Causal Subordination in English and Norwegian

Causal relations are central to the way humans perceive and interpret the interplay between states of affairs in the everyday world. This is revealed in the multitude of forms the expression of a causal relationship can take in different languages. The purpose of the present article, which is based on my MA thesis (see Meier 2001), is to compare and contrast causal linkage in a corpus material of original and translated English and Norwegian texts. The main focus is on the various correspondences of English causal subordinators in Norwegian, and the correspondences of Norwegian causal subordinators in English; in other words, what happens when the causal links are translated from one language into the other? More specifically, I aim to answer the following questions:


Aim
Causal relations are central to the way humans perceive and interpret the interplay between states of affairs in the everyday world.This is revealed in the multitude of forms the expression of a causal relationship can take in different languages.
The purpose of the present article, which is based on my MA thesis (see Meier 2001), is to compare and contrast causal linkage in a corpus material of original and translated English and Norwegian texts.The main focus is on the various correspondences of English causal subordinators in Norwegian, and the correspondences of Norwegian causal subordinators in English; in other words, what happens when the causal links are translated from one language into the other?More specifically, I aim to answer the following questions: • How is the causal link represented in translation?
• To what extent is the syntactic construction and causal ordering of the original text retained in translation, and what contextual factors intervene to bring about changes?
• To what extent do the causal links differ with respect to the sequencing of causal members, information structure, and their ability to express indirect reason?

Material and method
In order to allow for a proper treatment of the examples analysed, the scope of the study has been restricted to cases where the causal relation is explicitly realised by a subordinator.More specifically, I will compare the English causal subordinators because, since and as with the Norwegian causal subordinators fordi, siden and ettersom.
Other possible forms of causal linkage will be discussed to the extent that they appear as correspondences of the causal subordinators in the translated texts.All examples with since (115), as (56), siden (44) and ettersom (12) found in the corpus original texts have been included in the study.Of the 547 because-clauses and 423 fordi-clames found, a representative subset of 100 examples of each type has been included.Regrettably, the Norwegian subordinator da is not included in the study.As will be shown later, it proved to be much more frequent as a correspondence of the English subordinators than was originally believed.
The examples used in this study are taken from the English-Norwegian Parallel Corpus (ENPC) (see Appendix and http://www.hf.uio.no/iba/prosjekt/ for more information).The ENPC is a 2.6-million-word computer corpus of English and Norwegian text samples with their translations to the other language.30 original fiction samples and 20 original non-fiction samples in English and Norwegian are included, each containing 10,000-15,000 words.The texts are aligned so that each orthographic sentence, or s-unit, in the original text is paired with the equivalent s-unit in the translated text.This allows for easy identification of correspondences across the languages.
Comprising both original and translated texts in English and Norwegian, the ENPC can be employed in several different types of studies.In addition to various types of translation studies, the corpus can be used for contrastive studies based on parallel original texts, as well as contrastive studies based on the original texts and their translations Qohansson and Ebeling 1996, 4).As a basis for contrastive studies, the corpus is a good source for establishing crosslinguistic equivalence, since it relies not so much on the analyst's subjective intuition as on the combined judgements of a range of language-users, as revealed by the choices made in translation.In this study of causal expressions I have compared English and Norwegian original texts with their corresponding translations.
The methodological approach taken in the study is that of contrastive analysis (CA).Briefly defined, CA can be said to be "a systematic comparison of the linguistic systems of two or more languages" (Ringbom 1997, 737).Traditionally a method used for predicting learner difficulties in second language acquisition, in areas where languages differ markedly (cf.James 1980), contrastive analysis has been increasingly used for more theoretical work and as a means of uncovering language-specific characteristics that are not as easily discovered in monolingual studies.
A central problem in CA is how to compare languages that might be divergent in ways we are yet unaware of.There is a need for some sort of cross-linguistic constant to ensure that the objects juxtaposed in the investigation are comparable.In this study, comparability is established on the basis of translation equivalence.Following Ebeling, we can say that "if two constructions [...] are the most frequent translations of each other, then they are the closest equivalents in the two languages in the text types represented in the corpus" (Ebeling 1999, 22).

Causal relations
As a working definition for this study, a causal relation will be said to exist between two states of affairs when the occurrence of one state of affairs, henceforth referred to as the cause, is perceived by the speaker/writer as leading to the occurrence of another state of affairs, henceforth referred to as the result.The cause and the result will both be referred to as members of the causal relationship.
As Altenberg (1984, 20) has shown, the En glish language allows for the members of the causal relation to be encoded as phrases, subordinate clauses, or as independent clauses.The following examples will serve to illustrate this:

3.1
Because of the rain, the leaves of the shrub are wet.

3.2
Because it has been raining, the leaves of the shrub are wet.

3.3
It has been raining.Therefore the leaves of the shrub are wet.
In the above examples, the causal relation has been explicitly marked, through a prepositional construction (3.1), a subordinator (3.2) or a cohesive adverb (3.3).Altenberg refers to this linguistic marker as a "causal link" (1984,21).In some cases, the causal link is integrated into the main clause structure itself (clause-integrated linkage), as in 3.4, where the causal link is realised as the subject in a copular construction.

3.4
Next, for her own good, Marie-Louise demanded that her daughter jump to the floor from a table, several times.The only result was exhaustion and despair in the victim.(RDA1) Verbs may also serve as causal links: 3.5 She caused his death.
Norwegian and English have very similar inventories of causal links available to the speaker/writer.Some differences do exist, however.In English, cohesive adverbials can only be used to link two causal members in cause-result order, as in example 3.3 above.By contrast, Norwegian has a cohesive adverb, nemlig, which allows independent clauses to be linked causally in result-cause order:

3.6
Bladene på busken er våte.Det har nemlig regnet.[The leaves of the shrub are wet.It has (nemlig) been raining.] A further formal difference between English and Norwegian concerns the cognate pairs so/sä and for/for.Whereas for and så are seen as coordinators in Norwegian, the English so, while proving notoriously difficult to classify, is most often seen as an adverb, and for is usually included among the subordinators (Quirk et cd. 1985, 928).
Although the causal relation may be signalled by a causal link, the causal relation in itself is a semantic relation between propositions, and needs not be overtly marked (Altenberg 1984, 21).In the following examples, the causal relation is implied through coordination (3.7) and juxtaposition (3.8):

3.7
It has been raining, and the leaves of the shrub are wet.

3.8
It has been raining.The leaves of the shrub are wet.Finally, the ordering of the causal members may be varied (Altenberg 1984, 21).The cause may precede the result, as shown above, or the result may precede the cause.

3.9
The leaves of the shrub are wet because it has been raining Similarly, both the reason and the result clause may be marked for subordination.
3.10 It has been raining, so the leaves of the shrub are wet.Causal relations, besides having a wide variety of possible formal realisations, also bear affinity with several other types of semantic relationships that can hold between clauses.I will not go into detail on this point here, but the following examples are meant to serve as an illustration.
Causal vs. temporal relation (ambiguous) 3.11He found his fellow conscripts charming (his favourite adjective), and, as he listened to their plans for the future, he began to dwell on the possibility of making plans of his own.( ABl 3.14b They must have been flesh and blood, even though they lived God knows how long ago.(ABR1)

Reason clauses and information structure
Previous linguistic work on the use of causal links has shown that the information value of the reason clause influences both its position in relation to its associated main clause, and the choice of the introductory causal link itself (Altenberg 1984, 54-55, Vandepitte 1993, 87).Central in this respect is the perceived dichotomy between given and new information.
Several attempts have been made to arrive at a definition of given and new information.Usually, these terms have been defined in psychological and cognitive terms.Chafe focuses on the notion of consciousness.He claims that Given (or old) information is that knowledge which the speaker assumes to be in the consciousness of the addressee at the time of the utterance.So-called new information is what the speaker assumes he is introducing into the addressee's consciousness by what he says.(Chafe 1976, 30) Halliday takes a somewhat weaker stance.Given information does not have to be in the consciousness of the listener at the time of speaking; it is presented by the speaker/writer as being "recoverable" to the listener, either from the preceding discourse, or from the extralinguistic situation.Or it may be "something that is not around at all but that the speaker wants to present as given for rhetorical purposes" (Halliday 1994, 298).New, or non-recoverable, information may be previously unmentioned, or "something unexpected, whether previously mentioned or not" (Halliday 1994, 298).
The notion of recoverability is reminiscent of the idea of "manifestness" advanced by Sperber and Wilson: A fact is manifest to an individual at a given time if and only if he is capable at that time of representing it mentally and accepting its representation as true or probably true.

And:
To be manifest, then, is to be perceptible or inferable.(Sperber and Wilson 1986, 39) Vandepitte expounds on this idea of manifestness, claiming that the speaker/writer will assess to what extent a fact or assumption is manifest to the listener before making an utterance (1993,88).In other words, the givenness, or manifestness, of the facts presented in an utterance can be said to be listener-oriented, but speaker-selected (Halliday 1994, 298).
Although the representations of givenness and manifestness given by Halliday, Vandepitte and Sperber and Wilson seem similar in nature, they differ with respect to the entities to which their distinctions apply.
Certain writers, such as Chafe and Halliday, focus on the distinction between new and given elements within information units, i.e. the information value of lexical units like nouns and verbs.On the other hand, Sperber and Wilson, as well as Vandepitte, apply the distinction between new and given information to what they speak of as facts or assumptions, i.e. the proposition underlying a clause or sentence.In other words, in a sentence like / saw your father yesterday, the main point is not whether your father is assumed to be familiar to the listener, but whether the whole clause proposition saw your father yesterday can be considered new.The new information is not conveyed by the lexical units alone, but by the relationship between them (Krogsrud 1980, 22).
In the present study, I follow Vandepitte and Sperber and Wilson, and apply the distinction between given and new information to causal clause propositions, and not their lexical clause elements.

Position and information structure
Although various factors intervene in the linear arrangement of a text, most linguists agree that in its basic, or unmarked, form a text unfolds by proceeding sequentially from the known ('given') to the unknown ('new'), thus forming a chain in which what was unknown becomes the known point of departure towards a further unknown item.(Quirk et al. 1985(Quirk et al. ,1430) ) Quirk et al. refer to this tendency as the Principle of End Focus, noting that "it is common to process the information in a message so as to achieve a linear presentation from low to high information value" (1985,1357).A full treatment of this topic falls outside the scope of this study, but in short, the prediction can be made that, in most cases, given reason clauses will occur in initial position, and clauses presenting new information will be found in final position.

Direct and indirect reason
Clauses of reason usually express a direct causal relationship between the two causal members in the subordinate clause and the main clause: 3.15 My bank account is empty because I spent all my money.The reason clause because I spent all my money gives the direct reason for the phenomenon expressed in the main clause.Vandepitte refers to this as a formal causal relation, "in which the propositional forms of the sentences are causally related" (Vandepitte 1993, 115).
In contrast, reason clauses may also express an indirect causal relationship, where "the reason is not related to the situation in the matrix clause but is a motivation for the implicit speech act of the utterance" (Quirks/.1985, 1104): 3.16 My bank account is empty, because I checked it this morning.
Here the reason clause does not provide the actual reason why my bank account is empty.The bank account did not suddenly become empty just because I checked it.Rather, the reason clause gives the reason why the speaker/writer can know or say; that the bank account is empty.In other words, the propositional forms of the reason and the result clause are no longer causally related.
Vandepitte refers to indirect causal relations as an attitudinal causal relations, claiming that what is conveyed is the relation between a cause and an implicit expression of the speaker/writer's propositional attitude (Vandepitte 1993, 115).The speaker/writer's propositional attitude is "the extent to which a speaker regards a given state of affairs as true or desirable" (Hasselgård 1995, 125).In example 3.16, the reason clause justifies the speaker/writer's belief that his bank account is empty.By rewriting this as a direct causal relation, and thus explicitly expressing the speaker/writers's propositional attitude, we can bring the underlying causal relationship into the open: 3.17 / know that my bank account is empty because I checked it this morning.
Vandepitte divides the attitudinal causal relations into two categories, depending on whether the result clause contains an implicit expression of the speaker/writer's propositional attitude of belief (as in example 3.16), or an implicit expression of the speaker/writer's propositional attitude of desire.If the result clause contains an implicit expression of the speaker/writer's propositional attitude of desire, the speaker/writer either desires to know whether something is true or not, or he desires that somebody does something.In the first case, the result clause is normally interrogative, in the latter it is usually imperative (Vandepitte 1993, 125).These two types are illustrated in 3.18 and 3.19, respectively: 3.18 As you're in charge, where are the files on the new project?(Quirk tf al. 1985(Quirk tf al. , 1104) ) 3.19 Don't think about it, because it's over and done with.(TH1)

Analysis
In this part of the study I will present the findings of my investigation of the ENPC corpus material.First, the Norwegian correspondences of because, since and as will be presented and discussed, and then the English correspondences of fordi, siden and ettersom.Next, I turn to an investigation of information structure and indirect reason, as reflected in the material.

Correspondences
In order to reveal any influence the positioning of the reason clause might have on the choice of correspondences, each subordinator type has been treated separately in initial, medial and final position in the causal clause complex.
The correspondences have been categorised according to whether they retain the causal meaning of the original link (Causal) or not (Non-Causal).Causal links are subclassified into categories based on their formal properties.Non-causal links include cases where the causal link corresponds to a link from another semantic category (e.g.temporal; cf.3.1, above), as well as cases of zero correspondence, i.e. instances  1 surveys the correspondences of the English subordinators because, since and as in the Norwegian ENPC translations (horizontal axis), and indicates whether the original reason clause occurred in initial, medial and final position in the causal clause complex (vertical axis).Note that those instances where the part of the text containing the causal construction was left out of the translation are not included (three instances with because and one with since).
Because corresponds most frequendy to the causal subordinator fordi.
Including clause-integrated fordi-claxxses, because is rendered by fordi in 60% of the examples analysed.The predominance of fordi is not surprising, as it resembles because in many respects.Fordi is the most where there is no form in the translated text that corresponds to the link in the original text.(1997: 1140).This is confirmed by the ENPC results.All in all, the coordinator for occurs as a correspondence of because, since and as in 39 cases.On all occasions the original reason clause contains new information.
Teleman points out that the Swedish causal coordinator ty, normally replaced by fir in spoken language, is frequendy used with indirect reason (1976,(393)(394)(395).This seems to be the case with the Norwegian coordinator for as well.Of the 39 occurrences of for, 5, or 12.8%, are translations of indirect reason clauses.One example of this will be given here:

4.2a
Do you ever forget about them, the parents, for a moment?
They are always there in the hesitations -whether you will obey or defy -the opinions -where did you get them from?-that decide what you're doing.Because even while you defy the parents, deceive them, you believe in them.(NG1) 4.2b Glemmer man sine foreldre, om bare for et eneste øyeblikk?
De er alltid tilstede enten du velger å lystre eller å trosse, i meningene dine -for hvor har du fått dem fra -som avgjør ditt valg.For seiv når du trosser dine foreldre eller lurer dem, så har du tiltro til dem.In 4.2b, for corresponds to because in the original text.Besides illustrating the correspondence, the above example also shows how reason clauses in some cases can be given a global, text-structuring, scope.The because-clause, as well as its Norwegian counterpart, seems to qualify both the rhetorical question being asked in the first sentence of the example, and the opinion or belief being expressed in the following declarative sentence.
Since is most often rendered by the causal subordinator siden, although there is more variation in the types corresponding to since than was found with because.Siden occurs in 42.6% of the translated examples.It is not surprising that siden is the most frequent single correspondence of initial since; they are similar in form, and correspond closely as markers of temporal subordination as well.Moreover, siden appears to be preferred as the Norwegian correspondence of since with indirect reason:
As displays a highly varied pattern of correspondence.Siden and ettersom are the two dominant correspondence types, both occurring in 25% of the Norwegian translations.It should be noted though that, as correspondences of as, siden is predominantly used in fiction texts (in 12 of the 14 occurrences), whereas ettersom mainly occurs in non-fiction texts (in 11 of 14 cases).This might suggest that ettersom is somewhat more formal than siden.Compare: I couldn't forewarn anyone as I didn't know the names, let alone the phone numbers, of the people who worked for him.(DF1) Jeg kunne ikke varsle noen på forhånd, siden jeg ikke visste navnene på noen av dem som arbeidet for ham, og slett ikke telefonnumrene deres.
Pino said that he was not in a position to give details as he was not an official spokesman for the bank.(LT1) Pino sa at han ikke hadde anledning til å oppgi detaljer ettersom han ikke var offisiell talsmann for banken.
The causal subordinator da is used on several occasions (12.5%), and there were also a number of occurrences oifor (10.7%).
These are the main correspondence types found in the material.There is a lot to be said about the other, and less frequent types of correspondence, but there will be room for only a few examples here.
In the first example, the causal subordinator because has been rendered by a prepositional construction in the Norwegian translation, and the propositional meaning conveyed by the original ^cvzz^-clause has been encapsulated in a nominal £#-clause:

4.6a
The rich were incensed because they did not get their money, the poor were incensed because they got no land.(JH1) 4.6b De rike var opprørt over at de ikke fikk pengene sine, de fattige var opprørt over at de ikke fikk noe jord.In the original sentence, the information conveyed in the becauseclause is already given in the preceding context.The nominalised clause seems to make the underlying proposition more noun-like, like a thing or an already established fact.It thus appears that the givenness of the original because-clauses is emphasised more strongly in the Norwegian texts.A translation with an adverbial fordi-clause would also be possible.

4.5b
The close connection between reason clauses and other adverbial clause types is seen clearly in some of the correspondences that fall within the "Other link" category.In 4.7 the since-clause is probably ambiguous, and can be interpreted either as a temporal clause or, as is more likely, a reason clause.The corresponding Norwegian expression etter at introduces a temporal clause, although a causal relationship between the two clauses is still felt to be present.

4.7a
Since his wife had started going to evening classes, he preferred to prepare something for himself.(ABl)
Turning now to the Norwegian subordinators, table 2 presents the correspondences of fordi, siden and ettersom in the English ENPC translations.Three untranslated examples with fordi and one with siden are not included.Overall, there is less variation in the correspondence types than with the English subordinators.Fordi corresponds to a very high degree to because in the translated texts, with 71 occurrences.If we add to that a number of prepositional {because of) and clauseintegrated expressions {this is because), different forms of because make up 81 % of the total correspondences of fordi.
In addition to because, a number of other subordinating expressions were found as correspondences of fordi, although none were very frequent.Among these are as (1.1%) and since (2.2%).Not that is used in two cases when the originaly*Wz'-clause is negated:

4.8a
Med Bente herself caused a fuss -she's the sort who can live out of a knapsack and thrive.The small category of prepositional linking consists of three instances where fordi is rendered by the prepositional phrase because of In all cases, the information conveyed by a full clause in the original version {hun er syk) has been condensed into a noun phrase in the translation {illness):
(SU) 4.9b If the mother has to stay at home from work because of illness during her pregnancy, this is regarded as normal absence due to illness, and must not be reckoned as part of the leave entitlement in connection with the birth.
Interestingly, in the "Other link" category, a prepositional manner expression was also found as a correspondence of final fordi.
Here, it seems, the manner expression renders explicit an aspect of the meaning of the original reason clause, which there is only implied.The fact that the subordinator fordi could easily have been replaced by the manner expression ved at shows just how closely these interpretations are related.
4.10a De besehet ofte en overlegen hende fordi de brukte list. (SHI) 4.10b They often defeated a superior enemy by their cunning.Siden is predominantly rendered by since (42.6%) in the English texts.However, bearing in mind the correspondence pattern found with the English subordinators, it is not surprising that as occurs in a number of translations as well (22.7%).The subordinator in that is used in the next example: 4.11a Nerdrums Hermafroditt fra 1992 er i saerlig grad av interesse siden han her så direkte utformer en helhetsog fullkommenhetssymbolikk. (JEEH1) 4.11b Nerdrum's Hermaphrodite from 1992 is especially interesting in that he so directly develops a symbolism of wholeness and perfection here.Besides designating a reason clause, in that signals the point of view taken by the speaker (Quirk et al. 1985(Quirk et al. , 1105)).That is, Nerdrum's Hermaphrodite from 1992 is especially interesting from the point of view that he so direcdy develops a symbolism and perfection here.Some of the same meaning appears to be conveyed by the adverb phrase i saerlig grad in the Norwegian original.
As for the "Other link" type, three conditional correspondences were found in the English translations of Norwegian siden.In all cases the original siden-chaise introduces a premise for the conclusion reached in the following resultive main clause.In example 4.12a, the necessary connection between cause and result is reinforced by the addition of sä as a marker of the resultive clause.plan since the full, initial quota of 3,500 subscribers was reached quickly.In one instance, there is no overt causal marker in the translation.Instead, the causal relation is structurally marked by means of a supplementive z>zg-clause in medial position.

4.12a
Since they contain no overt marker of the semantic relationship expressed, clauses of this type can be used in a variety of functions, and the specific meaning of each clause must be interpreted from its context.However, when the zVzg-participle verb is realised by a stative verb, as in the following example, the clause usually has a causal function (Quirk et al. 1985(Quirk et al. ,1124)).A final point should be made as regards the structural stability of the original construction in translation.This is illustrated in table 3.Although there is some variation, the syntactic realisation of the causal relationship is very stable across the languages.The formal difference between subordination and coordination seems to be of little importance here.It is also interesting to note that the linear ordering of the causal members found in the original texts has been preserved in an overwhelming majority of cases.
Reinforcing this image of cross-linguistic structural stability is the fact that the causal relationship as such is retained in all but a few instances.Even though a non-causal link occurs sporadically in the translated text, the causal relation is often implied even where another semantic relationship is explicitly marked.Presumably, this reflects both the importance of the causal relation in the successful processing of the text as a coherent whole, and the translators' desire to remain faithful to the content of the original text.

Information structure
The pattern of correspondences found for each link type is related to the typical position and information structure of the clauses it introduces.All original reason clauses have been categorised as given or new, taking into account both linguistic and extralinguistic information.Clauses are said to be given when they contain information that can be derived from the preceding context of the clause, either through literal repetition or paraphrase of the preceding discourse, or through implication.In addition, clauses containing situational reference, information which is shared by persons in a particular situation, or information which is assumed to be general knowledge, have been classified as given.Table 4 correlates the English causal link types with the position and information status of the reason clauses examined.As table 4 shows, because-clauses occur almost exclusively in final position (93%).Since-and ^-clauses show a greater propensity for initial position (39.1% and 37.5%, respectively).As for information structure, there is a relatively clear disparity between the predominantly new because-clauses (90%) and the more frequently given since-and ^-clauses (44.3% and 28.6% given, respectively).Altenberg, who observed the same positional tendencies in material from the London-Lund Corpus of Spoken English and the Lancaster-Oslo/Bergen Corpus of British English, attributes the differences in placement to a principle of thematic ordering.Because-clauses are usually asserted, and present new information.As a consequence, they are naturally more frequent in final, focused or rhematic position.Since-and as-clauses, on the other hand, typically present information that is presupposed, i.e., assumed to be known to the addressee.They therefore show a greater preference for initial placement (39.1% and 37.5%, respectively), and function as a given premise to the conclusion in the following main clause (Altenberg 1984, 54).This point is also emphasised by Krogsrud, who claims that preposed because-, since-and AEf-clauses "provide information that is assumed to be either known/given to the addressee or to have less information value than the subsequent main clause" (Krogsrud 1980, 77), thereby providing a background for the more salient main clause which follows.The correlation between position and information structure is very clear in the ENPC material.Overall, 39 out of 71 (54.8%) initial reason clauses are given, compared to only 36 of 159 (18.5%) final clauses.
Following the same procedure, all original Norwegian reason clauses in the material were marked for position and information status.The results of this survey are presented in table 5. We note that jvrdi-clauses are predominantly found in final position (93%), while ettersom-cXanses, and particularly siden-cXaxxses, are much more frequent in initial position (41.7% and 68.2%, respectively).
The pattern of placement of the Norwegian causal links resembles the pattern found with the English causal links in table 4. Fordi and because are strongly right-tending (both 93% final), and ettersom shows roughly the same positional tendencies as since and as (39.1% and 37.5% initial, respectively).
What is interesting to note, however, is the strong inclination towards initial position found with siden-cXauses compared to the English subordinators.Related to this is a notable tendency for siden to introduce given information.Of the 44 occurrences of causal sidenclauses in the corpus, 26 (59.1%) are given.This is considerably higher than the corresponding figures found for since and as (443% and 28.6%, respectively).A possible interpretation would be that siden has a more restricted function within the set of Norwegian causal subordinators than is the case with since and as in English, and that it is more specialised towards a thematic anchoring function within the causal clause complex.
A point should also be made as regards the information status of ettersom-and ^rafz'-clauses.Ettersom shows largely the same pattern as since and as, introducing given material in 41.7% of the examples analysed.
Fordi-cXauses, although predominantly new, still have a higher propensity for givenness (21%) than was found with because (10%).As was shown above, the Norwegian causal coordinator for is often found as a correspondence of because in clauses with new information, and it is regularly employed as a causal link in Norwegian.In an investigation of spoken Norwegian, Guldal found four times as many occurrences of for (591) as of fordi (141) (Guldal 1977, 162).The fact that as high a proportion as 21% of all fordiclauses found in the corpus convey given information may indicate that for is preferred in place of fordi in many instances when the reason clause introduces new material.

Indirect reason
Following the method proposed by Vandepitte (1993, 120), all original examples in the material have been tested in order to see whether they express direct or indirect reason.For each causal relation, the following question was asked: Why + proposition expressed in the result clause?
If the question could not be answered by means of the reason clause, the propositional forms of the reason and the result clause were considered not to be causally related, and the causal relation was classified as indirect.Furthermore, all indirect clauses were subdivided according to whether the result clause was found to express an implicit propositional attitude of belief, or an implicit propositional attitude of desire (cf.3.4 above).
During this investigation, a number of problems, both expected and unexpected, cropped up.Firstly, a number of examples proved to be impossible to classify.This was only to be expected, as Vandepitte had the same experience during her investigation (1993,122).Of more concern was the fact that a number of occurrences, although undoubtedly showing some sort of indirect reason, did not appear to fit into the categories proposed by Vandepitte.Consider the following example: 4.15 "And since you inquired on my wheretos and whereabouts..." "I did no such thing." He continued as if she hadn't spoken, "I'm on my way to the low ground to pick me some wild herbs.And then I plans to stop by the Morgans' sugar cane field near the levee."(GN1) This excerpt is taken from a stretch of dialogue between the two characters Butch and Mattie.Butch, who is the first speaker, uses an initial since-claxxse to justify the main clause, which follows after an interruption by Maggie.Here, the since-clause does not provide the reason for the claim expressed in the related clause.The reason why Butch is on his way to the low ground is not that Mattie has inquired on his wheretos and whereabouts, which she even denies having done.The since-clzuse does not express direct reason, but it does not appear to be related to an implicit propositional attitude in the main clause either.In fact, it seems totally unrelated to the content of the second clause.Rather, the reason clause in this example serves to justify the fact that Butch is speaking at all.For another example of this type, see example 4.3 above.
All in all, six examples of this type were found in the material, and all in the English originals.Four are introduced by since, and one each with because and as.In this study, the examples of this type have been included among the indirect reason clauses and are referred to as the Utterance type.
In all, 37 occurrences of indirect reason were found.Ten occurrences were left undecided.The results of the survey are presented in table 6.Since seems to be the preferred subordinator with indirect reason in English, with 19 instances.Of the 115 -rø^-clauses in the corpus, 16.5% express indirect reason.It is interesting to note, though, that all the English subordinators can be used to express all types of indirect reason.
On the whole, the number of indirect reason clauses found is lower in the Norwegian material (5.8% of all clauses) compared to the English texts (10.3%).No occurrences of the Utterance type were found, and, interestingly, neither were any occurrences with fordi.Ettersom and siden are the only Norwegian subordinators used with indirect reason in the material.The fact that as many as one third of the occurrences of ettersom in the corpus convey indirect reason suggests that this may be one of the main functions of this causal subordinator.In this respect, ettersom resembles the Swedish subordinator eftersom.Teleman et al. claim: "För att ange talarens skäl till att anta något eller till att utföra en viss språkhandling används hellre eftersom än därför att"* (Teleman et al. 1999, 630).
Only siden is used to introduce a causal relation where the result is an implicit propositional attitude of desire.All types found are illustrated below.The result is an implicit propositional attitude of belief: 4.16 The system is not entirely failsafe, as ichthyologist Ted Pietsch recently found a male of one species attached to a female of a different species -a fatal mistake in evolutionary terms [...].(SJG1) 4.17 Har man store lese-og skriveproblemer, så utfordrer jo dette oppfinnsomheten, og ettersom Leonardo i høy grad var en oppfinner, klarte han vel etterhvert å takle sine rent praktiske problemer.(ANR1) The result is an implicit propositional attitude of desire: The result is a justification of an utterance act: 4.20 "If you wants the bus depot, you walkin' in the wrong direction, 'cause nobody in their right mind would be trying to walk to the train station."(GN1) Fordi is not found with indirect reason clauses in the corpus, not even in translation.This substantiates the claim made by Guldal that, whereas j^r-clauses can express both direct and indirect reason, fordiclauses are only used to express direct reason.By way of illustration, replacing for by fordi in the following examples of indirect reason would not be possible (both quoted from Guldal 1988, 137): 4.21 Da var vi i marka.En av markene.For det var i grunnen like mye Østmarka som Vestmarka og Nordmarka.

Conclusion
The causal clauses analysed show a remarkable structural stability in translation.In the great majority of cases, the translation preserves the syntactic construction, the order of the causal members and the causal relation.There are, however, some notable differences between the causal links in the two languages.Figures 1 and 2 attempt to capture the main correspondences graphically.The figures help to illustrate what appears to be two somewhat different paradigms of causal conjunction, as the bulk of the causal constructions are realised by a larger set of causal conjunctions in the Norwegian translations than in the English ones.This certainly seems to give the Norwegian language-user a wider range of stylistic alternatives, but presumably also leads to a more restricted or specialised function for each of the Norwegian link types.This is seen particularly with siden, which is more specialised as an initial marker of background information than since, and with fordi, which, due to the availability of the causal coordinator for, is more restricted than  As for the English causal subordinators, the limited number of unmarked alternatives gives each of the three link types a somewhat wider range of typical uses.It should be noted, though, that the total number of subordinating link types found actually exceeded the number of subordinator types in the Norwegian translations.A number of types occurred in addition to because, since and as: not that, that, now that, in that and with.However, their low frequency suggests that their use is limited, and they all appear to be stylistically marked, often blending the meaning of circumstantial reason with another semantic relationship (cf. example 4.11).In future work it would be desirable to extend the comparison to the full range of causal connectors in the two languages.

4
.18 I said, "Could you ask him to leave me a prescription somewhere, because I've fallen on my ankle and twisted it, and I'm running out of Distalgesic."(DF1) 4.19 Hva var det nå med dette kvinnemennesket som skulle ha den, siden tante Linn tok slik på vei? (KF2) because.

Figure
Figure 1.Norwegian correspondences ø/because, since andas Because ^^^^^^^^^^^^^^^ förf The present study is limited to causal relations explicitly realised by one of the subordinators because, since, as, fordi, siden and ettersom.

Table 1 . Norwegian correspondences of because, since and as
causal subordinator in Norwegian, it is stylistically unmarked, and shows roughly the same tendencies as because as regards position and information status (see below).It is therefore the most likely candidate for translation equivalence with because.In 4.1, the original clause-Jammen jeg skjønner ikke hvorfor det er så nyttig med sagflis når du seiger brukte biler, pappa.-Det er fordi du er en uvitende liten tufs, svarte faren.The only other sizable correspondence type with because is the causal coordinator for, which occurs in 20% of the Norwegian translations.Faarlund et al. note that ^ør-clauses, in contrast to clauses introduced by fordi, display a main clause word order pattern, and normally only contain new information frequent
Ikke fordi Bente selv laget bråk -hun hører til dem som kan bo i en ryggsekk og trives.(EHA1) 4.8b With Bente there a new life began, refreshing, varied, strenuous -all at once.It was as if there were always a gust blowing through an open window.Not that Da han kom inn i stuen igjen, hørte han fru Liens litt barnslige stemme: "Siden politiet er blandet opp i dette, så må dere vel tro at stakkars Brårud er død?" Ettersom is the least frequent of the causal links investigated, with only 12 occurrences in the Norwegian originals.In translation, ettersom is most often rendered by since (5 examples), while as is found in 2 examples.In 4.13, the causal clause expresses indirect reason (belief).

Table 5 .
Position and information status of fordi-, siden-and ettersomclauses in the ENPC