comedy (which is a masterpiece of translation in its own right, with extensive commentary on the text). Semantically, this line in his translation is somewhere between the two English translations, nearer to Singleton's. It runs as follows:

Míg többre ment az éhség mint a bánat!

In a literal translation this means something like "While hunger came (amounted) to more than grief". Still, this Hungarian version is much more powerful than Singleton's, owing to the organisation of its sound patterns, resulting in the overdetermination of its words. First, it conforms with the iambic pentameter pattern (and also with the penultimate-stressed rhyme pattern); and second, it is infused with an exceptionally intense and sonorous pattern of consonants. The same consonants occur in the same order in the following three words: ment... mint... bánat if we recall Kenneth Burke's notion of "colliteration", that is, that the bilabial [m] is replaced by the bilabial [b], rendering the sound pattern active but less obtrusive. In his commentary, Babits makes it quite clear that, irrespective of in what other ways this line could be construed, he understands it according to the hunger-rather-than-grief-killed interpretation. Now in light of Crisafulli's analysis we must come to one of two conclusions: either that the Hungarian situational context is as good as the Irish situational context to account for the same translation decision, or that Crisafulli's account of this line is somewhat questionable.

Now it should be noted that interpretations are usually not offered to isolated verse lines, but to whole poems, or at least to wider situations in longer poems. There is nothing in this isolated line to make the interpreter think of cannibalism -- it is suggested by the wider context. Ugolino's kids see him helplessly bite his own fists out of helplessness and desperateness (rather than of spiritual strength). While this is what we might term an expressive gesture, they misread it as eating his own body, and offer him to feed on their body. I suggest that some translators may be reluctant to adopt the cannibalism interpretation for nonaesthetic reasons: the moral shock in front of the possibility that Ugolino fed on the flesh of his own children; but also for some aesthetic reasons: the pathetic affect of this scene in which the kids offer their own flesh to their father might be impaired by the more brutal interpretation.

There are two more marginal pieces of evidence in favour of this interpretation. In line 23 Ugolino indicates that the tower in which they died was named after their ordeal "the tower of hunger" (rather than "of cannibalism"), and that Dante placed Ugolino in that part of the Inferno in which traitors (not cannibals) are punished. The notion of cannibalism in this context could be outrageous enough to the sensibilities of Dante or Ugolino's contemporaries so as to override hunger or treason in choosing their focus of emphasis. Thus, Crisafulli is perfectly right in pointing out on the descriptive level that in Heaney's version all suggestion that Ugolino fed on his children is eliminated. This, however, need not indicate, on the interpretation level, his spiritual strength or the influence of the Irish situational context on the translator's decision.

Consider now example 5:

Ché se 'l conte Ugolino aveva voce
d'aver tradita te delle castella
non dovei tu i figliuori porre tal croce. (Dante, Canto XXXIII, 85-87)

For if Count Ugolino had the name of betraying you of your castles,
you ought not to have put his children to such torture. (Singleton)

                     For the sins
Of Ugolino, who betrayed your forts,
Should never have been visited on his sons.(Heaney)

Crisafulli's discussion of this example is really a gross misreading of the situation. Singleton's "had the name of betraying you of your castles" would sound extremely awkward in any poetic translation. If I were to translate this passage into Hebrew or Hungarian (the two languages into which I have ever translated poetry) I would have, very likely, chosen a solution similar to Heaney's, and so would, doubtless, many translators I can think of, without having anything to do with the "Irish context". Crisafulli's suggestion that "Heaney reaches beyond an exclusive Roman Catholic audience by ascribing the betrayal of the 'forts' to his heroic figure" is based on a hypothesis hardly warranted by the text or by what we "know" about the Irish context, if all the evidence for it is this slight linguistic evidence (unless our position is that anything can be adapted to it). Simply, the main tenor of the passage focuses on the monstrous injustice committed against the children, and insisting on the nuance of "having the name of having betrayed" rather than "having betrayed" would not only render the passage stylistically cumbersome, but would also distract attention from this injustice.2

Consider in this context the following speculation of Crisafulli's:
Heaney's own poetry is concerned with the "typical strains which the consciousness labours under" in Ireland. One of these strains corresponds to a religious dimension which also pervades Dante's work. More precisely, Heaney says that his own conscience is torn between "the claims of orthodoxy and the necessity of refusing those claims". I suggest that this opposition is reflected in Heaney's selection of lexical items for instance in the foregrounding of "sins" in example 5. Since the concept of sins is closely connected with the idea of individual responsibility, it contrasts with the notion that history is responsible for the bad deeds of individuals.

There is nothing easier than to quote a poet on his own poetry. But it is dangerous too, if one does not have the necessary safeguards against falling into the trap of one's own making. Crisafulli is perfectly right on the descriptive level in saying that Heaney's version "nominalizes the event and the 'sins' become the subject of the clause which is foregrounded and thematised" (ibid., 199). But what is it he has proved by this? On the interpretative level there is very little in the word "sin" to justify the efforts to relate Heaney's version of Dante to his own musing on orthodoxy. This is a plain instance of making an accurate description and attributing an unwarranted significance.

On the other hand, it would appear that Crisafulli overlooked one important ingredient of the "Irish context", the English Bible, and the intertextual processes arising from a comparison of Heaney's version of Dante to the first of the Ten Commandments: "for I the Lord your God am a jealous God, visiting the iniquity of the fathers upon the children to the third and fourth generation" (Exodus, 20: 5; the prosodic reasons for the replacement of iniquity and children by sins and sons are obvious). This is, precisely, of what Dante accuses Pisa in his invocation in the case of Ugolino and his sons, irrespective of whether the accusation that Ugolino betrayed the fortresses is based on fact, or only on hearsay. Curiously enough, even Singleton in his prose translation simplifies the text for some reason. Where he writes "you ought not to have put his children to such torture", he substitutes the abstract torture for Dante's own visual image "porre a tal croce". The cross is obviously a metonymy of torture, with an allusion to Christ's sufferings. Why should Singleton eliminate this visual image and Biblical allusion where there are no prosodic constraints to compel him to do so?3   I think, Dante's expression here is too roundabout, quite difficult to follow; Singleton put his argument more bluntly, sparing the reader some mental effort, resulting in a more straightforward impact. Heaney's solution, by contrast, both achieves the straightforward impact and resorts to a (different) Biblical allusion. This use of the Biblical phrasing renders Dante's accusation exceptionally forceful; and it needs no special justification, since the Ugolino tragedy is, really, a particular instance of the Biblical threat (only that here the sin of the father is visited upon the sons not by God, but by flesh-and-blood agents, which renders Dante's indignation even more justified).

Perhaps, two prosodic observations too might illuminate the translator's decision here. First, the use of the Biblical allusion affords a brilliant off-rhyme: "sins" -- "sons". From the little we know on Heaney's poetics from Crisafulli's paper, this may be one of his favored prosodic devices, as we find such off-rhymes in his original poetry as well, as "bits" -- "nights", "rent" -- "tents", and "country" -- "sorry" on page 201. Second, in the iambic pentameter line, syntactic breaks are most natural when they occur at the line ending; syntactic breaks in mid-line, after positions 4, 5, or 6 are more tense, but still natural; syntactic breaks between the sixth and tenth positions are increasingly tense; when a run-on sentence begins in position 8, 9, or 10, it displays increasing impetus. In example 5, Dante's phrase begins at the beginning of the line, minimising tension as much as possible. In Heaney's version, the same sentence is much shorter (hence, perhaps one possible need to avoid "has the name of having betrayed"), and begins in position 7 of the line, resulting in a run-on line that generates more than usual tension: "For the sins / Of Ugolino, who betrayed your fort, / Should never have been visited on his sons". One apparent source of the exceptional vigour of Heaney's version seems to be precisely this prosodic device. At any rate, Crisafulli's paper offers an additional instance, in example 1: "I would sow / my words like curses that they might increase...". Notice the sentence beginning in position 7, and the emphatic syntactic break in the next line, after position 5.

Let us turn now to Crisafulli's example 4.

Ahi Pisa, vituperio delle genti
del bel paese là dove 'l si sona (Dante, Canto XXXIII, 79-80)

Ah, Pisa! shame on the people of the fair land where the si' is heard ... (Singleton)

Pisa! Pisa, your sounds are like a hiss
Sizzling in our country's grassy language.(Heaney)

At first glance it would appear that Heaney's translation broke here loose of the source text completely. Only the words Pisa, sona, and paese are translated with some accuracy into English; and in the target text, the words hiss, sizzling, and grassy seem to have no equivalent in the source text. Consequently, the contents too is very different in the two texts. On the thematic level, however, from the emotional and rhetorical point of view, both are a harsh apostrophe to Pisa (and its people), followed by the reprimand expressed in example 5 (discussed above), concerning the unjustifiable torture of the innocent children. The sequel "For the sins / Of Ugolino, who betrayed your forts, / Should never have been visited on his sons" seems to wholly justify this harsh apostrophe, and requires no further search for reasons. What Crisafulli does here, instead, appears to be quite typical of his handling Heaney's translation. He breaks the text up into discontinuous fragments, and inserts a longish passage conveying some historical information concerning political strife in Dante's Italy, beginning with "Dante is condemning the people of Pisa just as he denounces the lack of vision of ambitious contemporary political leaders and of clergymen who did not hesitate to stir up civil wars within the small communities in the pursuit of personal interest" (ibid., 197-198), and culminating in the sentence "Dante portrays the tragic condition of personalities caught in a web of political struggles and specious religious conflicts, which bring discord and division in society" (ibid., 198). For all we know, these musings are not inconsistent with the Irish situation either. The only reason why he ought to have spared us from it is this: if we read lines 79-90 in one continuum, there is no reason whatever to doubt that it was the cruel fate of Ugolino's kids that prompted these hard words of Dante's.

Further on, Crisafulli construes these two lines in Heaney's translation in a way that refers to two different languages spoken in one region. "I suggest that in order to adapt Dante's denouncement to the Irish situation Heaney here shapes his translation so that it carries an implied reference to the clash between two entirely different linguistic and cultural traditions: the Gaelic and the English" (ibid., 198). In isolation, with some effort, we can indeed construe these verse lines as if the speaker alluded to two different languages. These English verse lines are rather vague, and some effort of construal is required in any case; but the two-languages construal is not the most obvious one. One quite obvious paraphrase of these lines would run thus: "Pisa! Pisa, your sounds are sizzling in our country's grassy language like a hiss". This, of course, may not prevent us from construing these lines as "your sounds are like our country's grassy language". Furthermore, it is not clear whether it is the sibilant in the name Pisa (did not Heaney redouble the vocative, instead of using an interjection plus a vocative?), or the sounds of the language spoken in Pisa that "hiss sizzling". Of all these construals the two-languages construal appears to be the least probable, though not altogether improbable. There are, then, at least three possible construals of this sentence, and it is Crisafulli rather than Heaney who "here shapes the translation so that it carries an implied reference to the clash between two entirely different linguistic and cultural traditions: the Gaelic and the English".

A scrupulous scrutiny of the source text and the target text may reveal a piece of the translator's mind, indicating "that a translator's choices are meaningful and can be accounted for". The phrase "the fair land where the si' is heard" alludes to the contemporary practice of distinguishing between the various vernacular Romance languages by their word for "yes": "langue d'oc" for Provençal, "langue d'oil" for French; Italian was marked by "si". The Randomhouse College Dictionary offers the following two definitions for the verb hiss: "to make or emit a sharp sound like that of the letter s when prolonged", and "to express disapproval or contempt by making this sound". A comparison between the source and target text reveals that the first meaning may refer to the sound of the word si (reinforced by sizzling); the second may capture the notion of "vituperation". From this point on, the translator seems to have been concerned with packing as many sibilants as possible in these two lines: sounds was introduced as an English equivalent of sona; Pisa was reduplicated; sizzling and grassy were fortuitously introduced. At this point, Heaney seems to have been bothered very little whether the sounds of the name Pisa, or of the word si spoken in Pisa, or both, are "Sizzling in our country's grassy language like a hiss". It is here where our distinction between description and interpretation comes useful. There is no linguistic reason to prevent us from the "two-languages" construal of these lines in isolation except that they will make no sense in their wider context. Furthermore, vague sentences are vague by definition: they lack constraints to eliminate unwanted construals; this elimination ought to be done by interpretation. If we use the interpretation of Dante's co-text as such a constraint, the two-languages reading must be eliminated; if we apply an interpretation based on the Irish situation to allow this reading, we are assuming that which in the beginning was set forth to be proved.

Let us step now one step back and have a look at my treatment of example 4. On the descriptive level, I have pointed out that from the emotional point of view, both the source text and the target text display similar emotional tendencies, while from the rhetorical point of view they both serve as a harsh apostrophe to Pisa; at the same time, from the point of view of explicit contents, I have pointed out considerable discrepancies between the two texts. From the phonetic point of view, I pointed out the intensive sound pattern generated by the sibilants; I have also pointed out that some of these sibilants are planted into the target text by introducing words that are fortuitous to some degree from the source text's point of view: the reduplication of Pisa, and the gratuitous introduction of hiss, sizzling, and grassy. From the semantic point of view, I have pointed out the ambiguity of the word hiss, accommodating, in spite of all, two words from the source text: si and vituperio. From the syntactic point of view, I have pointed out three possible construals of the message, one of them (though not the most plausible one) being the two-languages construal. On the interpretive level, I have offered a parsimonious hypothesis, in an attempt to account for processes in the translator's mind in the course of achieving an "elegant" and powerful solution of his translation problems, starting with the accommodation of two words from the source text in hiss. I don't claim that things actually took place in this order; only that such an hypothesis may explain the translator's conception that led to his end-product, assuming "that a translator's choices are meaningful and can be accounted for". Furthermore, "such an assertion presupposes that the translator's choices show internal consistency": even the few short passages presented from Heaney's translation indicate a strong inclination to pack his lines with rich semantic information and intensive phonetic patterns.

Crisafulli invokes an essay by Heaney in which he clarifies how he relates to Dante's world. He says that he is struck by two features in Dante, which also inspire his own work:

1.The celebration of bonds of friendship and bonds of enmity.
2.The way in which Dante could place himself in a historical world and yet submit that world to scrutiny to a perspective beyond history; the way he could accommodate the political and the transcendent (ibid., 200).

Here again we come up against the issue of what Heaney's statement says, and what its significance is for the present argument. Insofar as Crisafulli claims that this may explain why Heaney has chosen to translate Dante in general and the Ugolino episode in particular, we may agree with him. But insofar as he implies that this may also show that the Irish situational context had an influence on the translator's choices, such a significance has not yet been demonstrated in a convincing manner.

Finally, I wish to make a comment on intertextuality. There are very few notions in present day literary theory that are so relentlessly abused as the notion of intertextuality. With the help of this notion one can "prove" anything. My recourse to Swift's "Modest Proposal" was originally ironical. But then I realised that this recourse could make out a better case for the relevance of the Irish context to the cannibalism-meaning of Dante's line than reliance on the rival meaning adopted by Crisafulli. The allusion to the Bible in example 5, by contrast, appears to be indispensable for a proper appreciation of the lines.

At the end of his paper, Crisafulli resorts to intertextuality to complete his argument. He proposes to read the Ugolino poem in the context of Heaney's original poetry. He points out that the poem immediately preceding the Ugolino poem ends with the word "underground". What can we learn from this about Dante's poem? At best, it merely reinforces what we already know, namely, that the Inferno's world is the "underworld"; at worst, it is quite irrelevant. Likewise, Crisafulli says, "In the poem 'Casualty' from the same collection we meet with cruelty and destruction" (and there follows a quotation). Now I don't believe that any amount of surrounding poetry can render Dante's description more cruel than it is unless the implication is that from now on we may read into Dante's poem anything we like. In the immediately preceding poem, on the other hand, we find the following sentence: "I am sorry / That party politics should divide our tents". With some good will, this sentence can be made relevant to the relationship between Ugolino and Ruggieri. Such intertextuality could help the reader to direct his attention away from the monstrous events narrated in the poem to what he may know about the former alliance between Ugolino and Ruggieri.

It should be noted that in these cases very little light is thrown upon Dante's poem by the translator's own poetry; and in no case did this intertextuality affect the translator's choice, to justify as it were Crisafulli's claim that "the relationship between the source and the target text is complex and must take into account the translator as a person whose values and culture affect and shape his products" (ibid., 202). Nonetheless, it would be quite unreasonable to argue that there is no interaction between the Ugolino poem and the other poems in Heaney's volume; only that this interaction seems to work, precisely, the other way around. Rather than looking for the ways in which the other poems affect the Ugolino poem, one should look for the ways Dante's poem affects the collection in which it is included. It is well possible that the world of Heaney's volume is a world of, e.g., "cruelty and destruction" that may be abstracted from a variety of poems in the collection."Ugolino" may serve as one more poem from which these qualities can be abstracted.

In principle, the intertextual influence may go in both directions. Here we should explicitly put the following theoretical question: How could, possibly, Heaney's original poems influence the reader's appreciation of Dante's poem included in his collection? The best answer I can think of is this: suppose there were certain salient themes and qualities in Heaney's original poems; suppose, again, that the same themes and qualities occurred in Dante's poem at a low level of salience; in such circumstances, the high-salient elements in Heaney's original poems would reinforce the low-salient elements in Dante's poem and promote them on the scale of salience. It could well be the case that the other poems promote in this poem some meaning that is low in the hierarchy of meanings. It merely happens to be the case that Crisafulli provides no such instances in his article.


1. I do not discuss the Singleton translation, or Heaney's poetry and his Dante translation, beyond what is written in Crisafulli's article. Thus, in the poem "Ugolino" there may be quite valuable information to support some of Crisafulli's claims; but since he did not bother to bring it to his readers' attention, I must ignore it. This may do some injustice to Crisafulli, but does not affect my methodological argument. [back]

2. Babits indicates the doubts insinuated by "having the reputation of" by using the rejected conditional:

Bár Ugolino vesztedet kereste
S minden várad eladta volna tényleg,
Nem vonhatnád fiait ily keresztre.

(Had Ugolino sought your destruction,
and really sold all your fortresses,
you ought not to have drawn his sons upon such a cross). [back]

3. This could be an excellent justification for looking into Singleton's "situational context" for some explanation for this deviation. [back]

Work Cited
Crisafulli, Edoardo. 1993. Culture and Text: Equivalence Revisited. Perspectives: Studies in Translatology 2, 194-206.

Back to home page
Back to "Occasional Papers"

Original file name: Cay - Dante 3 - converted on Tuesday, 10 March 1998, 12:41

This page was created using TextToHTML. TextToHTML is a free software for Macintosh and is (c) 1995,1996 by Kris Coppieters